Article pubs.acs.org/ac
Rational Design of Fluorescent Phosgene Sensors Pradip Kundu and Kuo Chu Hwang* Department of Chemistry, National Tsing Hua University, Hsinchu 30043, Taiwan R.O.C. S Supporting Information *
ABSTRACT: Phosgene is a very toxic gas, which was used as a chemical weapon in World War I, and is currently widely used in industrial processes. So far, no any phosgene fluorescent sensor has been reported. In this study, we report rational design of unimolecular fluorescent phosgene sensors for the first time. Phosgene was used to initiate intramolecular cyclization and convert nonfluorescent molecules to highly fluorescent products. Bright blue fluorescence of phosgene reaction products can be easily visualized by naked eye. The detection limit for phosgene is as low as 1 nM in solutions at room temperature. line diamond thin film.13 The phosgene detection limit is about 5 ppm. In the first two of the above methods, triphosgene was used, instead of phosgene itself. Although the above methods have been reported, there is still a lack of fluorescent sensing molecules which can detect phosgene directly, especially by commercially available molecules. In the view of increasing demand of phosgene in chemical industry and also more spreading terrorist activities in the world nowadays, it is necessary to develop a very concrete, extremely sensitive, and easy way to detect phosgene. In this paper, we report a rational design of fluorescent phosgene sensors. The key feature of the current method is to utilize the capability of a phosgene molecule to react with two adjacent reactive functional groups in the same molecules. Intramolecular cyclization was previously demonstrated to increase in the molecular rigidity and thus fluorescence intensity.14 Alternatively, intramolecular cyclization might lead to an increase in the conjugation length and thus a shift in the fluorescence emission maximum wavelengths of the reactant molecules. On the basis of this idea, we looked for and found several commercially available potential phosgene fluorescent sensor molecules. Scheme 1 shows the chemical structures of six fluorescent phosgene sensing molecules and their corresponding phosgene reaction products. All these sensing molecules are weakly or nonfluorescent, whereas their phosgene reaction products are highly fluorescent. The phosgene detection limits for all six molecules are in the range of 1−18 nM. The presence of phosgene leads to clear fluorescence color changes in solutions which can be clearly visualized by the naked eye.
P
hosgene is a very toxic gas, which was used in World War I as a chemical weapon. The lethal dose for human being is 2 ppm.1 Phosgene exhibits several lung irritant effects, including pulmonary edema and asphyxia.2,3 It is very difficult to notice when the colorless phosgene gas was inhaled, since the irritant effects (or symptoms) of phosgene appear a few hours later.4,5 Phosgene having a molecular formula of OCCl2 is chemically very reactive toward amines, alcohols, carboxylic acids, and moisture.6 Phosgene has many important industrial applications in the synthesis of many organic compounds, particularly in the preparation of plastics and pesticides. Although many fluorescent sensors have been developed for a variety of chemical warfare agents,7−9 chemical sensor molecules able to detect or respond to the presence of phosgene, instead of its stimulant, are still very rare. The extremely high toxicity of phosgene retards its investigation and development of detection methods. In the literature, it was reported a bimolecular phosgene colorimetric sensor, where gold nanoparticles were surface-functionalized with cysteine. In the presence of triphosgene (a simulant of phosgene), coupling of −SH groups on different Au NPs was initiated, leading to aggregation of gold nanoparticles and thus a color change from red to blue.10 When gold nanoparticles were replaced by two different fluorescent chromophores, the presence of triphosgene leads to coupling the two different chromophores and enables the fluorescent resonance energy transfer (FRET) between the two chromophores.11 The detection limit of triphosgene for the above bimolecular colorimetric and fluorescence resonance energy transfer sensors are 420 nM 10 and 50 μM,11 respectively. In another study, it was reported that doping of HCl generated by reaction of phosgene and moisture can increase the electrical conductivity of the polyaniline film.7 The phosgene detection limit for the polyaniline conductivity method was reported to be 0.01 ppm in the gas phase.12 Note that 0.01 ppm in the gas phase is equivalent to ∼80 nM in solutions, assuming an average density of 0.8 g/mL for organic solvents. In another report, phosgene was found to be able to change the surface conductivity of a H-terminated nanocrystal© XXXX American Chemical Society
Received: March 15, 2012 Accepted: April 10, 2012
A
