Synthesis and structural studies of N-heterocyclic carbene Ag(I) and Hg(II) complexes and recognition of dihydrogen phosphate anion

Bis-benzimidazolium salt (S)-2,2′-bis[2″-(N-Et-benzimidazoliumyl)ethoxy]-1,1′-binaphthyl hexafluorophosphate [(S)-L1H2]·(PF6)2 and bis-imidazolium salts (S)-2,2′-bis[2″-(N-R-imidazoliumyl)ethoxy]-1,1′-binaphthyl hexafluorophosphate [(S)-L2H2]·(PF6)2 and [(S)-L3H2]·(PF6)2 (R = ethyl or benzyl), as well as their five N-heterocyclic carbene Hg(II) and Ag(I) complexes such as [(S)-L1Hg(HgBr4)] (1), [(S)-L2Hg(HgBr4)] (2), [(S)-L2Hg(HgI4)] (3), {[(S)-L2Ag](PF6)}n (4) and [(S)-L3Ag](PF6) (5) have been prepared and characterized. Each of complexes 1–3 consists of two rings (one 6-membered ring and one 11-membered ring), in which the oxygen atom in the ligand participates in coordination with Hg(II) ion. In complex 4, 1D helical polymeric chain is formed via biscarbene ligand (S)-L2 and Ag(I) ion. A 15-membered macrometallocycle is constructed through a ligand (S)-L3 and a Ag(I) ion in complex 5. Additionally, the selective recognition of H2PO4 − using complex 5 as a receptor was investigated on the basis of fluorescence and UV/vis spectroscopic titrations. The results indicate that complex 5 can distinguish effectively H2PO4 − from other anions.

Crystal structure analysis of 4 reveals the formation of 1D helical polymeric chain via NHC ligand (S)-L 2 and silver(I) ion ( Fig. 7(a)), in which Ag-π interactions are observed (π system being from naphthalene ring, and the separation of Ag-π being 3.591(1) Å) 67,68 . There exists a cavity of about 3.35 Å × 3.60 Å in the center of the chain by viewing from b axis ( Fig. 7(b)). The distance between adjacent two silver(I) ions in the chain is 8.177(4) Å. The coordination geometry of each silver(I) ion is approximately linear with 175.0(2)° angle of C(3)-Ag(1)-C(30) and 2.081(6) Å-2.084(6) Å bond distance of Ag-C (Fig. 7(c)). Similar observations were also reported for known NHC silver(I) complexes 69 .
In complex 5 (Fig. 8), one 15-membered macrometallocycle is formed by one ligand (S)-L 3 and one silver(I) ion, in which silver(I) ion is di-coordinated with two carbene carbon atoms to adopt an approximately linear geometry. The bond angle of C(8)-Ag(1)-C(37) is 173.4(1)°. The two Ag-C bond distances are 2.086(4) Å and  2.090(4) Å respectively. Ag···O separation of 3.2 Å is longer than the sum of the van der Waals Radii between Ag(I) ion and oxygen atom (3.1 Å) which indicates the absence of Ag···O interactions.
Powder X-ray diffraction. In order to establish their crystalline phase purity, powder X-ray diffraction (PXRD) experiments were carried out on complexes 1-5. As shown in the PXRD patterns ( Figure S1-S5), the excellent agreement between the experimental PXRD patterns of the bulk samples 1-5 and the patterns simulated from the single-crystal data proved the crystalline phase purity of the corresponding 1-5.
