Linkage Isomerism in Silver Acylpyrazolonato Complexes and

May 13, 2016 - Only for complexes 7–9 a clear shift of the carbonyl ν(C═O) band to ..... that can contribute to the conductivity but unrelated to...
5 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/IC

Linkage Isomerism in Silver Acylpyrazolonato Complexes and Correlation with Their Antibacterial Activity Fabio Marchetti,*,‡,§ Jessica Palmucci,‡ Claudio Pettinari,†,§ Riccardo Pettinari,†,§ Stefania Scuri,⊥ Iolanda Grappasonni,⊥ Mario Cocchioni,⊥ Mario Amati,*,∥ Francesco Lelj,∥ and Alessandra Crispini∇ †

School of Pharmacy and ‡School of Science and Technology, Chemistry Section, University of Camerino, Via S. Agostino 1, 62032 Camerino (MC), Italy ⊥ School of Pharmacy, Hygienistic, Environmental and Health Sciences Research Centre, University of Camerino, Via Madonna delle Carceri 9, 62032 Camerino (MC), Italy § ICCOM, CNR 62032 Camerino, Italy ∥ Dipartimento di Scienze, LASCAMM, CR-INSTM Unità della Basilicata and La.M.I., Via dell’Ateneo Lucano, 10, 85100 Potenza, Italy ∇ Centro di Eccellenza CEMIF.CAL-LASCAMM, CR-INSTM (Unità della Calabria, Dipartimento di Chimica e Tecnologie Chimiche), Università della Calabria, I-87030 Arcavacata di Rende (CS), Italy S Supporting Information *

ABSTRACT: Novel silver(I) acylpyrazolonato coordination polymers of formula [Ag(QR)]n (1−3) have been synthesized by interaction of silver nitrate with HQR in methanol in the presence of an equivalent quantity of KOH (in general HQR = 1-phenyl3-methyl-4-RC(O)-5-pyrazolone, in detail HQfb, R = -CF2CF2CF3; HQcy, R = -cycloC6H11; HQbe, R = -C(H)C(CH3)2). [Ag(QR)]n react with 2-ethylimidazole (2EtimH), 1-methylimidazole (Meim), and triphenylphosphine (PPh3), affording the mononuclear Ag(Qfb)(EtimH) (4), Ag(Qcy)(Meim)2 (5), Ag(Qbe)(Meim) (6), and Ag(QR)(PPh3)2 (7−9). All complexes have been analytically and spectroscopically characterized, and for some of them the X-ray crystal structure has been resolved. In particular, the single crystal molecular structure determination of Ag(Qfb)(EtimH) and Ag(Qbe)(PPh3)2 has confirmed the different coordination modes of the HQfb and HQbe acylpyrazolone ligands, the former being bound to the silver(I) ion in a monodentate fashion while the latter in the O2-chelating mode. Density functional theory computations suggest new insights about metal−ligand interactions and the observed linkage isomerism. While phosphine-containing complexes Ag(QR)(PPh3)2 (7−9) seem not to be able to efficiently inhibit the growth of Escherichia coli and Staphylococcus aureus, the polynuclear complexes [Ag(QR)]n (1−3) and the mononuclear Ag(Qfb)(EtimH) (4), Ag(Qcy)(Meim)2 (5), and Ag(Qbe)(Meim) (6) show a high and almost steady in time antibacterial activity, comparable to that of AgNO3. This activity is likely related to the degree of saturation of the silver center and to the presence of different ancillary ligands in the diverse typologies of complexes.

■

INTRODUCTION Silver(I) complexes bearing nitrogen and/or phosphorus donor ligands exhibit a wide range of applications in the fields of medicinal and analytical chemistry, catalysis, and the polymer industry.1,2 The biomedical uses of Ag(I) compounds are mainly related to their well-documented antibacterial and antifungal actions.3−5 The antimicrobial activity and other necessary properties (e.g., water solubility and light stability) of silver complexes can be adjusted by varying the number and type of ligands in the silver atom coordination environment. It has been established that many silver complexes typically exhibit superior antimicrobial activity in comparison with simple Ag(I) salts, but their frequent insolubility in an aqueous medium limits their uses in bactericidal compositions, ointments, and creams.6,7 In this context we have previously reported several examples of hydrosoluble and bioactive silver © XXXX American Chemical Society

complexes containing tris(pyrazolyl)methanesulfonate and polypyridine ligands together with 1,3,5-triaza-7-phosphadamantane (PTA) and its cationic complex N-methyl-1,3,5-triaza7-phosphaadamantane (mPTA).8−10 Because of our longstanding interest in acylpyrazolone ligands11−13 and their silver(I) coordination chemistry,14−17 we have likewise addressed our attention to some silver(I) acylpyrazolonates as potential additives to polyethylene (PE). We previously reported a number of mono-, di-, and polynuclear Ag(I) complexes containing acylpyrazolonates having a different 4acyl moiety and their embedding in a PE matrix to afford composite plastics able to exhibit potent antibacterial activity against three bacterial strains, Gram positive and Gram negative Received: February 28, 2016

