Random Copolymers of Styrene with Pendant Fluorophore Moieties: Synthesis and Applications as Fluorescence Sensors for Nitroaromatics

Five random copolymers comprising styrene and styrene with pendant fluorophore moieties, namely pyrene, naphthalene, phenanthrene, and triphenylamine, in molar ratios of 10:1, were synthesized and employed as fluorescent sensors. Their photophysical properties were investigated using absorption and emission spectral analyses in dichloromethane solution and in solid state. All copolymers possessed relative quantum yields up to 0.3 in solution and absolute quantum yields up to 0.93 in solid state, depending on their fluorophore components. Fluorescence studies showed that the emission of these copolymers is highly sensitive towards various nitroaromatic compounds, both in solution and in the vapor phase. The detection limits of these fluorophores for nitroaromatic compounds in dichloromethane solution proved to be in the range of 10−6 to 10−7 mol/L. The sensor materials for new hand-made sniffers based on these fluorophores were prepared by electrospinning and applied for the reliable detection of nitrobenzene vapors at 1 ppm in less than 5 min.


Introduction
The development of new chemosensors and applications thereof for the trace probing of various explosives is an important task for researchers, considering their potential for improving anti-terrorism measures and homeland security [1][2][3].In addition, these technologies could help to enhance the environmental and forensic investigations associated with explosive substances [4,5].Accordingly, the design and synthesis of prompt and highly sensitive devices for the detection of explosive compounds is crucial for safeguarding global security.Nitroaromatic compounds (NACs) are the principal ingredients in many explosive blends, and they are also widely exploited by the agrochemical, dye, and pharmaceutical industries.In particular, 2,4,6-trinitrotoluene (TNT); 2,4-dinitrotoluene (2,4-DNT); and 2,4,6-trinitrophenol (picric acid, PA) are widely used in military ammunition, being present in unexploded landmines around the world [6][7][8].
Although many precise techniques exist for NAC recognition and quantification (e.g., surface-enhanced Raman spectroscopy, ion mobility spectrometry, gas/liquid chromatography/mass spectrometry), they are generally quite expensive, time-consuming, and/or lack the portability necessary for real-time application [9,10].
On the other hand, various types of sensors [11], such as biosensors, electrochemical sensors, colorimetric sensors [12], and fluorescent sensors, have been developed and

