Preparation of anthracene-based tetraperimidine hexafluorophosphate and selective recognition of chromium(III) ions

A novel anthracene-based tetraperimidine hexafluorophosphate 3 was prepared, and its structure was determined through X-ray analysis, HRMS as well as 1H and 13C NMR spectroscopy. In the cationic moiety of 3, two (N-ethylperimidinyl–C2H4)2NCH2– arms were attached to the 9- and 10-positions of anthracene. In addition, compound 3 was used as a chemosensor to research the ability to recognize Cr3+ through fluorescence and UV titrations, HRMS, as well as 1H NMR and IR spectroscopy. The results indicate that 3 is an effective chemosensor for Cr3+.


Introduction
Fluorescent chemosensors are an attractive and efficient tool for the detection of metal ions in environmental and biological science because of their high sensitivity, selectivity, and simple usage [1][2][3][4][5][6][7][8]. Among metal ions, the detection of Cr 3+ ions occupies an important position. Chromium(III) is an essential microelement for humans and animals, and it plays an important role in glucose and lipid metabolism in the body [9,10]. The deficiency of chromium(III) in the human body leads to various diseases, such as diabetes as well as autoimmune and cardiovascular diseases [11]. On the other hand, the excessive incorporation of chromium(III) is toxic to humans, and can cause cancer through the oxidation of DNA and some proteins [12][13][14]. Therefore, the detection of chromium(III) has a vital practical significance for human health monitoring.
In the process of our research, a tetradentate compound bearing a fluorophore aroused our interest. In this paper, we report the synthesis of a novel anthracene-based tetraperimidine hexa-Scheme 1: Synthetic route to compound 3. fluorophosphate 3, and its structure was determined by X-ray analysis as well as 1 H and 13 C NMR spectroscopy. Particularly, compound 3 was tested as a chemosensor for the recognition of Cr 3+ through fluorescence, UV, IR, and 1 H NMR spectroscopy along with HRMS. Altogether, the results indicate the utility of 3 as an effective chemosensor for Cr 3+ .

Results and Discussion
Synthesis and characterization of 3 As displayed in Scheme 1, paraformaldehyde was reacted with anthracene to give 9,10-di(chloromethyl)anthracene in 82% yield, which reacted further with HN(CH 2 CH 2 OH) 2 to form 9,10-bis{[N,N-di(2-hydroxyethyl)amino]methyl}anthracene (1) with a yield of 58% [41]. Compound 1 was then treated with SOCl 2 to generate 9,10-bis{[N,N-di(2-chloroethyl)amino]-methyl}anthracene (2) in 75% yield, which reacted with 1-ethylperimidine in the presence of KI to afford the analogous iodide salt to tetraperimidine 3. Subsequently, an anion exchange reaction with NH 4 PF 6 was performed to generate tetraperimidine hexafluorophosphate 3 with a yield of 85%. Compound 3 was stable to heat, moisture, and air, and it had a good solubility in DCM, DMSO, and CH 3 CN. In turn, it had a poor solubility in benzene and petroleum ether. In the 1 H NMR spectrum of 3, the proton signal corresponding to the NCHN motif in perimidine was present at δ = 8.69 ppm [42].
To further investigate the recognition of Cr 3+ by 3, fluorescence titrations were carried out ( Figure 3). The fluorescence intensity of 3 in the region of 388-500 nm increased gradually upon addition of Cr 3+ (c = 5.0⋅10 −6 M). Titration was continued until no more notable changes in emission intensity occurred. In the inset of Figure 3, it is shown that when the molar ratio of Cr 3+ to 3, i.e., c Cr 3+/c 3 , was below 1:1, fluorescence intensity enhanced sharply. However, when the molar ratio exceeded 1:1, the rate of fluorescence enhancement gradually slowed down until no more changes were noticeable. The limit of detection (LOD) was calculated to be 2.33⋅10 −7 M (Supporting Informa-tion File 1, Figure S3). This value is analogous to the lowest corresponding value that has been reported in the literature (9.40⋅10 −7 -5.55⋅10 −6 M) [44][45][46]. The association constant K SV was calculated to be 6.6⋅10 4 M −1 (R = 0.998) for 3·Cr 3+ using Equation 1 (Supporting Information File 1, Figure S4) [47]. (1) In Equation 1, the fluorescence intensities of 3 in the presence of Cr 3+ and in its free form are represented by F and F 0 .
In the UV titration experiments, the absorption band in the region of 240-265 nm increased gradually upon addition of Cr 3+ to a solution of 3 (c = 5.0⋅10 −6 M) in CH 3 CN/DMSO, 9:1, v/v at room temperature (Supporting Information File 1, Figure  S5). To evaluate the stability of 3·Cr 3+ , the stability constant K for the complex was computed as 8.23⋅10 4 M −1 (R = 0.999) at 258 nm using Equation 2 (Supporting Information File 1, Figure S6) [48][49][50][51]. ( In Equation 2, the absorbances of 3 in presence and absence of Cr 3+ are represented as A 0 and A. The discrepancy in absorbance in the presence and absence of Cr 3+ is represented through the expression A 0 − A (i.e., ΔA). The molar extinction coefficients of Cr 3+ and the complex 3·Cr 3+ are represented by ε r and ε c .
To measure the selectivity of Cr 3+ complexation by 3, displacement experiments were carried out (Supporting Information File 1, Figure S7). Firstly, 30 equiv of Cr 3+ were added to solutions of 3 containing 30 equiv of K + , Na + , Li + , Ag + , NH 4  In order to probe whether the anions of the chromium(III) salts had effects on the binding of 3 and Cr 3+ , other chromium(III) salts, CrCl 3 , CrBr 3 , Cr 2 (SO 4 ) 3 , Cr(NO 3 ) 3 , and Cr(OAc) 3 , were tested. As displayed in Figure S8, Supporting Information File 1, when 30 equiv of any of these were added to 3, similar fluorescence intensities could be detected. A reversible binding experiment was also carried out (Supporting Information File 1, Figure S9). Therein, 30 equiv of ethylenediaminetetraacetic acid (EDTA) were added to a solution of Cr 3+ (c = 1.5⋅10 −4 M) and 3 (c = 5.0⋅10 −6 M), which led to a reduction of the fluorescence intensity at 388-500 nm. This reduced fluorescent intensity was analogous to that of free 3, displaying that 3 was regenerated in its uncomplexed form. When Cr 3+ was added anew, the fluorescent intensity increased again. These results show that the binding process of 3 and Cr 3+ has good reversibility and highlights the regenerative capacity of the 3⋅Cr 3+ complex.

