key: cord-0752547-iccs9wzr
authors: Cao, Hongyan; Li, Bin; Jiang, Xubao; Zhu, Xiaoli; Zheng Kong, Xiang
title: Fluorescent linear polyurea based on toluene diisocyanate: easy preparation, broad emission and potential applications
date: 2020-06-13
journal: Chem Eng J
DOI: 10.1016/j.cej.2020.125867
sha: 18527a0c5d7765f40c616c340a394d4ee166b782
doc_id: 752547
cord_uid: iccs9wzr

In contrast to conventional fluorescent polymers featured by large conjugation structures, a new class of fluorescent polymers without above conjugations are gaining constant interest owing to their significant academic importance and promising applications in diverse fields. These unconventional fluorescent polymers are in general composed of heteroatoms (e.g. N, O, P, and S) under different forms. Here we report our recent study on polyurea, prepared by a very simple one step precipitation polymerization of toluene diisocyanate in a binary solvent of water-acetone. This polyurea, basically consisting of phenyl ring and urea group, shows fluorescent emission in a broad concentration range, from very low (10(-5) mg/mL) to its solubility limit (50 mg/mL), and in a wide range of emission wavelength from UV to visible regions of up to 500 nm under varied excitation wavelength. The emission behaviors were fully studied under different concentrations and excitations. It was concluded that the emission in UV region was intrinsic due to the conjugation between the phenyl and the adjacent urea unit; while the emission in visible region, strongly excitation dependent, was caused by the cluster formation of the molecular chains, in accordance with the cluster-triggered-emission (CTE) mechanism. The formation of the cluster was tested through dynamic light scattering, FTIR and UV absorbance. Tested in presence of different metal ions, Fe(3+) demonstrated a quenching effect with high selectivity. Based on this study, different paper-based sensors were designed to detect Fe(3+), H(2)O(2) in bioanalysis and for data encryption. This work provides a simple way to prepare a polyurea, a novel type of unconventional fluorescent polymer, with high emission performance distinct from its known analogues.

2-phenethyl isocyanate with 2-phenylethyl amine) (Fig. S2 ). All syntheses were done at 30 ºC with equivalent amounts for all the starting reagents. Taking the synthesis of the compound M3 as an example: n-butylamine (0.53 g, 7.2 mmol) was dissolved in acetonitrile (28. 58 g) at 30 ºC, and p-toluene isocyanate (0.96 g, 7.2 mmol) was added in under stirring. The reaction was allowed to proceed for 4 h. The product M3 was seen to precipitate out during the reaction. At end of the synthesis, the reaction mixture was filtrated to separate out the product, which was dried under vacuum at room temperature. The synthesis protocol for the other three compounds was the same, except for M2 where a binary solvent of water/acetonitrile (3/7 by vol.) was used instead of pure acetonitrile.

Chemical structures of the four organic model compounds (M1 to M4) were characterized by 1 

The reactions involved in the synthesis of polyurea TPU, by dissolving TDI in the binary mixture of water-acetone, are well known [24, 28, 29] : in a first step, TDI reacts with water to yield toluene diamine (TDA) with the release of CO 2 , the in-situ formed TDA reacts then with TDI to form TPU (Fig. S1 ). TPU as prepared was characterized in previous studies, which showed that this polyurea was a typical porous material, with the specific surface area of 161.68 m 2 /g, the pore volume of 2.20 cm 3 /g. The pore size distribution, determined by Hg intrusion and BET tests, was quite broad, with the pore size varied from a few nanometers to dozens of micrometers [29] . TPU was also used for enzyme immobilization and kinetic resolution of racemic molecules [24] .

Under mercury lamp (λ ex =365 nm), the white TPU powder exhibited strong blue emission at room temperature (Fig. 1A) . From its excitation spectrum with emission λ em =440 nm, a maximal excitation, λ ex =370 nm, was indeed confirmed (Fig. S3A ). In addition, another emission at λ em =355 nm was also detected at excitation λ ex =320 nm (Fig. S3B ).

( Fig. 1.) Under irradiation of λ ex =365 nm, TPU solution in DMSO exhibited blue emission, and the emission was enhanced with TPU concentration (Fig. 1B) . With different excitation wavelength λ ex , TPU solution was emissive from 10 -4 to 20 mg/mL, although the emission intensity and its maximal emission wavelength changed with its concentration (Fig. 2) . Under excitation of λ ex =300 nm, the maximal emission of TPU was at 355 nm, detected from a concentration of as low as 10 -5 mg/mL ( Fig. 2A) , followed by a sharp increase in emission intensity with the concentration; The emission reached its maximal intensity at 0.05 mg/mL, and declined quickly with the increase in concentration and became non-emissive at 0.4 8 mg/mL ( Fig. 2A, D) . With the excitation wavelength increased to λ ex =320 nm, the maximal emission of TPU remained at 355 nm (Fig. 2B) . However, the emission was detectable from 0.01 mg/mL to 5.0 mg/mL, with the maximal intensity observed at 1.0 mg/mL (Fig. 2B, D) .