Synthesis and Thermal Properties
The monomers with pendant fluorophore moieties, 4-arylstyrenes (I-V), were synthesized by reacting 4-bromostyrene with the corresponding arylboronic acid through the microwave-assisted Suzuki coupling reaction (Scheme 1) [25].The polymerization of vinyl monomers is usually carried out in a radical manner by the action of a thermal initiator.As is well-known, 2,2′-azobis(isobutyronitrile) (AIBN) is the key initiator of thermal radicals, and it can be introduced into the reaction in rather small quantities [26].The polymerization of styrene with 4-(9-phenanthrenyl)styrene (III) by the action of AIBN, as the initiator, was recently described for the manufacture of or-Scheme 1. Syntheses of 4-arylstyrenes (I-V) through the microwave-assisted Suzuki coupling reaction.
The polymerization of vinyl monomers is usually carried out in a radical manner by the action of a thermal initiator.As is well-known, 2,2 -azobis(isobutyronitrile) (AIBN) is the key initiator of thermal radicals, and it can be introduced into the reaction in rather small quantities [26].The polymerization of styrene with 4-(9-phenanthrenyl)styrene (III) by the action of AIBN, as the initiator, was recently described for the manufacture of organic fieldeffect transistor memory materials [27].We chose this reaction as an appropriate method to prepare the sensitive fluorophores (Scheme 2).The conditions of radical polymerization in THF were optimized to obtain the highest yield of the copolymer and to reach the lowest polydispersity (PDI).The impact of changing the amount of AIBN and THF, the temperature, and the reaction time was studied (Table 1).When the polymerization took place in toluene, the yield of compound P3 was two-times lower than that in THF (Table 1, Entry 2).Unfortunately, polymerization under microwave irradiation did not improve the yield of the desired product P3 (Table 1, Entry 9).The analysis of the results revealed that 1.5 mol.% initiator in 0.5 mL THF at 80 • C for 3 h proved to be the optimal conditions, providing the best yield of polymer P3 and the lowest polydispersity PDI (see Table 1, Entry 7).General conditions: 4-arylstyrenes I-V (100 mg,1 equiv.),styrene (10 equiv.), and AIBN (1.5 mol.%) in 0.5 mL THF at 80 °C/3 h.
The thermal stability of the synthesized copolymers was investigated by thermogravimetric analysis (TGA) under argon flow.The Td for all synthesized polymers proved to be higher than 385 °C (Figure 2, Table 2).These data show the good thermal stability of the obtained polymers for applications as sensor materials.According to the shape of the thermogravimetric curves, it is most likely that the thermal decomposition of all the copolymers P1-P5 proceeded along the path of polymer depolymerization, similarly to styrene.General conditions: 4-arylstyrenes I-V (100 mg,1 equiv.),styrene (10 equiv.), and AIBN (1.5 mol.%) in 0.5 mL THF at 80 The thermal stability of the synthesized copolymers was investigated by thermogravimetric analysis (TGA) under argon flow.The T d for all synthesized polymers proved to be higher than 385 • C (Figure 2, Table 2).These data show the good thermal stability of the obtained polymers for applications as sensor materials.According to the shape of the thermogravimetric curves, it is most likely that the thermal decomposition of all the copolymers P1-P5 proceeded along the path of polymer depolymerization, similarly to styrene.
The structures of the obtained vinyl monomers (I-V) and copolymers (P1-P5) were also proved by Fourier transform infrared spectroscopy (FTIR) (Table S1, Figures S6-S16).All IR spectra of I-V contained the bands that are characteristic for vibrations of vinyl groups at ~1626 cm −1 .In addition, the out-of-plane =CH 2 wagging vibrations gave rise to a strong band at ~900 cm −1 .These bands were logically absent in the FTIR spectra of P1-P5, in which vibrational modes of methylene groups were observed.The FTIR spectra for copolymers P1-P5 were similar to the spectrum of polystyrene [28].The main difference concerned the presence of CH wagging vibrations among adjacent hydrogens of the aryl substituents in the range of 890-760 cm −1 .

Photophysical Studies of the Obtained Fluorescent Polymers
The photophysical properties of the prepared polymers P1-P5 were investigated at room temperature using UV-vis and photoluminescence (PL) spectroscopy in dichloromethane (DCM) solutions and in solid state (Table 3, Figures 3 and 4, and Figures S17-S27).The UV-vis spectra of polymers P1-P5 contained maximum absorption bands at 350-270 nm (ε = 132,600-858,800 M −1 •cm −1 ), corresponding to the absorption to 4-arylphenyl substituents.An intensive band at 260-240 nm (332,700-1,318,300 M −1 •cm −1 ) could be attributed to the absorption of the phenyl ring in polystyrene [29].The long-wavelength absorption maximum was a bathochromic shift in the series P1 ≈ P2 < P3 < P5, P4 due to the increasing conjunction of the aromatic system.
Figures 4 and S17-S21 show the steady-state fluorescence spectra of P1-P5 in DCM solution.All polymers had a fluorescence band in the range of 350-450 nm, which was most likely associated with emission from the monomeric 4-arylphenyl moiety in the polymer chain [25].The PL spectra of P1, P3, P3*, and P4 (though not P2 and P5) had a vibrational structure in this fluorescence band, probably due to steric factors.Figure 4 demonstrates that the PL spectrum of P4, in addition to the monomeric structured bands (III, V), showed a longer-wavelength structureless fluorescence band centered at 478 nm, which was attributed to emission from the excimer of pyrene molecules [30].
At the transition from solution to solid powder, the emissions changed noticeably for the P4 and P5 polymers.The emission spectrum of each polymer consisted of one band at 457 and 420 nm, respectively.Tables S2 and S3, as well as Figures S28-S48, demonstrate the fluorescence decay time for all polymers in DCM and solid powders.The experimental results show that the decay process of P2 and P5 in DCM consisted predominantly of one fast component, 1.80 ns (92.2%) and 1.09 ns (98.0%), respectively, which could be assigned to 4-arylphenyl substituents.This was also observed in the solid state: τ 1 = 2.04 ns (97.5%) for P2 and τ 1 = 1.08 ns (77.3%) for P5.In addition, a small percentage of the short-lived component was observed for P4: 1.7-2.2ns (~8%).The well-known excimer conformation of pyrene in P4 at 478 nm showed 100% content with τ = 25.23 ns.An increasing average lifetime was observed for the solid powders of P1, P3, P3*, and P4.