Interactions between 3 and Cr 3+
Looking at the structural characteristics of 3, the nitrogen atoms of the tertiary amines, and the π systems, were most likely the binding sites for Cr 3+ through Cr 3+ ···N and Cr 3+ ···π interactions (Scheme 2). In order to obtain further information on the binding pattern between 3 and Cr 3+ , 1 H NMR titration studies were done in DMSO-d 6 . The spectra are depicted in Figure 4.
Upon incremental addition of Cr 3+ to 3 (from 0.0 to 1.0 equiv), the proton signals a and b, corresponding to anthracene, shifted downfield by 0.03 ppm in total while the proton signals f and l, corresponding to perimidine, shifted downfield between 0.02 and 0.07 ppm. Further, the proton signals c, d, and e of the CH 2 groups beside perimidine and anthracence shifted downfield by 0.03 ppm while the chemical shifts of other protons did not undergo visible changes. These experimental results suggest Cr 3+ ···π interactions as the most likely binding mode between Cr 3+ and 3, as illustrated in Scheme 2. In 3, each perimidine moiety is electron-rich due to the existence of a bond (Sup-Scheme 2: Illustration of interactions between 3 and Cr 3+ in 3⋅Cr 3+ . porting Information File 1, Figure S10), and the strong affinity of the π system of each perimidine and the anthracene motif towards Cr 3+ resulted in the facile formation of Cr 3+ ···π interactions. It is worth noting that Cr 3+ ···π interactions are not uncommon, and they have been reported in diaryl chromium complexes [54,55]. Besides, Cr 3+ ···N interactions in 3·Cr 3+ did not appear relevant for complexation. The reasons were that (1) the signal m, corresponding to the CH 2 fragment beside the nitrogen atom of the tertiary amine, did not shift discernibly during 1 H NMR titration; (2) if the tertiary amine groups coordinated to one Cr 3+ ion each, a 1:2 binding mode would have been determined for 3⋅Cr 3+ ; and (3) from the molecular structure of 3, due to the distance, it is spatially impossible that one Cr 3+ ion is bound by both tertiary amine functions at the same time. All these arguments underline the absence of Cr 3+ ···N interactions. The selectivity of the Cr 3+ binding process by 3 may be mainly due to the metal ion's size, which could have been particularly suitable for coordination between anthracence and four perimidine groups, whereas the sizes of other metal cations were unsuitable. The fact that chromium(III) is triply charged may not have been a key influence because otherwise, other metal cations M 3+ would have also been bound by 3.
Furthermore, looking at Figure 4, the proton signals a to l remained unchanged after 1 equiv of Cr 3+ had been added to 3.
That is, the signals in spectra (iv) and (v) have the same positions. This again illustrates 1:1 complexation between 3 and Cr 3+ . This result is consistent with the conclusions obtained from Job's plot. In addition, HRMS analysis of 3·Cr 3+ (Supporting Information File 1, Figure S11) produced a distinctive signal at m/z = 587.1086, matching (3⋅Cr 3+ )/3, again indicating 1:1 complexation. In the IR spectra of 3 and 3⋅Cr 3+ (Supporting Information File 1, Figure S12), the C-C absorption band of a benzene moiety in 3 shifted from 1170 to 1185 cm −1 upon complexation, and the C=N absorption band at 1664 cm −1 shifted to a value of 1672 cm −1 .

