Colorimetric sensing for fluoride.
Chapter 4
Colorimetric sensing for fluoride.
4.1aIntroduction.
The recognition and sensing of anions has emerged as a key area of interest in supramolecular chemistry and consequently it has been subjected to considerable research effort.1a,b In particular fluoride recognition has received great attention, not only due to its established role in many biological processes2a-d, but also for its potential in industrial applications.2e
Binding of anions is generally achieved employing various combination of pyrroles, guanidiniums, Lewis acids, amides and urea/thioureas moieties.1a,3 However a sensor is something more than a receptor. The binding has to be followed by the generation of a detectable signal which reveals the presence of the targeted analyte and which permits to determine its concentration. To a binding site therefore has to be added a second site which is sensitive to the presence of the analyte. Redox or photoactive sensors have been developed. Their coordination to anions is detected by changes in the physical properties of the host. For example the ditopic boroncontaining Lewis acid 1 binds two equivalents of fluoride in CHCl3 or CH2Cl2. The binding can be monitored by the 1H-NMR, but more easily, if the sample is exposed to air the ferrocene moiety is oxidized to ferrocenium and the solution colour changes from orange to pale green. Only the fluoride complexed specie is sensible to the molecular oxygen and other anions like the halogens, H2PO4-, HSO4-, BF4-, PF6-, or NO3- are not complexed. The process can be followed in Figure 1.4
Figure 1: Selective boron-based receptor 1.
Boron is a hard Lewis acid and it is not surprising that it can bind fluoride. Thus boronbased compound are widely used. If a boronic acid group is covalently linked to a cromophore or luminophore a selective fluorescent probe for fluoride can be obtained. It is the case of 2 (Fig. 2 left) which displays highly intense emission spectra in the visible region associated with the binding of fluoride (K ~ 103 M-1), with high selectivity over Br- and Cl- in a 2:1 water-methanol solution. Emission profiles are shown in Figure 2, right.
Figure 2: Fluorescent probe for the recognition of fluoride 2, left and emission spectra, right.
More elaborate structures can be thought in order to improve the binding. The trianthranylborane 3 (Fig. 3a) shows high affinity towards fluoride in THF (K=2.80.3x105 M1), but the most important features in this system consists in a dramatic change in the spectroscopic properties of the host upon complexation (Figure 3b).5 The planarity of the host, which allows for an effective orbital overlap and for absorption in the visible region, is lost in the fluoride-bound adduct and a marked colour change is therefore visible by the “naked-eye”, as shown in Figure 3c.
Figure 3: Trianthranylborane 3,loss of conjugationand colour change upon fluoride binding.
Naked-eye sensors are very valuable systems. They provide a direct detection without the use of any spectroscopic instrumentations, but to date only few examples are reported in the literature. In this context Sessler’s work is worth noting.6a-e Dipyrrollylquinoxaline 4a was found to bind F- with K=1.8x104 M-1 in dichloromethaned2, yet higher affinities are desirable for real applications. The introduction of strong electron withdrawing substituents enhances the binding, boosting the association constant (K= 6.1x104 M-1). 4b retains such binding properties even in DMSO. 4a and 4b have the advantage of possessing a built-in chromophore and readily accessible from a synthetic point of view (Figure 4a). The fluorophore can be also appended to the host non-covalently. Compound 5 is a Cobalt complex derived from 4a. In this case the electron withdrawing group is the coordinated metal center.6a This compound displays an enhanced selectivity for F- over commonly interfering anions such as H2PO4-, being KF/KH2PO4 = 103. Moreover, the spectral changes upon complexation are so intense that it can be employed as a naked-eye sensor for F-, as documented in Figure 4b.
Figure 4: a) Receptors 4a-b and 5 b) spectral changes upon addition of several anions to a host 5 solution in DMSO,[5]=5x10-5 M.
Colometric ON-OFF switches can be obtained if the cromophore moiety can act as hetero ditopic receptor. 6 is a good example of that.7 It works in acetonitrile and it alone is colourless. The addition of 10 eq. of F- results in a clear change to yellow colour in the solution (ON signal). The OFF signal can be induced by adding K+ ions (Figure 5).
Figure 5: Ditopic receptor 6 (left) and colour changes in a 2.5x10-4 M solution of 6 after addition of fluoride and then potassium K+(right).
These examples demonstrate that recognition of fluoride is an extremely active area of research, yet the main goal is to achieve such recognition in water with high selectivity. To the best of our knowledge such ambition remains still unfulfilled.