Detection of Nitroaromatic Compounds in Dichloromethane Solution
Various nitroaromatic compounds, such as nitrobenzene, 2,4-dinitrotoluene, 2,4,6trinitrotoluene, and picric acid, are not only components of explosives and industrial reagents, but also hazardous ecotoxicants that can cause serious harm to human health (Figure 5).Therefore, we must pay great attention to detecting these dangerous chemicals both in solution and in the vapor phase [31][32][33].To estimate the ability of the copolymers P1-P5 to detect nitroaromatic analytes, the fluorescence titration of P1-P5 was performed in dichloromethane solution using a method similar to that previously described in [34].As depicted in Figure 6, all fluorophores P1-P5 acted as efficient detectors of the considered nitroaromatics in solution.The fluorescence quenching efficiency was quantified, as is typical, by the Stern−Volmer constant (K SV ); the fluorophores showed good linear quenching responses in the low concentration range of NB, DNT, TNT, and PA, as indicated by the regression factors (R 2 ) being close to 0.99 in most cases (Figures S49-S54).The Stern-Volmer equation is defined as where I 0 and I are the fluorescence intensities in the absence and presence of a quencher and [Q] represents the quencher concentration.The detection limits (DLs) were determined according to the well-known equation DL = 3σ/k, where σ is the standard deviation of the fluorophore intensity in the absence of an analyte and k is the slope of the calibration curve [8].The data on the obtained Stern-Volmer constants and the detection limits for P1-P5 copolymers are summarized in the Table 4.   S4 in Supplementary Materials).To confirm this suggestion, the quenching rate constant (k q ) value was calculated using the Stern-Volmer equation: Accordingly, the quenching rate constant value could was determined as k q = Ksv/τ 0 , where τ 0 is the average lifetime of the copolymer in the absence of an analyte.For the DNT titrated dichloromethane solution of P5, the Ksv value was found to be 4.66 × 10 4 M −1 , and the average lifetime value (τ 0 ) was calculated as 1.12 × 10 −9 s.Evidently, the calculated value of k q = 4.16 × 10 13 was much higher than the maximum scatter collision quenching constant, which is 2 × 10 10 M −1 × s −1 for dynamic quenching [35].Therefore, the detection mechanism was undoubtedly static quenching.Thus, we believe that the formation of the non-fluorescent complex P5-DNT took place for the pendant fluorophore moiety due to the π-π interaction, as was previously proved for pyrimidine polymers [33] and push-pull systems [8].The proposed mechanism of complexation is shown in Figure 8.

Preparation and Morphology of Electrospun Sensing Materials
The on-site detection of nitroaromatic vapors is critical for security providers and for the protection of human health.The electrospinning technique was used to produce sensing materials based on the synthesized polymers.As is common, polystyrene (PS) was used for the electrospinning, and the process parameters were selected according to the literature [36][37][38].Melamine formaldehyde foam cut into 2 mm sheets was selected as the substrate for material deposition due to its permeability, low fluorescence under 365 nm UV excitation, chemical inertness, availability, and ease of manipulation.The electrospinning solution was prepared by dissolving the sensor polymer in THF to obtain approximately 500 µL of 5% w/v solution.The molecular weight of the polymers was not sufficient to form fibers [36][37][38], and the deposition was achieved by the spraying of polystyrene beads.Sensing materials M1, M2, and M3 were prepared from polymers P4, P1, and P5, respectively.All electrospun polymers produced similar structures, featuring beads approximately 10 µm in diameter (8.65 ± 1.99 µm for P4 (M1), 9.48 ± 1.07 µm for P1 (M2), and 10.10 ± 2.72 µm for P5 (M3), see Figures 9 and S59).The shape of the beads resembled structures obtained for 5% solution w/v in the literature [36][37][38].Electrospun copolymers P3, P3*, and P2 emitted negligible fluorescence intensities in a solid state under 365 nm UV irradiation and were not included for further examination.Out of the remaining materials, M1 was the brightest and M2 was the dimmest, which was in good agreement with the higher absolute quantum yield in the solid state for copolymer P4 and the much lower yield for copolymer P1 (see Table 3).Figure 9 shows the obtained M1 material under magnification.

