Mechanochemical synthesis of 2,2-difluoro-4, 6-bis(β-styryl)-1,3,2-dioxaborines and their use in cyanide ion sensing

Introduction

The design and development of 1,3,2-dioxaborine based fluorophores is a growing area of research with applications in sensing and biomedical imaging [1–4]. 1,3-Diketones form stable 2,2-difluoro-1,3,2-dioxaborines, also termed as difluoroboron complexes, that exhibit interesting photophysical properties. Curcuminoids are 1,3-diketones, existing preferentially as enol tautomers [5], and belong to the diarylheptanoid class of natural products. These (1E,4Z,6E)-1,7-diaryl-5-hydroxyhepta-1,4,6-trien- 3-ones have served as useful precursors for the synthesis of 2,2-difluoro-1,3,2-dioxaborine scaffold based cyanide sensing fluoroprobe [6], NIR probes for amyloid β-deposits [7–9], inhibitors for HIV proteases [10] and NIR luminophores [11]. Recently, BF₂ complexes of π-extended curcuminoids have been described that show solvatochromic and solvatofluorochromic properties [12].

Reported methods for the synthesis of curcuminoid-BF₂ complexes include the reaction of the natural curcuminoids, or synthetic ones obtained from arylaldehyde, boric acid and 1,3-pentanedione, with boron trifluoride etherate. These complexes can also be prepared by condensation of arylaldehydes with 2,2-difluoro-1,3-dioxaborylpentadione. Both routes however, afford these complexes in generally low yields. For example, Ran and co-workers [7, 8] have reported yields in the range of 14–20% for 2,2-difluoro-1,3,2-dioxaborines obtained by the bis-condensation of arylaldehydes with 2,2-difluoro-1,3-dioxaborylpentadione at 60°C overnight. More recently, Bai et al. [12] have prepared BF₂ complexes, including π-extended systems, by reacting arylaldehydes, hetarylaldehydes and cinnamaldehydes with 2,2-difluoro-1,3-dioxaborylpentadione in toluene at room temperature for 2 h in 15–45% yield. Feloaut and co-workers [13] obtained two curcuminoid-BF₂ complexes in 61 and 74% yield by reacting the curcuminoids and boron trifluoride etherate in dichloromethane at reflux for 24 h. In a similar way, a long reaction time has been reported by Sui and co-workers [10] in seemingly the first report of the synthesis of a curcumin-BF₂ complex. With our interest in the synthesis and bioactivity of curcuminoids [14–16], we now report a mild two-stage synthetic route to access curcumin templated 1,3,2-dioxaborines.

Results and discussion

The (1E,4Z,6E)-1,7-diaryl-5-hydroxyhepta-1,4,6-trien-3-ones 1a–i were prepared by microwave irradiation of a mixture of arylaldehyde, acetylacetone and boric acid in a mixed solvent consisting of a small amount of N,N-dimethylformamide (DMF), glacial acetic acid and 1,2,3,4-tetrahydroquinoline (THQ) as catalyst for 20 min. This method is a great improvement over the method reported previously which required heating at 85°C for 4 h [14]. Among different solvents tried, including DMF, dimethyl sulfoxide, N-methylpyrrolidine and N,N-dimethylacetamide, DMF was the best and with various amines tried as catalyst, THQ was found to work best, followed by diethanolamine, on the basis of yield and purity of the crude product. This use of THQ-acetic acid in curcumin synthesis has been adopted by others [9, 17]. We also observed that boric acid could also serve as the boron source and the addition of a tri-n-butyl borate did not improve the yield [9, 17]. The reaction mixture was worked up by a dropwise addition to 20% aqueous acetic acid under very rapid stirring to afford a crude product, which was further purified by planar preparative chromatography (PTLC). The curcuminoids 1a–i, obtained in yields of 61–79%, were then efficiently converted to the corresponding BF₂ complexes in yields of 87–93% by neat grinding of these compounds with BF₃-etherate for a few minutes (Scheme 1 and Table 1). The resultant complexes 2a–i were characterized by FTIR, ¹H NMR, ¹³C NMR and HRMS. The products 2b–d, 2h and 2i have not been described previously. The synthesized BF₂ complexes are excellent D-A-D molecules having electron donor (D) moieties at both ends of the conjugated system and an electron acceptor (A) moiety in the middle of the molecule.