4.2 Salophen-UO2 based Colorimetric Sensing of Fluoride.
A salophen-UO2 complex acts as Lewis acidic center embedded in an organic framework. UO2 is an hard species and its salophen-complexes have been documented as effective receptors for hard anions.8 Most of the work on this systems have been carried out by Reinhoudt’s group.9 However detailed spectroscopic studies have not been undertaken. We therefore carried out a spectroscopic analysis of the properties of such receptors starting from an interesting finding. In Figure 5a the picture of two vials containing the host C alone (1x10-3 M,10 left) and in the presence of TBAF (right) in DMSO is reported. The neat colour variation from dark orange to yellow is evident by the naked-eye. This finding led us to embark upon a broad screening on several anionic species to evaluate the selectivity of receptor C towards them. The advantage of receptor C is that it can be synthesized in gram quantities with ease. The salophen ligand is a chromofore itself and it is clear that any change on the coordinative environment of the uranyl center can induce energetic variations on the metal-ligand electronic transitions. Preliminary studies in acetonitrile demonstrated that the association of C with F- is practically quantitative. However, in these conditions the association with chloride is quite strong as well. Therefore, to increase the selectivity among halides, DMSO has been chosen as the solvent for such studies. In Figure 5b the titration profile between C and TBA F is shown.
Figure 5: a) colour change of a 1 mmolar solution of C upon addition of F-; b) spectrometric profile of the titration of C with TBAF.
The addition of increasing amount of fluoride (the TBA cation is considered to be solvated and not interfering with the C-F- association) produces the appearance of a spectral band around 420 nm and the decrease in intensity of the 340 nm band along with a small ipsochromic shift. Three clear isosbesthic points are found at 298, 390 and 470 nm. Even if the presence of isosbestic points commonly suggests the formation of a 1:1 complex, this is not the case. Multiple equilibria are present as highlighted by the titration profile. In Figure 6 A() vs. fluoride concentration have been plotted for selected wavelengths. A slight sigmoidal behaviour is found.
Figure 6: Titration profiles and data points for selected wavelengths (400-450 nm)
The salophen ligand in receptor C does not bear any substituent or group that could hinder the complexation site. After the formation of the C-F- complex, in presence of an excess of receptor C, a second species (C2F-) can be formed as shown in Scheme 1. This behaviour has been already documented in similar cases with related compounds by NMR.11 Unfortunately it has not been possible to obtain good fit for such behavior and only qualitatively informations have been obtained for K1.
Scheme 1: Multiple equilibria involving C and fluoride ion.
No spectral changes after the addition of 200 eq. of TBACl, TBABr and TBAI have been observed. It can therefore be concluded that in DMSO receptor C is selective for F- over the other halogens. A broader screening is however necessary. For this purpose other anions as their TBA or K salt form have been tested. No spectral changes upon addition of a large excess of SCN-, PF6-, HSO4- and ClO4- have been found. The situation is different with AcO- and H2PO4-. They induce evident changes on the absorption spectrum of C. The UV-vis profiles are reported in Figure 7 in comparison with fluoride.
Titration experiments have been therefore carried out to quantitatively assess the association constants for such interfering anions. The results compared with those obtained for fluoride are shown in Table 1. Dihydrogenphosphate and acetate displays a 1:1 binding stoichiometry with C and no secondary equilibria are observed. The data shows that no selectivity for fluoride exists over H2PO4-, but the association with acetate is 5 times weaker than with F-.
Table 1: Association constant K (M-1) in DMSO at 25°C for host C.H2PO4- / AcO- / F-
11000700 / 3000100 / 14000a
a: calculated considering 98% of complexation in the spectral profile plateau
Obviously a major goal in such area of research consists in the recognition in water. Our Salophen-UO2 receptors are not soluble in such medium. However we discovered that suitable micellar systems can interact with receptors such as C and bring themin aqueous media. Hence, it was interesting to test the capability of C to bind fluoride in such conditions. A certain quantity of receptor C has been added to a 10mM solution of CTABr (cetyltrimethylammonium bromide) in water. The initial uncoloured solution gains a yellowish colour due to the presence of C. 1H-NMR and UV-vis spectra of such sample confirms the presence of the salophen-UO2 receptor in solution. The profile of the titration with TBAF is shown in Figure 8.
Figure 8: Spectral profile for the association of C with fluoride in the presence of CTABr micelles and titration profiles for selected wavelength (400-450 nm).
Again a slight sigmoidal behaviour prevents from an optimal fit of the experimental data. Therefore only a rough estimate of K=4500 M-1 for the binding constant is given. Its value is calculated considering the equilibrium completed for 98% in the plateau and ignoring any species but the 1:1 complex. The complexation with fluoride is confirmed by 1H-NMR. A moderate shift on the host signals upon addition of such anion is observed (Figure 9).