Sensor Cartridge and Fluorescence Recorder
To adapt the sensor materials for gas-phase sampling, a cartridge was manufactured from ABS plastic via 3D printing (Figure 10).The cartridge featured radially positioned sensor material slots with through-holes and was installed perpendicular to the airflow.The remainder of the cartridge's surface blocked the passage of air to force the airflow through the slots.Foam sheets carrying electrospun sensor materials were cut into fragments and installed into the slots.The cartridge design enabled the housing of an array of materials for the parallel evaluation of their responses to gas-phase analytes.For the detection of vapors, the cartridge was put into an original fluorescent recorder (Figure 11).The main design idea of this portable fluorescence recorder was the use of a compact camera for the parallel recording of the fluorescence intensities of sensing materials in the visible band (Figure S60).A UV-light-emitting diode (λ em = 365 nm) was applied for the fluorescence excitation of materials on the cartridge (Figure S61).The camera was fitted with a UV-suppressing optical filter to prevent the interference of the excitation light in the fluorescence recording.During the measurement, the video stream was transmitted for processing to the PC.Custom software was used to operate the device and extract fluorescence intensity data from the video stream.

The Sensing Performance of Materials for Various Analytes in Vapor Phase
The evaluation of the materials was performed as follows.Two measurement scenarios imitating possible sampling cases were formulated: (1) the detection of saturated vapors accumulated in a limited volume (saturated vapor scenario), and (2) the detection of diluted vapors emanating from a vapor source (diluted vapor scenario).Accordingly, the exposure to analytes was achieved either by inserting the vessel with saturated vapors into the airflow, imitating sampling inside a limited volume, or by dosing saturated vapors into the airflow to produce a constant analyte vapor concentration.The duration of exposure was set to 50 s and 100 s for the two scenarios, respectively.The UV illumination intensity was at a constant level during the test.Figure 12 shows an example of a saturated vapor case record (scenario 1); the 50 s exposure interval is marked by dashed lines.According to the Stern-Volmer equation, the fluorescence attenuation was described by the fluorescence intensity before and during exposure, defined as I and I 0 , respectively.During the exposure to vapors, I was the observed intensity; the base level I 0 had to be predicted.In the experiment, a linear model trained on pre-exposure intensity data was used for I 0 prediction.The primary task of the prediction was to account for the loss of intensity due to the increase in photobleaching over time.Differences between I and I 0 at the end of exposure to vapors I exp and at the end of post-exposure recovery under clean air I recov were taken as metrics of the sensor response to vapors and are shown in Figure 12.The fluorescence intensity responses to saturated vapors of interferents and nitroaromatic analytes are presented in Figure 13.The amplitude of the quenching by the nitroaromatics reflected the saturated vapor pressures of the analytes (372 ppm for NB, 411 ppb for DNT, 9.15 ppb for TNT, and 971 ppt for PA at 25 • C) [39,40].In the case of analyte vapors, the fluorescence response of the polymers was determined not only by the electron affinity of the analytes towards the fluorophores but also by the permeation dynamics of the vapor molecules into the depths of the polymer.For glassy polymers such as PS, the analyte permeated into the polymer as a saturated front, and the higher vapor pressure allowed for a faster front advance into the interior of the polymer [41].No significant signal was produced upon exposure to picric acid, presumably due to the saturated vapor pressure at the level below 1 ppb.The material M1 was capable of detecting saturated TNT vapors at 5.8 ppb from a 160 mL sample volume, and saturated DNT vapors were detected at 282.4 ppb and 386.1 ppb from a 160 mL sample volume by M1 and M3, respectively.All materials produced a signal of at least 30% fluorescence quenching upon exposure to NB.A summary of the detection times for the analytes is provided in Table S5.
To study the selectivity of the sensors M1-M3, the fluorescence responses towards interferent vapors were evaluated (see columns 1-10 in Figure 13).Sensor materials M1-M3 demonstrated good selectivity for NB compared to the other analytes and NACs (see column 11 in Figure 13), since their quenching sensitivities were much weaker in these vapors than in NB.
To determine the sensing performance for diluted nitro compound vapors, the materials were exposed to NB and DNT according to the diluted vapor scenario 2 (Figure S46).Picric acid and 2,4,6-TNT were not selected for the test regarding the magnitude of the response to saturated vapors because their saturated vapor levels were too low.At this dilution level, 2,4-DNT was registered by M3, and NB was registered by M1 and M3.These measurements demonstrated the applicability of the obtained sensory materials for the direct vapor-phase sensing of explosives with the use of a pre-concentrator or via closed packages allowing vapor accumulation.The lack of fluorescence recovery under clean air after exposure to nitroaromatics was common to all sensor materials.PS is known to have a sorption capacity linked to the π-π interactions of its benzene rings with aromatic analytes, which contributes to the retention of analytes in the polymer, especially those that are weakly volatile and have an electron affinity towards PS moieties [42].The tendency to accumulate analytes can be utilized for the detection of low concentrations of vapor contaminants within long exposure intervals.In order to simulate the NB detection at the maximum permissible level [43], a 1 ppm concentration was created in the airflow via a syringe driver and a 20 mL syringe containing saturated NB vapors.The sensor cartridge carrying materials M1 and M3 was used in the fluorescence recorder to sample the contaminated air; M2 was ruled out due to low brightness.
The accumulation of nitroaromatic analytes implied that the resulting fluorescence attenuation would grow as the exposure interval duration increased.Therefore, the detection times were calculated as the times required to produce the 3σ threshold signal at a certain analyte concentration so that it would be registered by the fluorescence recorder.The signal value was calculated as the difference between I and I 0 within the exposure intervals, and the noise was estimated as the difference between I and I 0 for the separate measurement under clean air.The brightness of the material dictated the gain used in the video capture, and the 3σ threshold was higher for dimmer materials due to noise amplification.
The 100 s interval before exposure to vapors was used to estimate the expected fluorescence intensity and the noise at the 3σ level (Figure 14).The results showed that both materials could detect NB at 1 ppm in less than 5 min and could be applied for air contamination monitoring.