That is to say that, with λ ex increased from 300 nm to 320 nm, both the concentration to have the highest emission intensity and the concentration range to be emissive were dramatically changed, though the maximal emission was kept at λ em =355 nm, the same as λ ex =300 nm ( Fig.   2A ). With the excitation wavelength increased further to λ ex =350 nm (Fig. 2C) , the emission at 355 nm disappeared, and instead a new emission appeared at around λ em =440 nm. In addition, this emission was undetectable at TPU concentration of 0.2 mg/mL and below, started to appear at 0.4 mg/mL with λ em =438 nm. And a slight red-shift of the emission with concentration was also seen: the maximal emission shifted to λ em =440 nm at concentration of 2.0 mg/mL, and it shifted further to λ em =444 nm with the concentration increased to 20 mg/mL (Table S1 ). It is highly important to note that, this emission under excitation at λ ex =350 nm was quite different from that at lower excitation (λ em =355 nm with λ ex at 300 & 320 nm): Not only the emission red-shifted with concentration, but also the absence of the upper limit for TPU concentration vis-à-vis the emission intensity, which was linearly increased with TPU concentration increase up to 50 mg/mL. Above this concentration, emission intensity was practically constant owing to the solubility limit. This was a strong indication that the emission center or species at λ ex =350 nm must be different from that at excitation with lower wavelength, i.e. with λ ex =300 and 320 nm.

These results revealed that the emission of TPU solution was significantly dependent on its concentration and excitation. Under short wavelength excitation (λ ex ≤320 nm), it was emissive in UV zone (λ em =355 nm), the emission was well seen at low concentration ( ≤0.1 mg/mL), and quickly faded out with increased concentration or excitation wavelength ( Fig.   2A, B) . Under longer wavelength excitation (λ ex =350 nm) of a TPU solution with higher concentration (≥0.4 mg/mL), the emission shifted to a higher wavelength in the visible region (λ em around 440 nm). By increasing TPU concentration, the emission was enhanced (Fig. 2C) rather than the inverse observed under shorter wavelength excitation (λ ex ≤320 nm) and at lower concentration ( ≤ 0.1 mg/mL). Taking into account of these observations, a TPU solution of 5.0 mg/mL was made, a concentration where emission was detectable under excitation with λ ex at 320 and 350 nm, and its emission was observed under different excitation from λ ex =320 to 440 nm. From the results given in Fig. 3 , it is seen that, under excitation of λ ex =320 nm, only one single emission was seen in UV zone at λ em =355 nm; with excitation at λ ex =325 nm, the emission at 355 nm was intensified, and a new broad and weak emission appeared in the visible region at λ em =418 nm; with the excitation wavelength increased to λ ex =330 nm, both the two emissions were distinct, the one at 355 nm was attenuated and that in the visible region (424 nm) boosted up; by shifting further the excitation wavelength to λ ex =340 nm, the emission at λ em =355 nm practically disappeared, leaving one single emission in the visible region at λ em =428 nm. This emission in visible region was in constant red-shifting with increased wavelength of the excitation. By changing excitation from λ ex =330 nm to 440 nm, the emission red-shifted from λ em =424 nm to 512 nm (Table S2, Fig. 3 ), an obvious excitation dependent emission (EDE). It is worth noting that a highest intensity of emission (λ em =438 nm) was observed at excitation of λ ex =360 nm, the emission intensity was attenuated with the excitation going to either direction from the mission under excitation λ ex =360 nm. (Fig. 3.) 

The above results demonstrated clearly that, under short excitation wavelength (λ ex ≤320 nm), TPU emitted in UV zone (λ em =355 nm), the position of the emission maximum was independent of the excitation wavelength, and the emission disappeared at a given higher concentration. With a longer excitation (λ ex ≥340 nm), TPU emission red-shifted to visible region (λ em ≥428 nm), and the emission was of typical EDE. Based on the studies on fluorescent polymers [5, 6, 10] , we strongly suggest that the emission of TPU in UV zone and that in the visible region came from different emission centers, or these two emissions must be ascribed to different forms of TPU chains. At concentration of 5.0 mg/mL, TPU chains were assumingly present as fully dissolved on one hand, and they might also form clusters on the other hand. It is conjectured that the emission in UV region at λ em =355 nm was caused by TPU chains individually dissolved, while the emission in visible region at λ em ≥428 nm was owing to the clustering of TPU chains, suggesting that the EDE in visible region was typical CTE emission as afore-mentioned [20, 21] .

As mentioned above, the emission of TPU in its solution ought to come from the molecules themselves when they are dissolved individually in DMSO. Originated from TDI, phenyl rings are present in TPU chains (Fig. S1 ). Independent benzene ring is believed either 11 non-emissive or to be emissive at λ em =280 nm with very low intensity [30, 31] . TPU did not show any emission at 280 nm, and emitted at λ em =355 nm instead. Benzene ring was reportedly red-shifted through two different ways [30] : One is their π-π stacking in aggregates or crystals, which was clearly not happening in TPU solution; the other is electron delocalization in conjugated structure which is unlikely from TPU structure.

