Fluoride-selective colorimetric sensor based on thiourea binding site and anthraquinone reporter

Fluoride-selective colorimetric sensor based on thiourea binding site and anthraquinone reporter

Spectrochimica Acta Part A 65 (2006) 633–637 Fluoride-selective colorimetric sensor based on thiourea binding site and anthraquinone reporter Fang-yi...

238KB Sizes 3 Downloads 149 Views

Spectrochimica Acta Part A 65 (2006) 633–637

Fluoride-selective colorimetric sensor based on thiourea binding site and anthraquinone reporter Fang-ying Wu ∗ , Mei-hua Hu, Yu-mei Wu, Xiao-fang Tan, Yong-qiang Zhao, Zhao-jun Ji The Center of Analysis and Testing, Department of Chemistry, Nanchang University, Nanchang 330047, China Received 12 August 2005; received in revised form 12 December 2005; accepted 13 December 2005

Abstract A structurally simple colorimetric sensor, N-4-nitrobenzene-N -1 -anthraquinone-thiourea (1), for anions was synthesized and characterized by H NMR, ESI mass and IR methods. In acetonitrile, the addition of F− changed 1 solution from colorless to yellow. In the presence of other anions such as CH3 CO2 − , H2 PO4 − , HSO4 − and Cl− , however, the absorption spectrum of 1 was slightly red shifted with no obvious color changes observed. The association constants of anionic complexes followed the order of F−  CH3 CO2 − > H2 PO4 − > HSO4 − > Cl− > Br− , which was different from the order of anion basicity. AM1 calculation results indicated that the most stable configuration of 1 existed in the Z–E-conformation with a sixmembered ring via intramolecular hydrogen bond. This made thiourea moiety of 1 in an unfavorable conformation to bond with oxygen-anionic substrates such as CH3 CO2 − and H2 PO4 − , thus leading to a high selectivity and sensitivity for the detection of F− . © 2005 Elsevier B.V. All rights reserved. 1

Keywords: Anion recognition; Colorimetric assay; N-4-nitrobenzene-N -1 -anthraquinone-thiourea

1. Introduction The design and synthesis of systems that are capable of sensing various biologically and chemically important anions are currently of major interest because anions play a fundamental role in chemical and biological process [1–5]. Among various important anionic analytes, fluoride anion is one of the most significant due to its role in dental care and treatment of osteoporosis [6,7]. Therefore, recently a large number of excellent examples for sensing fluoride anion have been reported [8–17]. Colorimetric sensors have attracted much attention for allowing the so-called ‘naked-eye’ detection and offering qualitative and quantitative information without using expensive equipments. A chemosensor usually consists of two parts, a recognition moiety and a signal reporter. They either link directly [8] to allow for a highly efficient communication between the recognition moiety and the reporter or are linked by a flexible spacer [18] to ensure an optimal approaching of them. Anion recognition in biological systems is very often achieved via hydrogen bond by highly preorganized proteins with sterically well-defined complex sites in the interior of proteins [19,20]. Hydrogen-bond donors such as



Corresponding author. Tel.: +86 791 830 4414. E-mail address: [email protected] (F.-y. Wu).

1386-1425/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2005.12.022

pyrrole/calixpyrrole [8,10,18,21], (thio)urea [11,14–17,22–24], guanidinium [25], azophenol [26] and amide [9,12,27] usually act as anion binding sites. Among a variety of possible hydrogenbond donors, thiourea derivatives have been shown to be particularly good. In this paper, we reported a new chromogenic sensor, N-4-nitrobenzene-N -1 -anthraquinone-thiourea, in which thiourea moiety is able to bond with anion guests via hydrogenbond interaction and the anthraquinone group serves as chromogenic signaling subunit. Aminoanthraquinone was chosen as the chromophore because it can act as an optically sensitive indicator for anion recognition especially for the detection of fluoride anion [17,28]. The experiment results showed that receptor 1 could distinguish F− from other anions like Cl− and Br− , which is of great significance for an assay of F− despite the large amount of Cl− existing in biological system.