Experimental Materials and Methods
The detailed specifications of the chemical materials and the methods used for their characterization are provided in the electronic Supplementary Materials.

Materials
Since we have already described the general details of the analytical equipment and methods [8,22,25] used, they are presented in the "General Information" section (see Supplementary Materials).

Conclusions
We described the synthesis and photophysical properties of five random copolymers comprising styrene and styrene with pendant fluorophore moieties, namely pyrene, naphthalene, phenanthrene, and triphenylamine, in molar ratios of 10:1.It was shown that the emission of these copolymers was highly sensitive towards various nitroaromatic compounds, both in solution and in the vapor phase.The detection limits of these fluorophores for nitroaromatic compounds in dichloromethane solution reached 10 −7 mol/L.The sensor materials for new hand-made sniffers based on these fluorophores were prepared by electrospinning and applied for the reliable detection of nitrobenzene vapors at 1 ppm in less than 5 min.

a
Φ F values were determined relative to 2-aminopyridine in 0.1N H 2 SO 4 as standard (Φ F = 0.60); excitation at 300 nm; b excitation at 345 nm.c Φ F values were determined using integrating sphere SC-30 of FS5 Edinburgh Instruments spectrofluorometer.

Figure 7 .
Figure 7. Fluorescence decay time curves of P5 with DNT in DCM at an excitation wavelength of 300 nm and an emission wavelength of 392 nm.

Figure 8 .
Figure 8.The proposed complexation mechanism of copolymer P5 with DNT.

Figure 11 .
Figure 11.Fluorescence recorder and cartridge used in the experiment: assembled (a) and disassembled (b).

Figure 12 .
Figure 12.Fluorescence responses of materials M1-M3 towards NB vapors in saturated vapor scenario.Vertical dashed lines mark the vapor exposure interval.

Figure 13 .
Figure 13.Fluorescence responses of sensor materials M1-M3 towards saturated vapors of interferents and nitroaromatic compounds in a 50 s exposure interval.The 3σ noise levels of registered fluorescence intensity differences are marked with horizontal dashed lines.

Figure 14 .
Figure 14.Fluorescence response of materials M1 and M3 to NB vapors at 1 ppm concentration.