In fact, conjugation or sharing of electrons between heteroatoms and phenyl ring have been proposed for different molecules [30] [31] [32] [33] [34] , in which at least one heteroatom was directly bonded to phenyl ring, including phenol, aniline and 4-(N,N′-dimethylamino)benzonitrile. It is suggested that, the electron donor substituents, such as OH and NH 2 , contribute to increase electron density in π conjugation of a phenyl ring, and this is achieved by twisted intramolecular configuration of the substituent when it is out of the plane of the phenyl ring [32, 33] . And the fluorescence behavior of these molecules is affected by the polarity of the solvent used. As to TPU, it has two urea units directly attached to phenyl ring through N atoms, of which the electrons are mostly in hybrid orbital of sp 2 [31] , conjugation is believed to occur between the lone-pair electrons of N atom and the π electrons on the phenyl ring, reducing the energy required for their excitation and red-shifting therefore the fluorescent emission [35, 36] . Seeing that each phenyl is substituted by two urea units, and their N atoms are separated by a carbonyl, it is therefore highly likely that the π electrons of the carbonyl are also conjugated, making the whole TPU chain like a conjugated one. This might be the reason for the emission at λ Em =355 nm, largely red-shifted in comparison with that around 280 nm expected for a bare phenyl ring [30, 31] .

To support to this conclusion, four model compounds were synthesized (Fig. S2) , and denoted as M1, M2, M3 and M4. Their structures were checked and confirmed by 1 H NMR ( Fig. S4 ). All the four model molecules are consisted of phenyl ring(s) and urea unit(s) but connected in different ways (Fig. 4) . From a view point of likelihood to conjugate for the phenyl ring, the similarity of these model molecules to TPU is summarized as follows: The phenyl ring of M1 is connected to two urea groups, a structure the closest to TPU; For M2, each of its phenyl rings has only one urea attached, which is connected to another methyl-substituted phenyl ring, the environment similarity of the phenyl ring to that of TPU is lessened compared to M1; For M3, the similarity of the chemical environment to TPU is further enlarged because one side of urea is replaced by a butyl; As to M4, the chemical environment for its phenyl ring is completely different from that in TPU, because of the insertion of ethylene spacer between phenyl rings and urea group, their possible conjugation, if any, must be disrupted. Instead, the chemical environment for its phenyl rings must be the most similar to that of benzene in comparison with M1 to M3. Each of the four compounds was dissolved in DMSO to have their solutions (0.01 mg/mL), and UV absorbance was measured for each, along with benzene and TPU (Fig. S5 ). From their absorbance bands (Table S3) , it was seen that only one single absorbance band at 261 nm was present for benzene, attributed to its π-π* transition [37, 38] , while there were two bands for TPU at 256 and 300 nm, attributed respectively to, π-π* and n-π* transitions of its phenyl ring, owing to conjugation with N atoms and even the carbonyl double bond [35] . In comparison, all the four model compounds exhibited an absorbance band around 260 nm in accordance with that of benzene (261 nm). Besides this band, other bands at longer wavelength were also observed for all the model compounds: an absorbance at 300 nm (with a shoulder at 287 nm) for M1 and at 296 nm for M2, in accordance with the absorbance at 300 nm for M1. For M3, a number of shoulder bands were seen from 271 to 296 nm; and for M4, this band shifted to 270 nm, the most separated from that of TPU (300 nm), in accordance with the distinct difference in the chemical environment of its phenyl ring from that of TPU. These results indicate that the absorbance bands of the model compounds were more similar to those of TPU when the chemical environment in the molecules was closer to that in TPU.

Based on the results, under the same excitation (λ ex =300 nm) as for TPU solution to exhibit emission in UV region, emission behavior of the model molecules were tested in their DMSO solution at different concentrations, including M4 despite its very low absorbance as shown above (Fig. S5) . Briefly, all the three compounds (M1, M2 & M3) were more emissive than M4 (Fig. S6 ). For M1, the maximal emission appeared at λ em =323 nm (Fig. S6A) , and was unchanged with concentration; M3 exhibited practically the same emission behaviors, with its maximal emission equally at λ em =323 nm, unchanged with regard to its concentration. For both, their maximal emission intensities appeared at 1.0 mg/mL (Fig. S6 ). In contrast, M2 behaved slightly different in two ways (Fig. S6B ): the maximal emission red-shifted to 352 nm, much closer to that of TPU, and the maximal intensity occurred at 2.0 mg/mL, instead of at 1.0 mg/mL in M1 and M3. This emission similarity of M2 to TPU might indicate the emission center was the urea unit rather than the phenyl. All these properties were very similar to TPU as shown above ( Fig. 2A, B) , except the maximal emission intensity (λ em =323 nm for M1, M3; λ em =352 nm for M2), which was different from TPU (λ em =355 nm). This may be a good support for the presence of extended conjugation of the phenyl ring to carbonyl through N atoms of its urea unit, making a more distinct red-shift for TPU emission.