2. Experimental Absorption spectra were recorded on Shimadzu-2501 UV–Vis spectrophotometer using 1-cm quartz cell. 1 H NMR spectra were carried out in DMSO-d6 on a Bruker Avance 400 MHz NMR and Varian Unity 300 MHz spectrometer using TMS as the internal standard. ESI mass spectra were obtained using a Waters ZQ4000/2695 LC–MS spectrometer. Infrared

634

F.-y. Wu et al. / Spectrochimica Acta Part A 65 (2006) 633–637

spectra were registered as KBr pellets on a Nicolet 5700 FTIR spectrometer. The receptor 1 (N-4-nitrobenzene-N -1 -anthraquinonethiourea) was obtained [23] by condensation of 1aminoanthraquinone with p-nitrobenzylisothiocyanate in THF and purified by recrystallization from absolute ethanol. The other compounds were synthesized using the similar method. The products were characterized by IR, 1 H NMR and ESI mass methods. The data were consistent with the proposed formula. 1: 1 H NMR: δ (ppm) = 7.02 (1H, d, J = 9 Hz), 7.37 (3H, d, J = 9 Hz), 7.48 (1H, d, J = 9 Hz), 7.83 (1H, d, J = 9 Hz), 8.21 (1H, d, J = 9 Hz), 8.33 (4H, d, J = 9 Hz), 10.25 (NH, s), 10.36 (NH, s); ESI mass: m/e calcd. for C21 H13 N3 O4 S [M + Na+ ] 426.06, found [M + Na+ ] 426. 2: 1 H NMR: δ (ppm) = 7.00 (3H, d, J = 8.6 Hz), 7.14 (1H, t), 7.32 (3H, t), 7.46–7.50 (5H, m), 9.74 (NH, s); ESI mass: m/e calcd. for C21 H14 N2 O2 S [M + Na+ ] 381.26, found [M + Na+ ] 381. 3: 1 H NMR: δ (ppm) = 7.56–7.62 (4H, m), 7.90–8.01 (5H, m), 8.22 (2H, d, J = 9 Hz), 10.30 (NH, s), 10.41 (NH, s); ESI mass: m/e calcd. for C17 H13 N3 O2 S [M + H+ ] 324.07, found [M + H+ ] 324. 4: 1 H NMR: δ (ppm) = 7.15 (t, 1H, J = 7.5 Hz), 7.37–7.40 (m, 2H, J = 4.5 Hz), 7.45 (d, 2H, J = 7.5 Hz), 7.56 (d, 2H, J = 5.5 Hz), 7.58–7.68 (m, 2H, J = 7.5 Hz), 9.95 (s, NH), 10.01 (s, NH); ESI mass: m/e calcd. for C13 H11 N3 O2 S [M + H+ ] 274.06, found [M + H+ ] 274. The tetrabutylammonium salts of the tested anions were products of Sigma Corp. Organic solvent was purified by distilling and checked to show no fluorescent impurity. 3. Results and discussion The UV–vis spectra of 1 in the presence and absence of anions were presented in Fig. 1. With Br− as an exception, addition of anions such as F− , CH3 CO2 − , H2 PO4 − , HSO4 − and Cl− to 1 solution made absorption spectra red shift, which was ascribed to the formation of hydrogen-bond anionic complex. The hydrogen-bond interaction between anion and the electron

Fig. 1. Absorption spectra of 1 (5.0 × 10−5 mol L−1 ) in the presence and absence of anions in acetonitrile. The concentrations of anions were 5.0 × 10−5 mol L−1 for F− and 1.0 × 10−4 mol L−1 for other anions.

donor of receptor increased the electron density in the donor group (thiourea group). This increase in charge density resulted in the red shift of absorption [16]. Actually introducing of one equivalent of F− with respect to the concentration of 1 causes 1 solution undergo a distinct color changes from colorless to yellow. However, addition of two equivalent of the corresponding anions such as CH3 CO2 − , H2 PO4 − , HSO4 − and Cl− led only to a new shoulder peak valued at ca. 360 nm. But no detectable color changes were observed even in the presence of a larger excess of ten equivalent of corresponding anions, which made it clear that 1 could highly and selectively sense F− over other anions. It was inferred that fluoride anion was one of the anions expected to give stronger hydrogen bonds due to its comparatively smaller size. Fig. 2 showed the absorption titration of 1 against F− . In the absence of anions, the spectrum of 1 was characterized by the presence of one band with maximum absorption at 304 nm (ε = 1.2 × 104 cm−1 mol−1 L). Upon addition of increasing amount of F− , the peak at 304 nm was blue shifted to 285 nm while a new peak appeared at 427 nm. Clear isosbestic points at 241 and 295 nm were observed, respectively, demonstrating the existence of a well-defined stoichiometry complex. The fact that the absorbance at 427 nm leveled off after introducing one equivalent of F− also indicated the formation of a 1:1 complex, as shown in the inset plot of Fig. 2. The long wavelength absorption band was due to the formation of hydrogen-bond complex between thiourea moiety and F− , which most probably affected the electronic properties of chromophore and facilitated the process of charge transfer from thiourea moiety to p-nitrobenzene. Introducing hydrogen-bond solvents such as ethanol or methanol into the yellow solution (the mixture of 1 and fluoride anion) led to the recovery of the absorption spectrum of 1 (shown in Fig. 3). This could be understood in terms of competition between the proton solvent molecule and F− for hydrogen-bond sites in 1 and suggested the hydrogen-bond nature of the anion-1