A

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry as well.18 Here we extend our investigation to the interaction of different acylpyrazolones with silver(I) salts and their reactivity with N- and P-donor ligands, in order to synthesize silver complexes with diverse structural and physicochemical features and to check how the latter may influence their antibacterial activity. The acylpyrazolone HQ proligands have been chosen with quite different electronic and steric features in the acyl moiety, ranging from a linear fluorinated propyl group in HQfb, to a cyclohexyl moiety in HQcy and a 2-methyl-2-butenoyl in HQbe (Chart 1). While the coordination chemistry of HQcy has

grade acetonitrile and methanol. For the ESI-MS data, mass and intensities were compared to those calculated using IsoPro Isotopic Abundance Simulator, version 3.1.26 Peaks containing silver(I) ions were identified as the center of an isotopic cluster. Thermal gravimetric analyses (TGA) were carried out in a N2 stream with a PerkinElmer STA 6000 simultaneous thermal analyzer (heating rate: 7 °C/min). Synthesis and Characterization of HQ fb, HQcy, and HQbe Proligands. The proligands HQcy (3-methyl-1-phenyl-4-cyclohexanecarbonyl-pyrazol-5-one), HQbe (4-(3-methylbut-2-enoyl)-3-methyl-1phenyl-pyrazol-5-one), and HQfb (3-methyl-1-phenyl-4-heptafluorobutanoil-pyrazol-5-one) were synthesized by a procedure analogous to that previously reported.12,19 The analytical and spectral data of HQfb and the IR data of HQcy and HQbe are reported in the Supporting Information, section 1. Synthesis of [Ag(Qfb)]n (1). A methanol solution (30 mL) of the ligand HQfb (0.185 g, 0.5 mmol) and NaOMe (0.027 g, 0.5 mmol) was added to a water solution (10 mL) of silver nitrate (0.085 g, 0.5 mmol). A white precipitate slowly formed. After 1 h, the solvent was removed almost on a rotary evaporator and the precipitate afforded, which was filtered off, washed with Et2O (10 mL). After filtration, the gray powder was dried in vacuo to constant weight and shown to be derivate 1. Yield: 75%. It is insoluble in all solvents, except DMSO. Mp: 271−273 °C. Anal. Calcd for C14H8AgF7N2O2: C, 35.25; H, 1.69; N, 5.87%. Found: C, 35.56; H, 1.54; N, 5.61%. ΛM (DMSO, 310 K, 10−3 mol/L): 27.4 S cm2 mol−1. IR (cm−1) data: 1634s, 1596w, 1533w, 1501m, 1429m ν(CO, CC, CN, C−N), 1353w, 1335w, 1222s, 1167s, 1115vs, 1019s, 761s, 618s, 530m, 503s, 448m, 347m, 307m, 295m ν(Ag−N), 218m, 191m ν(Ag−O). 1H NMR (DMSO-d6): δ 2.24s (C3−CH3), 7.02t, 7.30t, 7.81d (5H, Harom of Qfb). 13C{1H} NMR (DMSO): δ 17.96s (C3-CH3), 100.2s (C4), 110.44s (CF2), 113.11s (CF3), 116.4s (COCF2), 119.7s, 123.4s, 128.4s, 139.8s (Carom of Qfb), 152.0s (C3), 162.6s (C5), 172.2s (CO). 19F{1H} NMR (CD3CN): δ −79.47s (CF2CF2CF3), −116.67s (CF2CF2CF3), −124.65s (CF2CF2CF3). TGA-DTA (mg % vs °C): heating from 30 to 600 °C with a speed of 8 °C/min; from 215 to 600 °C progressive decomposition, with a final black residual of 26.6% weight. Synthesis of Ag(Qfb)(EtimH) (4). 2-Ethylimidazole (0.020 g, 0.20 mmol) was added to a methanol solution (30 mL) of [Ag(Qfb)] (1) (0.048 g, 0.10 mmol). The yellow solution was stirred at room temperature for 24 h, the solvent removed on a rotary evaporator and the brown oil product dissolved in Et2O (5 mL), and then H2O was added (30 mL). A white precipitate afforded, which was filtered off, washed with H2O (20 mL), dried under reduced pressure (20 °C, 0.1 Torr) to constant weight, and shown to be compound 4. Yield: 65%. It is soluble in acetonitrile and DMSO. Mp: 235−237 °C. Anal. Calcd for C19H16AgF7N4O2: C, 39.81; H, 2.81; N, 9.77%. Found: C, 40.18; H, 2.63; N, 9.44%. ΛM (DMSO, 310 K, 10−3 mol/L): 19.4 S cm2 mol−1. IR (cm−1) data: 3676sbr, 3142sbr ν (N−HimH) 2989m ν(C−H), 1649m, 1619s, 1589m, 1568m, 1475s, 1421m ν(CO, CC, CN, C−N), 1338m, 1215s, 1179vs, 1059m, 804s, 591s, 507m, 441m, 334m, 286s ν(Ag−N), 251m ν(Ag−N), 226w. 1H NMR (CD3CN): δ 1.22t (C2-CH2−CH3 of EtImH), 2.33s (C3−CH3), 2.71q (C1−CH2-CH3 of EtimH), 6.95s (2H, H4 and H5 of EtimH), 7.13t, 7.35t, 7. 83d (5H, Harom of Qfb), 11.02 (N-H of EtimH). ESI-MS (+, CH3CN) m/z (%): 97 (100) [(EtimH2)]+, 245 (8) [Ag(EtimH)(CH3CN)]+, 300 (44) [Ag(EtimH)2]+, 573 (1) [Ag(HQfb)(EtimH)]+, 777 (4) [Ag2(Qfb)(EtimH)2]+. TGA-DTA (mg % vs °C): heating from 30 to 600 °C with a speed of 8 °C/min; from 160 to 600 °C progressive decomposition, with a final black residual of 34% weight. Synthesis of Ag(Qbe)(PPh3)2 (9). PPh3 (0.052g, 0.20 mmol) was added to a chloroform suspension (30 mL) of [Ag(Qbe)] (1) (0.036g, 0.10 mmol). The yellow solution was stirred at room temperature for 2 h, the solvent was removed on a rotary evaporator and the brown oil product dissolved in CH2Cl2 (5 mL), and then n-hexane was added (30 mL). A little brown precipitate afforded, which was filtered off, washed with n-hexane (20 mL), dried under reduced pressure (20 °C, 0.1 Torr) to constant weight, and shown to be compound 9. Yield: 78%. It is soluble in methanol, acetonitrile, chloroform and DMSO. Mp: 160−162 °C. Anal. Calcd for C51H45AgN2O2P2: C, 69.00; H,