In addition, M4 emission was also examined at different concentrations. A very weak emission was observed even under optimal conditions, the maximal intensity was about 1/10 of those observed for M1 and M3 (Fig. S6D ). This was highly expected because, different from TPU and the other three model molecules, the phenyl rings in M4 were separated from its urea unit by an ethylene spacer, the conjugation of the phenyl ring with the urea unit was completely cut off. Through the comparisons of TPU structure, its absorbance and emission behaviors with those of the model compounds, it was concluded that, the emission of TPU in UV region (λ em =355 nm) was caused by its phenyl ring conjugated with the adjacent N atoms and the carbonyl groups of its urea units.

As shown in Fig. 3 and Table S2 , under excitation of different wavelength (λ ex ≥330 nm), TPU was highly emissive in visible region (λ em ≥424 nm), with obvious EDE behavior. We suggest that this emission belonged most likely to the afore-mentioned CTE mechanism, i.e.

owing to the formation of clusters upon aggregation of heteroatoms on the polymer chains [4] [5] [6] 21] . In the clusters, the electron clouds of these so-called unconventional or intrinsically fluorescent groups are overlapped, offering therefore the molecules with extended conjugations (lowered energy gaps) and rigidified conformations. This type of emission and its intensity are usually dependent on the size of the cluster [5, 6] , leading to EDE behavior because the cluster size is concentration dependent. EDE has been well previously reported, mainly on carbon-based nanomaterials [39] [40] [41] [42] [43] [44] , commonly featured by their C atoms with hybrid orbital of sp3 or sp2, and chemical groups with O atoms [41, 43] . Size effect has been the widely accepted theory for this behavior [41] [42] [43] , i.e. the energy band gap is determined by the size and size distribution of a carbon nanomaterial, and emission is determined by this band gap. In contrast, EDE has been reported in a lessened extent for fluorescent polymers [21, 23, [45] [46] [47] [48] [49] [50] . Among those reported, a great part has been focused on those without conventional chromophores [4] [5] [6] [45] [46] [47] [48] [49] [50] , namely non-traditional intrinsic luminescent polymers [5] or intrinsically fluorescent [4] . For examples, EDE behavior was reported on poly(amino ester) and attributed to the cluster formation owing to the ester and tertiary amine groups, making n-π* electron transition possible from the tertiary amino to aggregated C=O [48] ; Similar EDE was also reported on poly(maleic anhydride-alt-vinyl pyrrolidone), and was ascribed to the intra-and/or interchain interactions between N and O atoms as the result of the aggregation or clustering of the polymer chains [49] . Up to date, a huge number of studies have been reported on diverse polymers, which contain heteroatoms (N, O, S, P etc.) under the form of carbonyl, ester, hydroxyl, ether, amino, amide etc [20, 21, 27, 42, 45] . These polymers, of which the monomers are not fluorescent, do not contain conventional chromophores, and therefore do not emit at low concentration [20] ; or, they contain in some cases conventional chromophores, such as phenyl group, and are not emissive or show weak emission in UV zone only [42, 46, 47] . However, they do show high emission in visible region at a higher concentration, and the emission is usually EDE featured. This type of emission behavior has been observed and widely studied. A variety of interpretations have been offered by different authors in different case studies. The most accepted and widely accepted mechanism is the CTE [20, 21, 45] , proposed by Tang et al based on a study on pure oxygenic nonconjugated poly(maleic anhydride-alt-vinyl acetate) [20] , they suggested that the formation of a heterodox cluster of carbonyl groups was the reason for the polymer to be emissive under UV excitation, and that the absorption and emission shifts to longer wavelength was due to the solvatochromism, i.e. the formation of polymer-solvent complexes due to the interaction of the polymer with electron-rich solvents. Under CTE mechanism, the photophysical behaviors of many polymers with neither conventional chromophores nor conjugated structures have been well interpreted [21, 45, [49] [50] [51] .

Taking into accounts these analyses, the structure of TPU and its emission behavior described above (Fig. 2 & 3) , it is strongly believed that the emission in the visible region 

As described above, TPU emission in visible region was believed to be owing to cluster formation of TPU chains due to interaction of their functional groups. To support this conjecture, TPU solution in DMSO at different concentrations was prepared, and DLS test was conducted to explore the presence of the expected clusters [54] . The results revealed that clusters or aggregates were indeed occurring in TPU solution under certain circumstances.

While DLS test did not provide any particles at concentration of 0.05 mg/mL or lower, particles or clusters of very small size (~0.8 nm) were effectively detected in TPU solution at concentration of 0.1 mg/mL, and this size was in regular increase with TPU concentration, as displayed in Fig. 5 : the size of the clusters was increasing quite slowly at low TPU concentration, to 2.4 nm and 4.3 nm, respectively, at TPU concentration of 0.2 and 0.8 mg/mL; This increase in size (to 27 nm) was much sharper once TPU concentration increased to 1 mg/mL, and further to 83 nm with TPU concentration at 2 mg/mL. The size increase in the clusters became slower after the concentration reached 5 mg/mL (Fig. S7) , and an average size of 403 nm was obtained at TPU concentration of 20 mg/mL. Comparing to the emission spectra of TPU solution at different concentrations (Fig. 2) , it was found that the emission at 440 nm (under λ ex =350 nm) started to be significant at TPU concentration around 1 to 2 mg/mL, which was in good consistence with the detection of large-sized cluster of about 83 nm, suggesting strongly that this emission at 440 nm was owing to the cluster formation, in accordance with the CTE mechanism given in the section above.