dx.doi.org/10.1021/ac300737g | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Scheme 1. Chemical Structures of Weakly Fluorescent Phosgene Sensing Molecules and Their Highly Fluorescent Reaction Products
■
EXPERIMENTAL SECTION Phosgene was generated in situ by reaction of triphosgene with Aliquat 336 in hexane in a cool trap (see the experimental setup shown in Scheme SI in the Supporting Information).13,15,16 Aliquots of triphosgene hexane solution were injected slowly with stirring at room-temperature to allow reaction with Aliquat 336 and in situ generation of phosgene. Phosgene so-generated in the cool trap A was brought to the cool trap B to react with sensing molecules (such as, 2,4-dihydroxy cinnamic acid, 1 μM concentration in CHCl3 in the presence of 1 mM pyridine) which contains at least two adjacent phosgene-reactive groups. The reaction mixture was removed from the reaction vessel for fluorescence measurements 10 min after every addition of triphosgene to the cool trap A. The fluorescence spectra of the reaction solutions were recorded using an Edinburgh FLS920 spectrometer. The presence of the third, fourth, and fifth cool traps guarantees the complete reaction of phosgene and avoids leakage of the highly toxic phosgene gas into the outside environment. The cool traps C, D, and E contain phosgene reactive methanol, which is highly reactive toward phosgene and can prevent escape of any unreacted phosgene gas to the outside environment. The chemical structures of the final phosgene reaction products were isolated by column chromatography (silica gel mesh 60−120) using 5% MeOH in dichloromethane as eluent and characterized by 1H NMR, 13 C NMR, and high-resolution mass spectrometry (HRMS) (see the Supporting Information). Warning: phosgene gas is highly toxic. The experiments should be done with extreme care and in a well-ventilated hood with a phosgene gas detector. General Data. NMR spectra were recorded on Mercury400 MHz and Varian-500 MHz spectrometers. The fluorescence spectra of the reaction products were recorded using an Edinburgh FLS920 spectrometer. HRMS spectra were recorded on a MAT-95XL instrument. Melting points were determined on a digital melting point apparatus. The fluorescence quantum yields of six phosgene reaction products, 1p∼6p, were determined using 9,10-diphenylanthracene (fluorescence quantum yield ΦF = 0.9) in cyclohexane as a reference standard.17
Figure 1. (a) Fluorescence emission spectra from a chloroform solution containing 1 μM of 2,4-dihydroxycinnamic acid (1) in the absence (black line) and the presence of different phosgene concentrations. The concentrations of phosgene are labeled in the figure. The excitation wavelength is 330 nm. The chemical structure of the coupling product, 7-hydroxycoumarin (1p), is shown in the figure. (b) The increase in the fluorescence intensity at 382 nm is plotted as a function of the phosgene concentration. The inset optical images of the blue fluorescence are from the excited 1p at different phosgene concentrations as labeled on top of each image. The inset plot shows the linear fitting of the experimental data.
infra), is highly fluorescent with an emission maximum wavelength at 382 nm and an additional small peak at 451 nm. Upon injection of aliquots of triphosgene to generate phosgene and subsequent reaction with 2,4-dihydroxy cinnamic acid, the overall fluorescence intensity of the solution in the cool trap B increases dramatically (see Figure 1a). The fluorescence intensity of 1 nM 7-hydroxycoumarin is higher than that of 1 μM 2,4-dihydroxy cinnamic acid. From the relative fluorescence intensities, the fluorescence quantum yield of the precursor, i.e., 2,4-dihydroxy cinnamic acid (1), is at least 3 orders smaller than that of the phosgene reaction product, i.e., 7-hydroxycoumarin (1p). The increase in the fluorescence at 382 nm was linearly proportional to the amount of phosgene generated (see Figure 1b). The blue fluorescence of the reaction product, i.e., 7-hydroxycoumarin, can be clear visualized by the naked eye (see the inset pictures in Figure 1b). At a higher dose of phosgene, more 7-hydroxycoumarin was formed leading to more intense blue fluorescence. The detection limit of the current fluorescence sensor system can be determined to be 9 nM (see also Table 1) using the value, 3 s/ m,10,19 where s is the standard deviation of a set of blank measurements (n = 12) and m is the slope of the linear regression fitting of the measurements in the presence of analyte. When o-hydroxy cinnamic acid was used to replace 2,4-
■
RESULTS AND DISCUSSION Phosgene was generated in situ by reaction of triphosgene with Aliquat 336 in hexane in a cool trap.15,16,18 The phosgene generated was introduced by a slow N2 gas flow to a second cool trap containing phosgene sensing molecules. As shown in Figure 1a, the fluorescence of 1 μM 2,4-dihydroxy cinnamic acid has a very weak emission maximum at 365 nm, whereas the phosgene reaction product, i.e., 7-hydroxycoumarin (vide B