Chart 1. HQ Proligands, Imidazoles, and Phosphine Used in This Work

been previously reported toward several metal acceptors,19−25 synthesis and coordination chemistry of HQfb and HQbe proligands have been reported here for the first time. Nine novel Ag(I) complexes have been synthesized and characterized, the X-ray crystal structures of some of them determined and the different metal−ligand connectivity investigated with density functional theory (DFT) techniques. All the Ag(I) complexes have been tested against two Gramnegative (Escherichia coli and Pseudomonas aeruginosa) and a Gram-positive (Staphylococcus aureus) bacteria, and the outcomes have been discussed in terms of the different physicochemical and structural features of the silver(I) systems.

■

EXPERIMENTAL SECTION

Materials and Methods. All chemicals were purchased from Aldrich (Milwaukee) and used as received. All of the reactions and manipulations were performed in the air. Solvent evaporations were always carried out under vacuum conditions using a rotary evaporator. The samples for microanalyses were dried in vacuo to constant weight (20 °C, ca. 0.1 Torr). Elemental analyses (C, H, N) were performed in-house with a Fisons Instruments 1108 CHNS-O Elemental analyzer. IR spectra were recorded from 4000 to 400 cm−1 with a PerkinElmer Spectrum 100 FT-IR instrument by total reflectance on a CdSe crystal. 1H, 13C{1H}, 19F{1H}, and 31P{1H} NMR spectra were recorded on a 400 Mercury Plus Varian instrument operating at room temperature (400 MHz for 1H, 100 MHz for 13C, 376.8 MHz for 19F, and 162.1 MHz for 31P). H and C chemical shifts (δ) are reported in parts per million (ppm) from SiMe4 (1H and 13C calibration by internal deuterium solvent lock), while F chemical shifts (δ) are reported in ppm versus CFCl3. Peak multiplicities are abbreviated: singlet, s; doublet, d; triplet, t; quartet, q; and multiplet, m. Melting points are uncorrected and were taken on an STMP3 Stuart scientific instrument and on a capillary apparatus. The electrical conductivity measurements (ΛM, reported as S cm2 mol−1) of acetonitrile, methanol, and DMSO solutions of the silver complexes were taken with a Crison CDTM 522 conductimeter at room temperature (r.t.). The positive and negative electrospray mass spectra were obtained with a Series 1100 MSI detector HP spectrometer, using an acetonitrile mobile phase. Solutions (3 mg/mL) for electrospray ionization mass spectrometry (ESI-MS) were prepared using reagentB