Scheme 1

Synthetic route for curcuminoid 1,3,2-dioxaborines.

Chaicham and co-workers have reported the cyanide ion sensing ability of curcumin-BF₂ complex 2a. These workers have attributed the selective detection of cyanide to the deprotonation of the two phenolic groups, followed by an intramolecular charge transfer process, as well as to the basicity of cyanide ion [6]. We examined the cyanide anion sensing ability of 2a–i and concluded that a hydroxy group on the aryl rings, para to the intervening seven carbon bridge, is a structural feature that imparts a good cyanide ion sensing ability.

Thus, the 2,2-difluoro-1,3,2-dioxaborines 2a and 2b, both having a 4-hydroxyaryl moiety, show an excellent ability to sense cyanide ion under the conditions of this work (Figures 1 and 2). Between compounds 2a and 2b, the former compound exhibits a stronger response. Since the response of 2a to cyanide ion is stronger than that of 2b, factors other than pKa of the phenolic hydroxyl group may also play a role in the sensing mechanism, since the phenolic group in 2b would be expected to be deprotonated easier. The compound 2c, with a 3-hydroxyaryl group, and 2d with a 2-hydroxyaryl group respond very poorly, whereas those 2,2-difluoro-1,3,2-dioxaborines with no hydroxyaryl group, such as 2e–i, do not respond markedly. The sensing of cyanide anion by 2,2-difluoro-1,3,2-dioxaborines 2a and 2b is insensitive to the presence of other anions including chloride, fluoride, phosphate, nitrite, nitrate and sulfate (Figure 3).

Figure 1

Absorption spectra of 5×10^-6m solutions of curcuminoid 2,2-difluoro-1,3,2-dioxaborines 2a–i in MeCN/H₂O (4:1).

Figure 2

Absorption spectra of 5×10^-6m solutions of curcuminoid 2,2-difluoro-1,3,2-dioxaborines 2a–i in presence of 100 molar excess of cyanide ion in MeCN/H₂O (4:1).

Figure 3

Absorption spectra of 5×10^-6m solutions of 2,2-difluoro-4,6-bis[β-(4-hydroxy-3-methoxystyryl)]-1,3,2-dioxaborine (2a) in presence of 100 molar excess of various ions in MeCN/H₂O (4:1).

We also examined the response of the curcumin-BF₂ complex 2a to pH using MeCN/H₂O solution in the pH range 1–14 (Figure 4). This study revealed that a change in pH is paralleled by a change in color. This observation indicates that the basis for cyanide ion sensing by curcumin-BF₂ complex 2a is probably due to an acid-base reaction between cyanide ion and 2a, as suggested by Chiacham and co-workers [6]. It appears that in order to analyze the presence of cyanide ion using 2a as a sensor, the pH of the analyte aqueous solution should preferably be in the range of 7–9.

Figure 4

Absorption spectra of 2,2-difluoro-4,6-bis[β-(4-hydroxy-3-methoxystyryl)]-1,3,2-dioxaborine (2a) in MeCN/H₂O (4:1) solutions of pH range 1–14.

Conclusions

In summary, we have synthesized several curcuminoids and converted these compounds to 2,2-difluoro-4,6-bis[β-(styryl)]-1,3,2-dioxaborines with a D-A-D motif under mild conditions in a very good overall yield. The structural feature required for conferring a cyanide ion sensing ability by these compounds is a 4-HOC₆H₄ functionality.

Experimental

Reagents and solvents used were of analytical grade. Analytical thin layer chromatography (TLC) and preparative thin layer chromatography (PTLC) were performed on glass plates (20 mm×75 mm and 20 cm×20 cm, respectively) coated with TLC grade silica gel-G (E. Merck, India). Preparative chromatography was conducted using dry columns as reported earlier by us [14]. The spots or bands were visualized directly or under UV light. The elemental analyses were carried out on a Vario EL elemental analyzer. The ¹H and ¹³C NMR spectra were recorded on Bruker Topspin and Bruker AV spectrometers at room temperature. HRMS were recorded on JEOL JMS 600H mass spectrometer. The UV-Vis absorption spectra of the complexes in the presence and absence of ions were recorded on a HALO DB-20 UV-Vis spectrophotometer.