dx.doi.org/10.1021/ac300737g | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
3-amino-2-naphthoic acid was used (under the same condition and at a concentration of 1 μM), strong fluorescence can be observed upon generation of phosgene in the cool trap A (see Figure 2a). The phosgene reaction product, i.e., naphtho[2,3-
Table 1. Detection Limits and Fluorescence Quantum Yields of Six Phosgene Fluorescent Sensor Molecules and Analogues
dihydroxy cinnamic acid, similar results were obtained with a detection limit of 18 nM (see Figure S1 in the Supporting Information). The fluorescence intensity of coumarin is, however, slightly less intense than that of 7-hydroxycoumarin. Phosgene is known to be chemically very reactive toward amine, alcohol, carboxylic acid, etc. The 2,4-dihydroxy cinnamic acid contains both hydroxy and carboxylic acid moieties. The chemical reaction between phosgene and o-hydroxy cinnamic acid could occur through two pathways (see Scheme 2). In path Scheme 2. Plausible Reaction Mechanisms for the Rapid Reaction of o-Hydroxy or 2,4-Dihydroxy Cinnamic Acid with Phosgene to Produce Coumarin or 7-Hydroxycoumarin
Figure 2. (a) Fluorescence emission spectra from a chloroform solution containing 1 μM of 3-amino-2-naphthoic acid (6) in the absence (black line) and the presence of different phosgene concentrations. The excitation wavelength is 380 nm. The concentrations of phosgene are labeled in the figure. The chemical structure of the coupling product, naphtho [2,3-b]azet-2(1H)-one (6p), is shown in the figure. (b) The increase in the fluorescence intensity at 430 nm is plotted as a function of the phosgene concentration. The inset optical images of the blue fluorescence are from the excited 6p at different phosgene concentrations as labeled on top of each image. The inset plot shows the linear fitting of the experimental data.
A, phosgene reacts with carboxylic acid moiety first to form the acid chloride 1Ai, which then further couples with the ohydroxy group and leads to formation of 7-hydroxycoumarin 1p. In the path B, phosgene reacts with the o-hydroxyl group to form the intermediate product 1Bi, which further couples with the carboxylic acid moiety to form product 1Bj. Because of large ring strain, thermal decomposition, via release of stable CO2, of 1p occurs easily and the same final coumarin product 1p was formed. When R is equal to −OH, reaction of phosgene can also occurs at the R site, which leads to formation of nonfluorescent product. From the 1H NMR, 13C NMR, and high-resolution mass spectrums, we identify the major phosgene reaction products with 2,4-dihydroxy cinnamic acid and 4-hydroxy cinnamic acid to be 1p and 2p, respectively (see Figure S2 in the Supporting Information for 1H and 13C NMR spectra). Besides 2,4-dihydroxy cinnamic acid and o-hydroxy cinnamic acid, four other compounds shown in Scheme 1 also match with our rational design of potential fluorescent phosgene sensors. These four compounds were also examined for their abilities to sense phosgene and generate fluorescent products. Indeed, as expected, they all can detect the presence of phosgene with detection limits of 1−14 nM. For example, when
b]azet-2(1H)-one (6p), is far more highly fluorescent than the precursor, 3-amino-2-naphthoic acid (6), which is very weakly fluorescent with an emission maximum at 488 nm. The phosgene reaction product, 6p, contains a strained four membered ring, and therefore the highest occupied molecular orbital−lowest unoccupied molecular orbital (HOMO− LUMO) energy gap becomes larger than the precursor, leading to a blue shift in the emission maximum to 432 nm. It is known that the bandgap between the HOMO and LUMO of molecules can be modulated by molecular strain.20−22 Since the product, 6p, is highly fluorescent with a very broad band emission band extended to 500 nm region, its intense blue fluorescence can be clearly visualized, even at a very low concentration of ∼1 nM (see the inset in Figure 2b). The other three sensing molecules, 2′-hydroxy-2-biphenylcarboxylic acid (3), (E)-3-(2-aminophenyl)-2-(naphthalen-1-yl) acrylic acid (4), and 2-aminocinnamic acid (5), all can detect phosgene with very high sensitivities (see Figures S3, S4, and S5 in the C
dx.doi.org/10.1021/ac300737g | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Supporting Information for fluorescence spectra and images). Table 1 lists the phosgene detection limits, fluorescence emission maxima, and quantum yields of all six sensing molecules investigated in this study. The common feature of all six unimolecular sensors is that the sensing molecules are very weakly or nonfluorescent, whereas the phosgene reaction products are highly fluorescent with fluorescence quantum yields at least 3 orders higher than the parent precursors. Since all these compounds are unimolecular, they all are highly sensitive and reactive toward phosgene. The phosgene limits are mostly near 1 nM or even lower.