Conclusion
In summary, a new anthracene-based tetraperimidine hexafluorophosphate 3 was prepared, and its structure was determined through X-ray analysis and 1 H and 13 C NMR spectroscopy. Compound 3 was proved to be a highly sensitive and selective chemosensor for Cr 3+ , and it can effectively distinguish Cr 3+ from other cations through fluorescence enhancement. Thus, complex 3 may have potential value for the application as a Cr 3+ detector.

Materials and instruments
The solvents and chemicals used for synthesis and analysis were analytical grade and obtained commercially. A RF-5301PC fluorescence spectrophotometer (Shimadzu) was used to record fluorescence spectra at room temperature. The excitation and emission slits were set to 10 nm. UV-vis absorption spectra were recorded at room temperature using a JASCO-V570 spectrometer. A Varian spectrometer was employed to record 1 H and 13 C NMR spectra. A Perkin-Elmer 2400C Elemental Analyzer was employed for elemental analyses. IR spectra were measured with a PerkinElmer Spectrum 100 FT-IR spectrophotometer. A Q-TOF LC/MS (Agilent) and a VG ZAB-HS (VG) mass spectrometer were applied for HRMS analysis.
Melting points were recorded employing a Boetius Block apparatus.

Analytical data
Synthesis of 1-ethylperimidine: Through a dropping funnel, a solution of perimidine (1.43 g, 8.5 mmol) in dry THF (30 mL) was added to a suspension of NaH (0.479 g, 20 mmol) in dry THF (10 mL), followed by stirring at room temperature for 1 h. Subsequently, bromoethane (1.308 g, 12 mmol) was added to the suspension, and the mixture was reacted for 24 h at ambient temperature. After filtration, the solvent was removed at reduced pressure, and water (50 mL) was added to the residue. This was extracted with CHCl 3 (3 × 20 mL), the organic layer was rinsed with water (3 × 30 mL), and dried over anhydrous

Fluorescence titrations
The concentration of 3 and the guest ions was 5.0⋅10 −6 and 0.0-45.0⋅10 −5 M, respectively, in the sample solutions. The excitation wavelength was set to λ ex = 258 nm and the widths of the emission and excitation spectral lines were adjusted to 5 and 10 nm. The emission spectra were recorded in the range of 375-500 nm. The program Origin 8.0 was employed for data processing.

UV-vis titrations
For UV-vis titrations, the sample solutions were prepared analogously to fluorescence titrations. The concentration of 3 was adjusted to 1.0⋅10 −6 M while the concentration of Cr 3+ ranged between 0.0 and 36.0⋅10 −6 M. The absorption spectra were recorded at 240-300 nm. Origin 8.0 was employed for data processing.

X-ray analysis
Diffraction data of 3 were collected by a Bruker Apex II CCD diffractometer [56]. SHELXS was used to solve the structure of 3 [57]. Other crystallographic data are shown in Supporting Information File 1, Table S1.

Supporting Information
Supporting Information File 1 Supporting crystallographic data, fluorescence, UV, HRMS, and IR spectra of 3 and 3·Cr 3+ , general considerations, characterization data, and copies of the 1 H and 13 C NMR spectra of all compounds.