Fig. 2. The changes in the absorption spectra of 1 (5.0 × 10−5 mol L−1 ) upon titration with F− (n-Bu4 N+ salt) in acetonitrile. Arrow indicates increasing of F− concentration. The F− concentration varied from 0 to 5.5 × 10−5 mol L−1 with an interval of 5.0 × 10−6 mol L−1 . Inset shows the absorbance at 427 nm as a function of F− concentration.

F.-y. Wu et al. / Spectrochimica Acta Part A 65 (2006) 633–637

Fig. 3. Absorption spectra of the 1:2 (1 and F− ) mixture in acetonitrile in the presence of various amounts of ethanol. Arrows show the direction of increasing amount of addition of ethanol. 1 concentration is 5.0 × 10−5 mol L−1 and the F− concentration is 1.0 × 10−4 mol L−1 .

interaction. In order to understand the bonding mode of 1-F− , 1 H NMR study was also carried out. In DMSO-d6 at 25 C◦ , the 1 H NMR signals of the thiourea –NH proton of 1 were at 10.25 and 10.36 ppm, respectively. The signal of NH appearing at downfield showed the high acidity and strong hydrogen-bonding ability. Upon addition of one equivalent of F− , the signal of NH disappeared. This result further indicated the formation of hydrogen-bond complex between 1 and F− [14,16,29]. To elucidate the substituent effect on the binding ability of receptor to anion, absorption titration of 2 against various anionic guests were also carried out. Upon addition of anions such as F− , CH3 CO2 − and H2 PO4 − , the spectra of 2 were only slightly red shifted from 278 to 279 nm though the absorbance increased. No new peak emerged. It exhibited that the nitro group played a fundamental role in inducing the intramolecular charge transfer process in the ground state and in increasing the acidity of NH in thiourea group. Hence the existence of electron withdrawing group such as nitro group enhanced the binding ability of receptor to anions and facilitated the process of charge transfer in the ground state bringing a vivid color change. Control compounds 3 and 4 were prepared to investigate the effect of the chromogenic signaling subunit on binding ability between receptor and anions. Figs. 4 and 5 showed the absorption spectral changes of 3 and 4 in the presence and absence of anions, respectively. Upon addition of anions, the absorption spectrum of 3 was red shifted with no obvious color change in the solution. However, in the case of receptor 4, it showed distinct spectral profile changes and the solution turned from colorless to yellow. The experiment results demonstrated that in contrast to 3, receptor 4 was highly sensitive to CH3 CO2 − . Compound 3 contains naphthyl group attached to the thiourea moieties. The contributing to the bulkiness of the rings makes it difficult for the receptor to bond with oxygenic anions such as CH3 CO2 − and H2 PO4 − [30]. All titration curves gave a satisfactory fit to a 1:1 binding model and the association constants (Kass ) of anion complexes

635

Fig. 4. Absorption spectra of 3 (2.0 × 10−5 mol L−1 ) in the presence and absence of anions (5.0 × 10−5 mol L−1 ) in acetonitrile.

Fig. 5. Absorption spectra of 4 (1.6 × 10−5 mol L−1 ) in the presence of anions (2.0 × 10−5 mol L−1 ).

were presented in Table 1. They were calculated by non-linear fitting [31] of the absorbance at the maximum wavelength of hydrogen-bond complexes as a function of anion concentration. The selectivity trend in binding affinities of anions for 1–2 was as follows: F−  CH3 CO2 − > H2 PO4 − > HSO4 − > Cl− > Br− , which was not completely in accord with the basicity of anionic guests. Nevertheless, the associate constants for 3–4 with anions were in the order of CH3 CO2 − > F− > H2 PO4 − , which was in Table 1 Association constants (L mol−1 ) of anion complexes in acetonitrile Anion

1

2

3

4

F−

3.39 × 106

1.89 × 104

2.20 × 104

6.76 × 103 2.00 × 105 8.76 × 103 n.d. n.d. n.d.

