An amino-substituted 2-(2′-hydroxyphenyl)benzimidazole for the fluorescent detection of phosgene based on an ESIPT mechanism

In this work, an ESIPT-based fluorescence probe, 5′-amino-2-(2′-hydroxyphenyl)benzimidazole (P1), was synthesized and explored for the ratiometric detection of phosgene. Compared to 2-(2′-hydroxyphenyl)benzimidazole (HBI), P1 exhibits high sensitivity (LoD = 5.3 nM) and selectivity toward phosgene with the introduction of the amine group. Furthermore, simple P1 loaded test papers are manufactured and display selective fluorescent detection of phosgene in the gas phase.


Introduction
Phosgene (COCl 2 ), also called carbonyl chloride, is a kind of colorless gas with a freshly cut grass-like odor and is widely used as an important industrial ingredient for the production of pesticides, pharmaceuticals, dyes, numerous chemicals and materials. [1][2][3][4][5][6] On the other hand, due to its fast reaction with alveoli protein, disrupting the blood-air barrier and resulting in suffocation, [7][8][9] phosgene is notorious for its abuse as a chemical weapon during World War I. [10][11][12][13][14] However, unlike other chemical weapons such as sarin, tabun, soman and VX, prohibited by laws all over the world, 15 phosgene can be easily generated by readily available triphosgene in the presence of tertiary amines or chloride ions. 2,16 The extensive industrial use, high toxicity and easy accessibility make phosgene a serious threat to public health and safety. Therefore, the development of facile and effective strategies for its rapid detection are urgent and necessary.
As to the detection of phosgene, conventional gas chromatography [17][18][19] and electrochemical technique 20 suffer from obvious drawbacks, such as bad portability, high cost and inconvenience in real-time detection. Nowadays, uorescent detection methods are well developed and widespread due to their easy manipulation, fast response, high sensitivity and selectivity, and the possibility of real-time detection. Among them, reports on the uorescent detection of phosgene are still limited [12][13][14] and mainly based on: (i) twice carbamylation reactions of uorescence probes, which employ o-phenylenediamine, 21 o-hydroxyaniline, 22 o-aminobenzyl amine, 23 catechol, 24 ethylenediamine, 25 ethanolamine 26 or other moieties 27 as reactive site; (ii) phosgene-promoted dehydration reaction of uorescence probes, which employ oxime 28 or amide 29 as reactive site; and (iii) several other phosgene-induced reactions, including intermolecular reaction of two uorophores, 30 intramolecular reaction of cinnamic acids, 31 ring opening reaction of benzimidazole-fused rhodamine dye 32 and Beckmann rearrangement of ketoxime 33 (Table S2 †). However, most of these uorescence probes are based on mechanisms of photoinduced electron transfer (PeT), intramolecular charge transfer (ICT) or aggregation-induced emission (AIE), which might be disturbed by acetylating, 21c phosphorylating agents 21h or oxidizing chemicals, 21e,27g resulting to false response. And several other drawbacks still need to be overcome, such as incapable of discrimination between triphosgene and phosgene 21b,21f and tedious preparation process. 22,24,25 Due to the promising advantages, such as high selectivity, high sensitivity and large Stokes shi, excited-state intramolecular proton transfer (ESIPT) based uorophores has attracted considerable attention in the development of new uorescence detection methods. 34 However, only two ESIPT based uorescent probes have been developed for the ratiometric detection of phosgene through twice carbamylation of their keto or enol from, i.e.  36 Despite successful application of the two ESIPT based probes, ABT is unable to discriminate between phosgene and triphosgene, and Pi suffers from high LoD in solution (0.14 mM).
In order to overcome these drawbacks and develop '6S' (simpleness, speedy, selectivity, sensitivity, stability and smart) uorescence probe for the detection of phosgene, we select a readily prepared small molecule, 2-(2 0 -hydroxyphenyl)benzimidazole (HBI), as our initially explored uorophore (Scheme 1, R ¼ H). Preliminary experiments in our group showed a fast and sensitive reaction between HBI and phosgene at a concentration of 10 mM, resulting to an obvious uorescent intensity increase at 351 nm and decrease at 462 nm. However, no obvious color change was observed under hand-held 365 nm UV light and the selectivity of HBI was not satisfactory (Fig. S1 †). Inspired by the fast and sensitive reaction of HBI toward phosgene, we continuously committed our efforts to develop new HBI derivatives with large Stokes shied emission and high selectivity. Herein, we report an amino modied 2-(2 0 -hydroxyphenyl)benzimidazole (P1) for the ratiometric detection of phosgene (Scheme 1, R ¼ NH 2 ). Through the introducing of the strong electron-donating amino group at the C5 0 position of HBI, the uorescent emission is red-shied to 540 nm. Upon addition of phosgene, P1 is converted to a six-membered ringcontaining product P1-CO, the ESIPT process is disrupted and the emission color changes from yellow to bright blue under hand-held 365 nm UV light. Moreover, compared to HBI, the selectivity of P1 is also well improved. Furthermore, simple and easily prepared P1 loaded test papers are fabricated and display an obvious uorescence color change upon exposure to phosgene in the gas phase.