Thermogravimetric analysis of complexes 1-5. To examine the thermal stability of complexes 1-5, the thermogravimetric analyses for crystal samples of 1-5 were performed under a simulated air atmosphere with a heating rate of 20 °C min −1 from ambient temperature up to 600 °C. As demonstrated in Figures S6 and S7, the TG curves of 1 and 2 revealed that complex 1 started to decompose from ambient temperature to 172.4 °C and complex 2 started to decompose from ambient temperature to 243.5 °C. The curves represented the losses of approximately 0.5 equiv. of solvent molecule (ClCH 2 CH 2 Cl) for complex 1 (calcd: 3.53%, found: 3.50%) and 1.5 equiv. of solvent molecules (DMSO) for complex 2 (calcd: 8.57%, found: 8.56%). 49.29% weight loss from 172.4 °C to 434.3 °C for complex 1 and 55.35% weight loss from 243.5 °C to 405.6 °C for complex 2 were experienced, which resulted from the thermal decomposition of the organic components and did not stop until the heating ended at 600 °C. From the TG curve of 3 ( Figure S8), it has been found that this compound decomposed from ambient temperature to 211.4 °C, which represented the loss of approximately 1 equiv. of solvent molecule (DMSO) (calcd: 5.13%, found: 5.12%). Further decomposition from 211.4 °C to 281.5 °C represented the loss 2 equiv. of iodide atoms (calcd: 16.73%, found: 16.78%). Complex 3 experienced weight loss of 45.61% from 281.5 °C to 402.4 °C due to the thermal decomposition of the organic components. It did not stop until heating ended at 600 °C. The TG curve depicted in Figure S9 indicated that complex 4 had a high thermal stability which remained unchanged up to 241.6 °C. Almost one-step weight loss of 56.05% was detected from 241.6 °C to 395.6 °C, which was attributed to the thermal decomposition of the organic components and did not stop until heating ends at 600 °C. As shown in Figure S10, compound 5 started to decompose from ambient temperature to 127.3 °C, which represented the loss of approximately 0.5 equiv. of solvent molecule (DMSO) (calcd: 4.12%, found: 4.20%). With temperature increasing to 165.1 °C from 127.3 °C, weight loss of 11.40% represented the loss of 1 equiv. of silver ion (calcd: 11.39%). This compound experienced weight loss of 25.29% from 165.1 °C to 436.4 °C, which was attributed to the thermal decomposition of the organic components and did not stop until heating ended at 600 °C.          Figure S17), the detection limit was estimated to be 4.9 × 10 −8 mol/L for 5 76 .
In UV/vis titration experiment (Figure 11), the UV/vis absorption spectra of 5 dropped gradually with the increase of the molar fraction of H 2 PO 4 − . It was notable that a 1:1 complexation stoichiometry for 5·H 2 PO 4 − was established by Job's plot analysis at 280 nm (inset of Fig. 11) 77,78 , where the products (χΔA) between molar fractions and the discrepancy of the absorption bands were plotted against molar fractions (χ) of 5 under the conditions of a constant total concentration. When the molar fraction of 5 was 0.5, the χΔA value for 5·H 2 PO 4 − reached maximum 79 .
According to the size of the cavity (8.4 Å × 6.7 Å) and structural characteristics of 5, the size of H 2 PO 4 − (radius of H 2 PO 4 − being ca. 2.9 Å) is able to match well with that of 5. Possible binding sites in 5 contain oxygen atoms and silver(I) ion. The acting force between H 2 PO 4 − and 5 might be the result of combined effects of several weak intermolecular interactions, such as O-H···O hydrogen bonds and Ag···O interactions. But no significant changes of proton signals were observed in terms of the 1 H NMR spectra of 5 and 5·H 2 PO 4 − . To further explore the utility of 5 as a selective fluorescence receptor for H 2 PO 4 − , the competition experiments were conducted, where 5 (1 × 10 −5 mol/L) was firstly mixed with 10 equiv. of various anions (F − , Cl − , Br − , I − , HSO 4 − , OAc − and NO 3 − ), and then 10 equiv. of H 2 PO 4 − was added. As displayed in Figure S18, no obvious interference was observed in the presence of 10 equiv. of various anions. In high resolution mass spectrometry (HRMS) analysis of 5·H 2 PO 4 − ( Figure S19), m/z (859.2) was observed which provided additional evidence for the formation of a 1:1 complex between 5 and H 2 PO 4 − . This result was consistent with the findings of the Job's plot analysis (inset of Fig. 11).