Figure 9: bottom spectrum: C signals in D2O with the presence of 10mM CTABr; upper spectrum: after the addition of an excess of TBAF.
Such receptors lack of selectivity towards hard anions. To improve this aspect the steric approach followed by Severin et al. could be taken into account for future improvements.12 Selectivity based only on steric requirements can be accessible to modified host based on the C structure. Work is in progress.
The experimental part concerning the synthesis of the receptor C employed in this chapter has been already reported (see chapter 2.6).
4.3 Conclusions.
In this study we focused on the spectroscopic properties of Salophen-UO2 receptor C in the binding of several anionic species. The data collected complemented the informations on such class of receptors already reported in literature and revealed that colorimetric sensing of fluoride can be achieved in millimolar concentration in DMSO with no interference by the halogen series and by other anions such as SCN-, PF6-, HSO4-and ClO4-. The simplereceptor C, however, binds with moderate to high affinities harder anions such as AcO- and H2PO4-. A different stoichiometry is displayed in the binding of fluoride. Recognition can be performed even in water. The supramolecular species formed between CTABr micelles and C is able to bind fluoride with moderately high association constant.
References:
[1]: a) J. W. Steed and J. L. Atwood, in Supramolecular Chemistry, Wiley, NY, 2000, pp 198-249 see also b) A. Bianchi, K. Bowman-James and E. Garcia-España (eds.) Supramolecular Chemistry of Anions, Wiley-VCH, NY, 1997
[2]: a) K. Kirk, Biochemistry of the Halogens and Inorganic Halides, Plenum Press: NY, 1991, p.58; b) B. L. Riggs, Bone and Mineral Research, Annual 2, Elsevier, Amsterdam, 1984, pp366-393.; c) A. Wiseman, Handbook of Experimental Pharmacology XX/2, Springer-Verlag, Berlin 1970, pag. 48-97; d) R. H. Dreibuch, Handbook of Poisoning, Lange Medical Publishers, Los Altos CA, 1980; e) J. Nomura, H. Imai and T. Miyaki in Emerging Technologies in Hazardous Waste Management, D. W. Tedder and F. G: Pohland (eds.), ACS, Symp. Ser. No..422, American Chemical Society, Washington, 1990.
[3]: Chemosensor of Ions and Molecular Recognition, J. P. Desvergne and A. W. Czarnik (eds.), Kluwer, Dordrecht, 1997, vol 492.
[4]: S. Aldridge, C. Bresner, I. A. Fallis, S. J. Coles and M. B. Hursthouse, Chem. Commmun.2002, 740-741.
[5]: S. Yamaguchi, S. Akiyama and K. Tamao, J. Am. Chem. Soc., 2001, 123, 11372-11375.
[6]: a) T. Mizuno, W. –H. Wei, L. R. Eller and J. L. Sessler, J. Am. Chem. Soc., 2002, 124, 1134; b) H. Miyaji and J. Sessler,Angew. Chem. Int. Ed. Engl., 2001, 40, 154; c) H. Miyaji, W. Sato and J. L. Sessler, Angew. Chem. Int. Ed. Engl., 2000, 39, 1777-1780; d) C. B. Black, B. Andrioletti, A. C. Try, C. Ruiperez and J. L. Sessler, J. Am. Chem. Soc., 1999, 121, 10438; e) P. A. Gale, L. J. Twyman, C. I. Hadlin and J. L. Sessler, Chem. Commun., 1999, 1851-1852.
[7]: H. Miyaji, S. R. Collinson, I. Prokes and J. H. R. Tucker, Chem. Commun., 2003, 64-65.
[8]: M. M. G. Antonisse and D. N. Reinhoudt, Chem. Comm., 1998, 443-448.
[9]: A. C. Ion, I. Ion, M. M. G. Antonisse, B. H. M. Snelink-Rüel, and D. N. Reinhoudt, Russ. J. Gen. Chem., 2001, 71, 159-161.
[10]: The concentration tested is in accordance with some related papers (see for example D. Jiménez, R. Martinez-Mañez, F. Sanncenòn and J. Soto, Tetrahedron Lett., 2002, 2823-2825), but is significatevely higher than related work by Sessler (reference 7a).
[11]: M. M. G. Antonisse, Anion Recognition and Sensing, 1998, PhD Thesis.
[12]: M. L. Lehaire, R. Scopelliti, H. Piotrowsky and K. Severin, Angew. Chem. Int. Ed. Engl., 2002, 41, 1419-1422.
1