Reagents and instruments
All reagents and solvents were obtained from commercial suppliers and used without further purication. 1 H and 13 C NMR spectra were recorded on a Bruker 400 spectrometer in DMSO-d 6 containing tetramethylsilane as an internal standard. Fluorescence emission spectra were collected by Hitachi F7000 uorescence spectrometer. UV-Vis absorption spectroscopy measurements were performed on Thermo Evolution 260 Bio at room temperature. High-resolution mass spectra (HRMS) were obtained with a Thermo LTQ Orbitrap mass spectrometer.

Synthesis and characterization of P1
5-Aminosalicylic acid (0.153 g, 1.0 mmol) and o-phenylenediamine (0.108 g, 1.0 mmol) were mixed and stirred in polyphosphoric acid (85% phosphorus pentoxide, 10 mL) at 160 C over a period of 3 h. 37 Then the reaction mixture was poured into distilled water (20 mL), and the precipitate were collected and washed by saturated NaHCO 3 aqueous solution to pH 8-9. The obtained crude product was further puried by column chromatography (ethyl acetate: petroleum ether ¼ 3 : 1) to give P1 (0.108 g, yield 48%) as a purple solid. 1  Synthesis and characterization of P1-CO P1 (0.113 g, 0.5 mmol) and triethylamine (0.1 mL) were dissolved in dry CH 2 Cl 2 (15 mL) at 0 C, then triphosgene (0.15 g, 0.5 mmol) in dry CH 2 Cl 2 (5 mL) was added over a period of 10 min. The mixture was continually stirred at 0 C until the completion of the reaction. Saturated NaHCO 3 aqueous solution was added into the mixture and extracted with CH 2 Cl 2 (20 mL Â 2). The organic phase was collected, dried over anhydrous Na 2 SO 4 and evaporated to give the crude product. The crude product was further puried by column chromatography (ethyl acetate: petroleum ether ¼ 1 : 2) to give P1-CO (0.098 g, yield 80%) as a brown solid. 1

General analytical experiment procedure
Phosgene was generated by addition of the less-toxic triphosgene to the solution of TEA (triethylamine). Stock solutions of P1 and triphosgene for UV-Vis and uorescence titrations were prepared in CH 2 Cl 2 (1.0 mM). Stock solution of triethylamine (TEA) was prepared in CH 2 Cl 2 (0.1 vol%). Relevant analytes ((COCl) 2 , CH 3 COCl, SOCl 2 , TsCl, DCP, HOAc, POCl 3 and SO 2 Cl 2 ) solutions were also prepared in CH 2 Cl 2 (10 mM). The test solutions were prepared by dilution of the stock solutions. For detection of phosgene in solutions: 20 mL probe solution and equal amount of TEA solution were added and mixed in 2.0 mL of CH 2 Cl 2 , then triphosgene stock solution (1-12 mL) was added to generate phosgene. For detection of relevant analytes in solutions: 20 mL probe solutions and equal amount of TEA solution were added and mixed in 2.0 mL of CH 2 Cl 2 and then 10 mL relevant analytes stock solutions were added. All solutions were kept for 10 min before the UV-Vis and uorescence testing.