Experimental
General procedures. All the reagents for synthesis and analyses were of analytical grade and used without further purification. Melting points were determined on an Digital Vision MP Instrument. 1 H and 13 C NMR spectra were recorded at 400 MHz and 100 MHz, respectively. Chemical shifts, δ, were reported in ppm relative to the internal standard TMS for both 1 H and 13 C NMR. J values were given in Hz. The elemental analyses of all compounds were obtained from the powder compounds recrystallised. The fluorescence spectra were performed using a Cary Eclipse fluorescence spectrophotometer. UV-vis spectra were recorded on a JASCO-V570 spectrometer. EI mass spectra were recorded on a VG ZAB-HS mass spectrometer (VG, U.K.). The powder X-ray diffractometry (PXRD) study was performed on a PANalytical X-Pert Pro diffractometer with Cu-Kα radiation. The thermogravimetric analysis (TGA) was performed with a NETZSCH STA 449 C instrument. IR spectra (KBr) were taken on an Bruker Equinox 55 spectrometer.

Fluorescence titrations.
A stock solution of the host was prepared in CH 3 CN as the concentration of 1.0 × 10 −4 mol/L. The stock solutions of the guests were prepared in CH 3 CN as the concentrations of 1.0 × 10 −3 mol/L and 1.0 × 10 −4 mol/L, respectively. The host solution (1.0 mL) was placed into a 10 mL volumetric flask, and the different amounts of the guest solutions (1.0 × 10 −3 mol/L or 1.0 × 10 −4 mol/L) were added using a microsyringe, and then diluted to 10 mL to prepare sample solutions. In the sample solutions, the concentrations of the host and the guest were 1.0 × 10 −5 mol/L and 0-40.0 × 10 −5 mol/L, respectively. After each addition, an equilibration time of 8-10 min was allowed before the fluorescence spectra were recorded. The fluorescence titration experiment was performed at 25 °C on a Cary Eclipse fluorescence spectrophotometer using a 1 cm path-length quartz cuvette. The sample solutions were excited at 280 nm, and the excitation and emission slits are 3 nm and 1.5 nm. The fluorescence emission spectra were recorded in the range of 300-500 nm. Statistical analysis of the data was carried out using Origin 8.0. CH 3 CN used in the titrations was freshly distilled.
Method for Job's plot. A stock solution of the host was prepared in CH 3 CN in the concentration of 1.0 × 10 −4 mol/L. The stock solutions of the guest were prepared in CH 3 CN in the concentrations of 1.0 × 10 −3 mol/L and 1.0 × 10 −4 mol/L, respectively. In the Job's plot experiment of 5 for H 2 PO 4 − , keeping the fixed overall concentration was 6.0 × 10 −5 mol/L, and the molar fraction of H 2 PO 4 − was changed from 0 to 1. In the course of preparation of sample solutions, the different amounts of host and guest solutions were placed into a 10 mL volumetric flask using a microsyringe, and then diluted to 10 mL. After each mixture, an equilibration time of 8-10 min was allowed before the absorption spectra were recorded. The absorption spectra were recorded in the range of 200-400 nm at 25 °C on a JASCO-V570 spectrometer using a 1 cm path-length quartz cuvette. Statistical analysis of the data was carried out using Origin 8.0. CH 3 CN used in the titrations was freshly distilled. X-ray data collection and structure determinations. X-ray single-crystal diffraction data for complexes were collected by using a Bruker Apex II CCD diffractometer at 296(2) K for [(S)-L 2 H 2 ]·(PF 6 ) 2 , 3 and 5, and 173(2) K for 1, 2 and 4 with Mo-Ka radiation (λ = 0.71073 Å) by ω scan mode. There was no evidence of crystal decay during data collection in all cases. Semiempirical absorption corrections were applied by using SADABS and the program SAINT was used for integration of the diffraction profiles 80 . All structures were solved by direct methods by using the SHELXS program of the SHELXTL package and refined with SHELXL 81 by the full-matrix least-squares methods with anisotropic thermal parameters for all non-hydrogen atoms on F 2 . Hydrogen atoms bonded to C atoms were placed geometrically and presumably solvent H atoms were first located in difference Fourier maps and then fixed in the calculated sites. Further details for crystallographic data and structural analysis are listed in Table 1 and Table 2. Figures were generated by using Crystal-Maker 82 .