(8) Cho, D. G.; Sessler, J. L. Chem. Soc. Rev. 2009, 38, 1647−1662. (9) Hill, H. H., Jr.; Martin, S. J. Pure Appl. Chem. 2002, 74, 2281− 2291. (10) Feng, D.; Zhang, Y. Y.; Shi, W.; Li, X.; Ma, H. Chem. Commun. 2010, 46, 9203−9205. (11) Zhang, H. X.; Rudkevich, D. M. Chem. Commun. 2007, 1238− 1239. (12) Virji, S.; Kojima, R.; Fowler, J. D.; Villanueva, J. G.; Kaner, R. B.; Weiller, B. H. Nano Res 2009, 2, 135−142. (13) Davydova, M.; Kromka, A.; Exnar, P.; Stuchlik, M.; Hruska, K.; Vanecek, M.; Kalbac, M. Phys. Status Solidi A 2009, 206, 2070−2073. (14) Zheng, S. W.; Swager, T. M. J. Am. Chem. Soc. 2003, 125, 3420− 3421. (15) Pasquato, L.; Modena, G.; Cotarca, L.; Delogu, P.; Mantovani, S. J. Org. Chem. 2000, 65, 8224−8228. (16) Eckert, H.; Auerweck, J. Org. Process Res. Dev. 2010, 14, 1501− 1505. (17) Hamai, S; Hirayama, F. J. Phys. Chem. 1983, 87, 83−89. (18) Trotzki, R.; Niicheter, M.; Ondruschka, B. Green Chem. 2003, 5, 285−290. (19) Sydoryk, I; Lim, A.; Jager, W; Tulip, J.; Parsons, M. T. Appl. Opt. 2010, 49, 945−949. (20) Zhang, D. B.; Dumitrica, T. Small 2011, 7 (No. 8), 1023−1027. (21) Yang, S. Y.; Prendergast, D.; Neaton, J. B. Nano Lett. 2010, 10, 3156−3162. (22) Ni, Z. H.; Yu, T.; Lu, Y. H.; Wang, Y. Y.; Feng, Y. P.; Shen, Z. X. ACS Nano 2008, 2 (11), 2301−2305. (23) Yamamoto, Y.; Kirai, N. Org. Lett. 2008, 10, 5513−5516. (24) Rao, H. S. P.; Sivakumar, S. J. Org. Chem. 2006, 71, 8715−8723.
■
CONCLUSIONS In summary, we have demonstrated a strategy of rational design for searching of fluorescent phosgene sensor, which works very well. The strategy utilizes phosgene to initiate intramolecular cyclization, extend the length of a conjugate system, and convert weakly fluorescent sensing molecules to highly fluorescent products. On the basis of the strategy, six molecules bearing two phosgene reactive groups were shown to be able to detect the presence of phosgene and form highly fluorescent products, with a phosgene detection limit of 1−18 nM at room temperature (see Table 1), which is 80−420 times lower than the reported literature values.9−12 The appearance of blue fluorescence can be easily visualized by the naked eye. The current method also demonstrate that phosgene, although very toxic, is an excellent “molecular glue” to trigger intramolecular cyclization, leading to an efficient one step synthesis of biomedically active coumarins and analogues, which are superior to many literature reported, multiple steps reactions.23,24
■
ASSOCIATED CONTENT
S Supporting Information *
Experimental procedures and 1H, 13C NMR spectra, and fluorescence spectra of four sensing molecules at different phosgene concentrations. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS The authors gratefully acknowledge the National Science Council, Taiwan, for financial support. REFERENCES
(1) Virji, S.; Kojima, R.; Fowler, J. D.; Villanueva, J. G.; Kaner, R. B.; Weiller, B. H. Nano Res. 2009, 2, 135−142. (2) Sciuto, A. M.; Stotts, R. R.; Hurt, H. H. J. Appl. Toxi. 1996, 16, 381−384. (3) Noort, D.; Hulst, A. G.; Fidder, A.; van Gurp, R. A. Chem. Res. Toxicol. 2000, 13, 719−726. (4) Kodavanti, U. P.; Costa, D. L.; Giri, S. N.; Starcher, B.; Hatch, G. E. Fundam. Appl. Toxicol. 1997, 37, 54−63. (5) Davydova, M.; Kromka1, A.; Exnar, P.; Stuchlik, M.; Hruska, K.; Vanecek, M.; Kalbac, M. Phys. Status Solidi A 2009, 206, 2070−2073. (6) Babad, H.; Zeiler, A. G. Chem. Rev. 1973, 73, 75−91. (7) Burnworth, M.; Rowan, S. J.; Weder, C. Chem.Eur. J. 2007, 13, 7828−7836. D
dx.doi.org/10.1021/ac300737g | Anal. Chem. XXXX, XXX, XXX−XXX