Preparation of P1 loaded test papers and detection of gaseous phosgene
Filter paper was cut into small strips (0.5 cm Â 2.5 cm) and dipped in the P1 stock solution (1.0 mM), then removed and dried under air. The dip-dry process was repeated for 2 times. For detection of gaseous phosgene: the pre-prepared lter paper was rstly pasted on the inner wall of a 10 mL centrifuge tube, and 20 mL TEA stock solution (0.1 vol% in CH 2 Cl 2 ) was added by the use of a HPLC needle. Following that, 2 mL, 4 mL, 8 mL, 12 mL and 16 mL of triphosgene solution (10 mM in CH 2 Cl 2 ) were added in the centrifugal tube, respectively. Thus, the concentration of phosgene was calculated to be 0.6 mg L À1 , 1.2 mg L À1 , 2.4 mg L À1 , 3.6 mg L À1 and 4.8 mg L À1 , assuming that triphosgene was completely decomposed into gaseous phosgene. The centrifugal tubes were sealed immediately and kept for 10 min, then imaged under 365 nm UV-light. For detection of other gaseous analytes: (COCl) 2 , CH 3 COCl, SOCl 2 , TsCl, DCP, HOAc, POCl 3 and SO 2 Cl 2 (4 mL, 0.1 M in CH 2 Cl 2 ) were added in the above mentioned 10 mL centrifugal tube with our pre-prepared lter paper. Then, the centrifugal tubes were sealed and kept for 10 min. The uorescent response to each analytes vapor was observed and imaged under 365 nm UV-light, respectively.

Results and discussion
Synthetic process of P1 and P1-CO As shown in Scheme 2, P1 was synthesized by one-pot condensation reaction of o-phenylenediamine with 5-aminosalicylic acid in PPA at 160 C for 3 h. 37 The related proposed sensing product P1-CO was synthesized through fast twice carbamylation reaction between P1 and triphosgene in the presence of TEA with good isolated yield. HPLC-MS analysis of the reaction mixture showed a complete conversion of P1 and the 5 0 -amino group did not react with phosgene under this condition (Fig. S7 †). The structures of P1 and P1-CO were fully characterized by 1 H NMR, 13 C NMR and highresolution mass spectroscopy (HRMS).

Photo physics property of P1 and P1-CO
As shown in Fig. 1, the photo physics property of the assynthetic two compounds were investigated. Due to the intramolecular proton transfer of the hydroxyl and imidazole moieties, P1 in CH 2 Cl 2 (10 mM) displays the absorption spectral maximum (l abs max ) at 305 and 355 nm and a mainly yellow uorescent emission maximum (l em max ) at 540 nm. While, the absorption and uorescence spectral maximum (l abs max /l em max ) of P1-CO were observed to be 305/358 nm with a blue uorescent emission. The uorescence quantum yields (F f ) of P1 (4.3%) and P1-CO (56%) were measured with quinine sulfate as the reference (F f ¼ 54% in 1.0 M H 2 SO 4 ) (Fig. S2 †). Therefore, P1 would be employed as a ratiometric uorescent probe for detection of phosgene with a uorescence-emission color change from yellow to blue under 365 nm UV lamp (Fig. 1, inset).

Fluorescence titration of P1 to phosgene
The uorescence titration analysis of P1 to phosgene was performed as shown in Fig. 2A. The uorescent spectra of P1 showed a regular change aer 0-6 mM triphosgene was added. The emission intensity at 540 nm gradually decreased while the peak at 358 nm increased. The ratio of the emission intensities at 358 and 540 nm was found to depend linearly on the triphosgene concentration over the 0-3.5 mM range (Fig. 2B). The triphosgene detection limit was determined to be 5.3 nM (LoD ¼ 3s/k, where s is the standard deviation of the blank experiment, and k is the slope of the relationship between the emission-intensity ratio and the phosgene concentration) (Fig. S4 †).