It is easy to conceive that interactions between TPU chains, in particular the segments with electron rich groups, must be affected by the cluster formation, and reflected in their spectroscopic characterization such as FTIR [45, [55] [56] [57] . Following the size detection by DLS, TPU solutions of different concentrations were subjected to FTIR analysis (Fig. 6) . The absorbance of TPU carbonyl (C=O) appeared around 1660 cm -1 , slightly red-shifted owing to likely the presence of amides at its two sides [58] [59] [60] . However, a new absorbance band appeared at 1700 cm -1 with TPU concentration increased to 2.0 mg/mL (Fig. 6B) , and it was significantly enhanced with increased concentration. This was indicating that TPU chains were presented under a new form, with higher electron density around their carbonyl groups.

By cluster formation, it is assumed that the electrons on each group, including the carbonyls, the amide N atom and the phenyl rings, previously limited to each group, are shared among these groups in a cluster because of their conjugation and the clustering. The appearance of this absorbance at short wavelength (long wave number) at 1700 cm -1 seemed to indicate an increase in electron density around the carbonyls. As to the broad absorbance at 1525-1550 cm -1 , there were basically two absorbance bands like in polyurethane [45] and peptide [61] ( Fig. 6C) . The one around 1533 cm -1 is attributed to the bending vibration of N-H (amide II) in urea group and the other at 1548 cm -1 to the stretching vibration of the phenyl ring [62] . At low concentration (.2.0 mg/mL), the two were of about equal intensity. However, with increased concentration in TPU, the one at 1548 cm -1 was attenuated relative to that at 1533 cm -1 , and the later was obviously enhancing, indicating a clustering with the amide as the center. In addition, the peak around 1533 cm -1 was the result of a slight blue-shifting with increased TPU concentration, from 1528 cm -1 at 2.0 mg/mL to 1533 cm -1 at 20 mg/mL, an indication of electron density increase over urea groups owing to the cluster formation.

Another quite distinct change was also observed in the broad absorbance band from 1170 cm -1 to 1250 cm -1 (Fig. 6D) , attributed to the stretching vibrations of amide III, i.e. the C-N bonds in the urea unit [61, 62] . There was practically no absorbance at low concentration (0.1 mg/mL), and two absorbance bands started to appear at 2.0 mg/mL, indicating that, not only the carbonyl but also the adjacent C-N bond vibration was affected. It is to note that these changes in the absorbance bands became obvious only when TPU concentration attained 2.0 mg/mL, which might be considered as the sign of the cluster formation, in agreement with the emission behavior seen above (Table S1 ), and at this concentration TPU became obviously emissive with discernable EDE behavior.

It is well known that, for a material under irradiation, the electrons on the out-layer of its atoms excite to a higher energy orbital by absorbing photons. Emission occurs when the excited electrons return back to their initial state by release of the absorbed energy through light radiation. UV absorbance is therefore often used to explore the mechanism of emission [23, 63] . TPU solution in DMSO was subjected to UV absorbance. The results (Fig S5B) revealed that all the solutions, with concentration varied from 10 -3 mg/mL to 20 mg/mL, had UV absorbance bands, and that the maximal absorbance wavelength red-shifted with increased concentration. At concentration of 10 -3 mg/mL, only one sharp absorbance at 256 nm was present, indicating likely that TPU chains were assumingly well dissolved; with concentration increased to 10 -2 mg/mL, a new absorbance appeared around 300 nm, making TPU emissive at 355 nm as shown in Fig. 2A . It is to note that this new absorbance around 300 nm was boosted quite fast with increased TPU concentration, leading to enhanced emission as seen in Fig. 2A , owing to the conjugation between the π electrons of the carbonyl and the phenyl ring through N atom in between, as discussed in section 3.2. It is particularly interesting that, starting from the concentration at 2 mg/mL, another new absorbance at 350 nm was detected, suggesting that a new type of matter was created, i.e. the formation of TPU clusters. It was exactly this cluster formation, which led to the new and excitation dependent emission at around 440 nm, shown in Fig. 2C and Fig. 3A . All these results are in good agreement, and give substantial support to the CTE mechanism elaborated in section 3.2.2 based on cluster formation.

As supplementary information, it is to point out that polyurea is insoluble in common organic solvents, such as hexane, toluene, THF, acetone, pyridine and acetonitrile [64] .