CH3 CO2 − H2 PO4 − HSO4 − Cl− Br−

4.34 × 104 3.94 × 104 1.45 × 104 4.32 × 103 n.d.

1.37 × 104 8.46 × 103 n.d. n.d. n.d.

5.90 × 104 4.71 × 103 n.d. n.d. n.d.

n.d. The association constants is too small to determine.

636

F.-y. Wu et al. / Spectrochimica Acta Part A 65 (2006) 633–637

Scheme 1. Structure of receptors 1–2 and 1:1 hydrogen-bond complex between 1 and fluoride anion.

Scheme 2. Structure of control compounds 3–4 and 1:1 hydrogen-bond complex between 3 and acetate anion.

good agreement with the basicity of anions. The different selectivity trends of 1–2 and 3–4 were assumed to be dependent on a conformation conversion of receptor. AM1 calculations indicated that the most stable configurations of 1–4 existed in a Z–Econformation and a conversion from the Z–E-conformation to the Z–Z-conformation must have occurred in order to bond with anions [32–34]. A six-membered ring formed via intramolecular hydrogen-bond interaction in 1 and 2 (the left in Scheme 1) and ˚ between the bond length was estimated to be around 2.071 A the oxygen atom of anthraquinone and the hydrogen atom of the thiourea moiety, which suggested that the hydrogen bond was very strong. Therefore it was difficult for 1 and 2 to convert from Z–E-conformation (the left in Scheme 2) to Z–Z-conformation (the right in Scheme 2) in the presence of anions. This made receptors (1 and 2) unfavorable in bonding with oxygenic anion substrates such as CH3 CO2 − and H2 PO4 − via multi-hydrogen bonds, which was assumed to be the reason for a high selectivity sensing of fluoride anion over other anions. But for 3 and 4, the conversion is easier than that of 1–2, so association constants between them and acetate anion were much larger than that with fluoride anion.

tive discrimination of F− , CH3 CO2 − and H2 PO4 − with similar basicity by naked-eye method. The selectivity trends in binding affinities of anions for 1 followed in the order of F−  CH3 CO2 − > H2 PO4 − > HSO4 − > Cl− > Br− . The observed binding sequence was not completely consistent with the anion basicity. It was reasoned that fluoride anion was one of the anions expected to give stronger hydrogen bonds due to its comparatively smaller size. AM1 calculations indicated that the most stable configuration of 1 was Z–E-conformation with six-membered ring via intramolecular hydrogen bond. This made it in the unfavorable conformation for thiourea moiety to bind oxygen-anionic substrates such as CH3 CO2 − and H2 PO4 − via multi-hydrogen bonds, which resulted in a high selectivity for the detection of F− .

4. Conclusions

References

In summary, we obtained a structurally simple colorimetric sensor (1) based on thiourea allowing for selec-

Acknowledgements The authors gratefully acknowledge the financial support of this study by the Jiangxi Province Natural Science Foundation (JXNSF No. 0420041) and Jiangxi Province Education Ministry Foundation (No. 2005-38).

[1] F.P. Schmidtchen, M. Berger, Chem. Rev. 97 (1997) 1609–1646. [2] T.S. Snowden, E.V. Anslyn, Curr. Opin. Chem. Biol. 3 (1999) 740–746.

F.-y. Wu et al. / Spectrochimica Acta Part A 65 (2006) 633–637 [3] [4] [5] [6] [7]

[8] [9] [10]

[11] [12] [13]

[14] [15] [16] [17]