Response time of P1 to phosgene
Time-dependent uorescence spectral was performed to assess the response time of P1 to phosgene. As shown in Fig. 3, the uorescence intensities at 358 nm for 10 mM P1 solution with 6 mM triphosgene in the presence (black line) or absence (red line) of TEA (20 mL, 0.1 vol%) was recorded. As it can be seen, the uorescence intensity at 358 nm rapidly increased (within 50 s) Scheme 2 Synthesis of P1 and P1-CO.  upon the addition of triphosgene and TEA (black line). On the contrary, there is no obvious change in the uorescence intensity at 358 nm without TEA (red line). Therefore, due to the different reactivity of P1 to phosgene and triphosgene, the designed probe P1 can efficiently discriminate between phosgene and triphosgene.
Sensing mechanism of P1 to phosgene Initially, P1 displayed dual uorescence emission at 412 nm and the mainly 540 nm (Fig. 1), which should be the sign of ESIPT and indicated the coexistence of the keto and enol form coexist of P1 in solution. 38 Upon addition of phosgene, the hydroxyl and imidazole moieties of P1 were connected by the carbonyl group, resulting to an intramolecular six-membered ring. The keto-enol tautomerization and the ESIPT process of P1 was blocked. The bluish uorescent emission of P1-CO could be attributed to the ESIPT 'enol' form. In order to conrm the proposed mechanism, P1 and the proposed sensing product P1-CO were characterized by 1 H NMR. As shown in Fig. 4, the peaks at 12.16 ppm for hydroxyl and 13.01 ppm for the imidazole amino groups disappear, which conrms that these groups serve as the phosgene-recognition sites. Moreover, the proposed sensing product was characterized by HRMS for C 14 H 10 N 3 O: M + H + : calculated 252.0768, found 252.0770 (Fig. S13 †), indicating that the molecular weight (m/z) of the puried product corresponds with that of P1-CO. These data strongly backed the proposed mechanism.

Sensing selectivity of P1 to phosgene
The selectivity of P1 for the detection of phosgene over various potential interferents were assessed. As shown in Fig. 5, 10 mM P1 solutions containing TEA (20 mL, 0.1 vol%) were treated with phosgene (triphosgene (6 mM)) and 50 mM interferents solutions. The uorescence spectroscopy reveals a clear change for phosgene at 358 nm (Fig. S5 †). The uorescence increments at 358 nm and 540 nm exhibits a highly specic and selective detection of phosgene over various acyl chlorides. Moreover, the selectivity of P1 can be obviously observed under 365 nm UV-light from yellow to blue only for phosgene (Fig. 5, inset). Furthermore, the determination of phosgene in the presence of these interfering compounds were performed. As shown in Fig. S6 and Table S1, † the uorescent intensities at 358 nm were not signicantly affected by these interfering compounds at 5.0 mM and good recovery rate was obtained (from 87.8% to 108.9%). However, increased concentration (10 mM) would cause obvious uorescent intensity decrease. This might be due to the block of recognition sites by interfering compounds at high concentration.

Detection of gaseous phosgene and other vapors
In order to investigate the possible application of P1 for the detection of gaseous phosgene, simple and low cost P1 test papers were prepared. As shown in Fig. 6, P1 test paper exhibited yellow uorescence under 365 nm UV light. Aer    exposure to various amounts of gaseous phosgene (0-4.8 mg L À1 ), a yellow-green-blue uorescence color change was clearly observed. Moreover, the selectivity of P1 loaded test paper to phosgene over other related analytes were also investigated (Fig. 7). The results showed that P1 test paper did not give false responses to other potential interferents. More importantly, the detection of gaseous phosgene were also founded effective even though the test paper was pre-exposed to the vapors of potential interferents. Therefore, based on all these results, P1 can be employed for selective detection of phosgene both in solutions and in gas phase.

Conclusions
In summary, we have designed and synthesized a new 5 0 -amino-2-(2 0 -hydroxyphenyl)benzimidazole (P1) for the uorescent detection of phosgene both in solution and in gas phase. This probe employs hydroxyl and imidazole moieties as recognition sites to trap phosgene and bears a strong electron donating amino group to expand the Stokes shied emission and improve the selectivity. Through twice carbamylation of the recognition sites, the ESIPT process of P1 is forbidden and an obvious uorescent color change from yellow to blue can be easily observed by naked eyes. This probe is sensitive (LoD ¼ 5.3 nM) and highly selective toward phosgene over triphosgene and other acyl chlorides. Furthermore, P1 loaded test paper was fabricated and successfully applied for selective detection of gaseous phosgene with an easily observed yellow to blue color change.

Conflicts of interest
There are no conicts to declare.