Among a dozen of solvents tested, the highest TPU concentration one can make is about 3 mg/mL in acetic acid and in m-cresol, which renders the comparison of the solvent effect on TPU emission difficult. TPU is only soluble in DMSO, N,N-dimethyl formamide (DMF) and N-methyl-2-pyrrolidone (NMP), up to concentration at 20 mg/mL or higher. Fluorescent emission was tested in DMF and NMP at different TPU concentrations (0.05, 1.0 and 5.0 mg/mL), which showed exactly the same emission profile as those in DMSO (Fig. 3) . The emission spectra of TPU at 5.0 mg/mL are given as an exemplar (Fig. S8) . These results indicate that no discernable solvent effect was observed. This insensitivity of emission versus the tested solvents suggests the negligible solvent effect on the cluster structure, i.e. weak interaction between solvent and TPU chains, which is different from hyperbranched polysiloxane, where obvious changes of the emission intensity was observed in different solvents [65] .

To see the possibility for TPU as a sensor to detect metal ions, its emission was evaluated in the presence of a number of metal ions, each at concentration of 0.3 mM. Under UV lump 21 (λ ex =365 nm, Fig. S8 ), the emission, in the presence of Ag + , Ni 2+ , Ca 2+ , Pb 2+ , Cd 2+ and Cu 2+ , was practically the same as that in the pure solution of TPU showed in Fig. 1 , and in the presence of Cu + and Fe 2+ , the emission was slightly attenuated. In contrast, an obvious quenching effect was well seen in the presence of the ions Fe 3+ under excitation of λ ex =350 nm, in comparison with the rest of the tested ions (Fig. 7A) . And the emission practically disappeared when Fe 3+ attained 0.54 mM (Fig. 7B) . The co-presence of these tested ions at the same concentration (0.54 mM) as Fe 3+ showed no disturbance on its quenching effect. The competitive ions appeared to behave as in the cases when they were individually present, with the emission slightly changed as seen in Fig. 7A . And the emission was practically turned off once Fe 3+ was added along with the competitive ions, indicating high selectivity of Fe 3+ quenching towards TPU emission (Fig. S9 ).

( Fig. 7.) The quenching effect of Fe 3+ has been often studied for different fluorescent materials [65] [66] [67] [68] [69] . Different mechanisms have been proposed for the selectivity of Fe 3+ . Its larger charge/radius ratio is believed to be one possibility [65, 69] , which renders it with the strongest electron-withdrawing ability among all the ions tested. Another widely accepted mechanism is its paramagnetic nature, with its five d orbits half-filled under electronic structure of 3d5-4s0. Fe 3+ is also featured by its outstanding standard electrode potentials [66] [67] [68] , higher than all the other metal ions examined but Ag + (Table S4 ). However, given the atomic numbers of Fe and Ag and their charges (3 to 1), Fe 3+ has the highest charge density among all the metal ions used. All these combined, it is obvious that Fe 3+ has the strongest electron-withdrawing ability among commonly seen metal ions, which makes it a common quencher via different quenching mechanisms [70] .

The detection limit (DL) of Fe 3+ was determined using a reported method [71], which gives: DL=K×σ/S, where σ is the standard deviation of the blank samples from at least 20 measures (σ=0.218 in this work); S is analytical sensitivity, i.e. the slope of the measure signal (the emission intensity) versus the concentration of Fe 3+ (the absolute value is used for negative slope); and K, a numerical factor chosen in accordance with desired confidence level. k<3 should not be used for DL calculations, and use of k=3 is suggested to allows a confidence level of 99.9%. In accordance, the emission intensity of TPU in DMSO (10 mg/mL) was recorded as a function of Fe 3+ concentration (mM), and the results given in Fig.   8 . It is seen that the emission intensity was linearly decreasing with increasing [Fe 3+ ] in the concentration range of 0~0.15 mM (Fig. 8A ). The linearity curve in Fig. 8B (Table S5) , DL of this work is similar to or better than all the reported data. Taking into account of the easy preparation, it is quite obvious that TPU is a material of choice for Fe 3+ detection in view of its easy fabrication and low DL. (Fig. 8.) Based on the quenching effect of Fe 3+ on TPU emission, a paper strip was dipped into a solution of TPU in DMSO at 30 mg/mL, TPU remained on the surface of the paper after drying up. Under UV lump (365 nm), this paper strip showed strong blue color as shown by Fig. 9A . By drawing a letter "k" using a pipette filled with Fe 3+ aqueous solution (0.03 mM), 23 that letter turned non-emissive because of the quenching effect (Fig. 9B ). This can be used for easy Fe 3+ detection in aqueous system. (Fig. 9.) Based on the quenching of TPU by Fe 3+ , a security paper for data encryption can be made from an ordinary (no fluorescent) paper by dipping it in TPU solution. This TPU coated paper was blue luminescent under UV lump (365 nm, Fig. 10A ). By writing "COVID-19" on the paper with an aqueous solution of disodium salt of ethylene diamine tetraacetic acid (Na 2 EDTA, 10 mg/mL) as the ink, no any changed was observed by naked eye under daylight or UV irradiation (Fig. 10B) , which meant that the information was successfully encrypted.