P.D. Beer, P.A. Gale, Angew. Chem. Int. Ed. Engl. 40 (2001) 486–516. J.L. Sessler, J.M. Davis, Acc. Chem. Res. 34 (2001) 989–997. R. Mart`ınez-M´anˇ ez, F. Sancen´on, Chem. Rev. 103 (2003) 4419–4476. T.J. Jentsch, Curr. Opin. Neurobiol. 6 (1996) 303–310. J.M. Tomich, D. Wallace, K. Henderson, K.E. Mitchell, G. Radke, R. Brandt, C.A. Ambler, A.J. Scott, J. Grantham, L. Sullivan, T. Iwamoto, Biophys. J. 74 (1998) 256–267. H. Miyaji, W. Sato, J.L. Sessler, Angew. Chem. Int. Ed. 39 (2000) 1777–1780. P. Piateka, J. Jurczak, Chem. Commun. (2002) 2450–2451. C.J. Woods, S. Camiolo, M.E. Light, S.J. Coles, M.B. Hursthouse, M.A. King, P.A. Gale, J.W. Essex, J. Am. Chem. Soc. 124 (2002) 8644– 8652. D.H. Lee, J.H. Im, J.H. Lee, J.-In. Hong, Tetrahedron Lett. 43 (2002) 9637–9640. S.J. Coles, J.G. Frey, P.A. Gale, M.B. Hursthouse, M.E. Light, K. Navakhun, G.L. Thomas, Chem. Commun. (2003) 568–569. S. Yun, H. Ihm, H.G. Kim, C.W. Lee, B. Indrajit, K.S. Oh, Y.J. Gong, J.W. Lee, J. Yoon, H.C. Lee, K.S. Kim, J. Org. Chem. 68 (2003) 2467– 2470. D.A. Jose, D.K. Kumar, B. Ganguly, A. Das, Org. Lett. 6 (2004) 3445– 3448. M. Boiocchi, L.D. Boca, D.E. G´omez, L. Fabbrizzi, M. Licchelli, E. Monzani, J. Am. Chem. Soc. 126 (2004) 16507–16514. V. Thiagarajan, P. Ramamurthy, D. Thirumalai, V.T. Ramakrishnan, Org. Lett. 7 (2005) 657–660. D. Jim´enez, R. Mart´ınez-M´an˜ ez, F. Sancen´on, J. Soto, Tetrahedron Lett. 43 (2002) 2823–2825.

637

[18] P. Anzenbacher Jr., K. Jurs´ıkov´a, J.L. Sessler, J. Am. Chem. Soc. 122 (2000) 9350–9351. [19] H. Luecke, F.A. Quiocho, Nature 347 (1990) 402–406. [20] J.J. He, F.A. Quiocho, Science 251 (1991) 1479–1481. [21] D. Curiel, A. Cowley, P.D. Beer, Chem. Commun. (2005) 236–238. [22] T. Gunnlaugsson, A.P. Davis, J.E. O’Brien, M. Glynn, Org. Biomol. Chem. 3 (2005) 48–56. [23] R. Kato, S. Nishizawa, T. Hayashita, N. Teramae, Tetrahedron Lett. 42 (2001) 5053–5056. [24] T. Hayashita, T. Onodera, R. Kato, S. Nishizawa, N. Teramae, Chem. Commun. (2000) 755–756. [25] K. Kobiro, Y. Inoue, J. Am. Chem. Soc. 125 (2003) 421–427. [26] K.H. Lee, H.-Y. Lee, D.H. Lee, J.-In. Hong, Tetrahedron Lett. 42 (2001) 5447–5449. [27] L. Fabbrizzi, F. Foti, A. Taglietti, Org. Lett. 7 (2005) 2603–2606. [28] H. Miyaji, J.L. Sesseler, Angew. Chem. Int. Ed. 40 (2001) 154–157. [29] J.Y. Lee, E.J. Cho, S. Mukamel, K.C. Nam, J. Org. Chem. 69 (2004) 943–950. [30] P. B˝ulmann, S. Nishizawa, K.P. Xiao, Y. Umezawa, Tetrahedron 53 (1997) 1647–1654. [31] B. Valeur, J. Pouget, J. Bourson, M. Kaschke, N.P. Ernsting, J. Phys. Chem. 96 (1992) 6545–6549. [32] J.L.J. Blanco, J.M. Benito, C.O. Mellet, J.M.G. Fernandez, Org. Lett. 1 (1999) 1217–1220. [33] K. Kumamoto, Y. Misawa, S. Tokita, Y. Kubo, H. Kotsuki, Tetrahedron Lett. 43 (2002) 1035–1038. [34] F.Y. Wu, Z. Li, Z.C. Wen, N. Zhou, Y.F. Zhao, Y.B. Jiang, Org. Lett. 4 (2002) 3203–3205.