However, "COVID-19" was clearly seen under UV lamp (Fig. 10C ) by brushing the security paper with an aqueous solution of FeCl 3 (1.0 mM). This is because Fe 3+ ions interacted in priority with Na 2 EDTA against TPU, protecting TPU fluorescence from quenching by Fe 3+ ions. FeCl 3 is used here as the key to decode the encrypted information on the security paper.

Therefore, TPU can be used as a tool to fabricate security paper for data encryption. Fig. 10A (the paper dipped in TPU without the presence of Fe 2+ ions). By writing any letters on this fluorescent strip using an aqueous solution of an oxidant (such as H 2 O 2 and tert-butyl hydroperoxide) as the ink, these letters were well readable under UV light because the paper covered by these letters became no fluorescent owing to Fe 3+ formation from Fe 2+ oxidation. This change in blue fluorescence was easily observable by naked-eyes with H 2 O 2 concentration at 0.04 mM (Fig. S10) . This is particularly interesting in biological detection of H 2 O 2 in human body, because H 2 O 2 is produced in metabolism process, and known as a precursor molecule of a number of reactive oxygen species, commonly known to cause oxidative damage on proteins and nucleic acids and to be correlated to cancer and Alzheimer's diseases [64] .

Polyurea, TPU, was synthesized through a very easy protocol of precipitation polymerization of TDI in a binary solvent of water-acetone, and was emissive as powder in its DMSO solution. In solution, TPU emission was covering a broad span from UV to visible region of up to 500 nm under different excitation. Under short wavelength excitation (λ ex ≤320 nm), the emission started from concentration of as low as 10 -5 mg/mL, passed a maximum at 355 nm, and faded off at higher concentration. The concentrations, where the emission started to appear, to maximize and to disappear, were closely dependent on the excitation. Under long wavelength excitation (λ ex ≥350 nm), the emission at 355 nm disappeared, and a new emission appeared at about λ em =440 nm, which is featured by constant increase in intensity with TPU concentration and constant red-shifting under excitation with extended wavelength.

Based on the results, it was proposed that the emission in UV region (λ em =355 nm) was ascribed to the chemical structure itself owing to an eventual conjugation between the phenyl ring and the urea carbonyl through n-π* transition of the N atoms in between, which was supported by emission behaviors of four organic urea model compounds synthesized for this study. As to the strongly excitation dependent emission appeared at about λ em =440 nm under λ ex ≥350 nm, it was attributed to the cluster formation of the TPU molecules, supported by DLS test for the presence of cluster particles, FTIR absorbance changes and UV absorbance at TPU concentration where this emission started to appear. By screening a number of metal ions, Fe 3+ was shown to have an obvious quenching effect on TPU emission. Simple paper strips were fabricated by coating TPU on their surface, followed by different treatments by aqueous solutions of Fe 3+ , Fe 2+ or Na 2 EDTA. These sensor paper strips were used for detections of Fe 3+ and H 2 O 2 in aqueous systems, and for data encryption.

aggregation-induced phosphorescence-active "turn-off" nanosensor based on ferric-specific quenching of luminescent and water-soluble Au(I)-cysteine nanocomplexes, Anal. Chem. 92 (2020) 6785-6791, DOI: 10.1021/acs.analchem.0c01358. 

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

• Toluene diisocyanate (TDI) based polyurea (TPU) is prepared by TDI reaction with 

The optoelectronic nose: Colorimetric and fluorometric sensor arrays

Electrochemical sensing based on carbon nanoparticles: A review

Facile synthesis of intrinsically photoluminescent hyperbranched polyethylenimine and its specific detection for copper ion

Non-traditional intrinsic luminescence: inexplicable blue fluorescence observed for dendrimers, macromolecules and small molecular structures lacking traditional/conventional luminophores

Clusterization-triggered emission: uncommon luminescence from common materials

Robust organic nanoparticles for noninvasive long-term fluorescence imaging

Synthesis of thermosensitive conjugated triblock copolymers by sequential click couplings for drug delivery and cell imaging

Highly fluorescent amidine/Schiff base dual-modified polyacrylonitrile nanoparticles for selective and sensitive detection of copper ions in living cells

An overview of nanoparticles commonly used in fluorescent bioimaging

Chromophore attachment to biliproteins: Specificity of PecE/PecF, a lyase-isomerase for the photoactive 3 1 -Cys-84-phycoviolobilin chromophore of phycoerythrocyanin

Triarylamine: a promising core unit for efficient photovoltaic materials

A simple pyrazoline-based fluorescent probe for Zn 2+ in aqueous solution and imaging in living neuron cells

Synthesis and optical storage properties of a novel polymethacrylate with benzothiazole azo chromophore in the side chain

Thermal, optical and electrochemical study of side chain liquid crystalline polymers bearing azo-benzothiazole chromophore in the mesogen

Fluorescence emission from dendrimers and its pH dependence

Strong blue photoluminescence and ECL from OH-terminated PAMAM dendrimers in the absence of gold nanoparticles

Poly [(maleic anhydride)-alt-(vinyl acetate)]: A pure oxygenic nonconjugated macromolecule with strong light emission and solvatochromic effect

Nonconventional macromolecular luminogens with aggregation-induced emission characteristics

Aggregation-induced emission of 1-methyl-1,2,3,4,5-pentaphenylsilole

Aggregation-induced emission: The whole is more brilliant than the parts

Immobilization of lipase from pseudomonas fluorescens on porous polyurea and its application in kinetic resolution of racemic 1-phenylethanol

Catalysis of isocyanate reaction with water by DMF and its use for fast preparation of uniform polyurea microspheres through precipitation polymerization

Easy preparation of porous polyurea through copolymerization of toluene diisocyanate with ethylenediamine and its use as absorbent for copper ions

Biocompatible polyurea dendrimers with pH-dependent fluorescence

Formation and shape transition of porous polyurea of exotic forms through interfacial polymerization of toluene diisocyanate in aqueous solution of ethylenediamine and their characterization

One step preparation of porous polyurea by reaction of toluene diisocyanate with water and its characterization

Polymerizing nonfluorescent monomers without incorporating any fluorescent agent produces strong fluorescent polymers

Effect of different substituent on three dimensional fluorescence properties of BTEX

Source of anomalous fluorescence from solutions of 4-N,N-dimethylaminobenzonitrilein polar solvents

Calculations of the absorption and emission spectra of p-N,N-dimethylaminobenzonitrile and analogues in solution

Theoretical study of the dual fluorescence of 4-(N,N-dimethylamino)benzonitrile in solution

Aggregation-induced long-lived phosphorescence in nonconjugated polyurethane derivatives at 77 K

Preparation of PEGylated polymeric nanoprobes with aggregation-induced emission feature through the combination of chain transfer free radical polymerization and multicomponent reaction: Self-assembly, characterization and biological imaging applications

Reaction mechanism of aqueous benzene with hydrogen peroxide by transient absorbance spectra

Selective aqueous benzene detection at ppb level with portable sensor based on pervaporation extraction and UV-spectroscopy

Hydrothermal route for cutting graphene sheets into blue-luminiscent graphene quantum dots

Wet chemistry route to hydrophobic blue fluorescent nanodiamond

Mechanism of photoluminescence from chemically derived graphene oxide: Role of chemical reduction

The mechanism of blue photoluminescence from carbon nanodots

Highly photoluminescent carbon dots for multicolor patterning, sensors, and bioimaging

Bioimaging based on fluorescent carbon dots

Synthesis, clustering-triggered emission, explosive detection and cell imaging of nonaromatic polyurethanes

Excitation dependence of the photoluminescence of poly(ethylene terephthalate)

Microenvironments in poly(ethylene terephthalate) film revealed by means of fluorescence measurements

Unprecedented multicolor photoluminescence from hyperbranched poly(amino ester)s

Highly emissive poly(maleic anhydride-alt-vinyl pyrrolidone) with molecular weight-dependent and excitation-dependent fluorescence

Fluorescent aliphatic hyperbranched polyether: Chromophore-free and without any N and P atoms

Unexpected strong blue photoluminescence produced from the aggregation of unconventional chromophores in novel siloxane-poly(amidoamine) dendrimers

Preparation of highly uniform and crosslinked polyurea microspheres through precipitation copolymerization and their property and structure characterization

Sticky ends in a self-assembling ABA triblock copolymer: the role of ureas in stimuli-responsive hydrogels

Aliphatic amide salt, a new type of luminogen: Characterization, emission and biological applications

In depth analysis on the unusual multistep aggregation process of oligo(ethylene glycol) methacrylate-based polymers in water

Role of water/methanol clustering dynamics on thermosensitivity of poly(n-isopropylacrylamide) from spectral and calorimetric insights

Investigating the effects of different natural molecules on the structure and oligomerization propensity of hen egg-white lysozyme

Effective enhancement of Cu ions adsorption on porous polyurea adsorbent by carboxylic modification of its terminal amine groups

Influence of mixed soft segments on microphase separation of polyurea elastomers

Phase separation of diamine chain-extended poly(urethane) copolymers: FTIR spectroscopy and phase transitions

Protonation sites and conformations of peptides of glycine (Gly 1-5 H + ) by IRMPD spectroscopy

Effect of backbone chemistry on the structure of polyurea films deposited by molecular layer deposition

Non-conventional fluorescent biogenic and synthetic polymers without aromatic rings

Preparation of highly uniform polyurea microspheres through precipitation polymerization and their characterization

Multiring-induced multicolour emission: hyperbranched polysiloxane with silicon bridge for data encryption

Emission and emissive mechanism of nonaromatic oxygen clusters

Ultra-sensitive chemosensors for

III) and explosives based on highly fluorescent oligofluoranthene

A turn-on fluorescent Fe 3+ sensor derived from an anthracene-bearing bisdiene macrocycle and its intracellular imaging application

Strong fluorescence of poly(N-vinylpyrrolidone) and its oxidized hydrolyzate

This research is financially supported by National Natural Science Foundation of China 

The authors declare no competing financial interest.

Supplementary data to this article can be found online at https://doi.org/XXXXXXXXX.