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Scheme 1. Synthetic Procedure for Complexes 1−9

5.11; N, 3.16%. Found: C, 69.07; H, 5.13; N, 2.97%. ΛM (DMSO, 310 K, 10−3 mol/L): 13.2 S cm2 mol−1. IR (cm−1) data: 3051wν (Carom− H), 1649m, 1603s, 1578s ν(CO), 1497m, 1479m, 1432m ν(CC, CN, C−N), 1393w, 1350m, 1095m 740s, 693s, 504vs ν(PPh3), 440m ν(PPh3), 427 ν(PPh3), 343m, 276w, 246w, 216m ν(Ag−O), 203m ν(Ag−O), 174w, 151m ν(Ag−P). 1H NMR (CDCl3): δ 1.60s (C−CH3), 1.89s (C−CH3), 2.34s (C3−CH3), 7.02m, 7.29m, 7.39m (36H, Harom of Qbe and of PPh3).1H NMR (DMSO): δ 1.78s (C− CH3), 2.05s (C−CH3), 2.21s (C3−CH3), 6.95t, 7.19t, 7.28t, 7.37t, 7.51t, 7.99d (36H, Harom of Qbe and of PPh3).13C{1H} NMR (CDCl3): δ 18.24s (C3-CH3), 28.12s, 29.28s [C-(CH3)2], 36.51s (N-CH3 of Meim), 47.50s [C-(CH3)2], 120.76s, 126.02s, 129.23s, 136.48s (Carom of Qbe), 122.57s, 123.23s, 139.28s (Carom of Meim), 147.2s (C3), 164.20s (C5). 31P{1H}NMR (CDCl3): δ 9.01s (T = 293 K), 10.19 2d [1J(31P - 109Ag) = 465.1 Hz, 1J(31P - 107Ag) = 392.9 Hz]. ESI-MS (+, MeOH) m/z (%): 633 (100) [Ag(PPh3)2]+, 995 (23) [Ag2(Qbe)(PPh3)2]+. TGA-DTA (mg% vs °C): heating from 30 to 600 °C with a speed of 8 °C/min; from 190 to 250 °C progressive decomposition with a weight loss of 77%, with a final white residue. The analytical, spectral and thermal data of complexes 2, 3, and 5−8 are reported in the Supporting Information, section 1. Crystal Structure Analyses. X-ray data for complexes 4 and 9 were collected on a Bruker-Nonius X8 Apex CCD area detector equipped with graphite monochromator and Mo Kα radiation (λ = 0.71073 Å), and data reduction was performed using the SAINT programs; absorption corrections based on multiscan were obtained by SADABS.27 Details of data collection and structure refinements for complexes 4 and 9 are reported in Table 1S of Supporting Information, section 2. Both structures were solved by Patterson method (SHELXS/L program in the SHELXTL-NT software package) and refined by full-matrix least-squares based on F2.28 All non-hydrogen atoms were refined anisotropically, and all hydrogen atoms were included as idealized atoms riding on the respective carbon atoms with C−H and N−H bond lengths appropriate to the carbon and nitrogen atom hybridization. Computational Methods. All the reported computations were performed by using the Gaussian09 rev. D.01 software.29 All the studies were performed by applying the DFT approach. The M06 xcfunctional 30 was used in all the computations; the ECP28MWB31 quasi-relativistic effective core potential (28 electrons in the core) and the associated 8s7p6d2f1g valence basis set contracted to 6s5p3d2f1g32 were applied to Ag in all the computations. On the other atoms (C, H, F, N, and O), the 6-311G(d) basis set was used (standard basis sets in Gaussian 09). In all the computations about the four model fragments of the polymer (Table 3 and Figure 4), a p

polarization function is added to the hydrogen atom involved into the hydrogen bond (6-311G(d,p) standard basis set in the used software). The Qfb perfluorinated chain was replaced by a single CF3 group for computational convenience in all the reported results. In all the computations, the integration grid was augmented in comparison to the default one (“int = ultrafine” key), as suggested for limiting the errors originated by small integration-grids in energy computations with meta-GGA xc functionals.33 SCF and structure optimization convergence criteria are the default ones. Solvation effects were included by the means of the IEFPCM method34−36 implemented in the standard software without any change with respect to the default parameters. The vibration frequencies discussed in the text and reported in Figure 5 are obtained by scaling the computed harmonic frequencies by a 0.950 factor. This factor produced a satisfying match between experimental and computational IR features in the 1400−1800 cm−1 spectrum of 9 (not reported in this paper). This value is identical to the one proposed for M06 and the 6-31+G(d,p) basis set37 and is in line to the ones suggested for the same xc functional with polarization-consistent basis sets (between 0.963 and 0.965)38 and, more recently, for the 631G(d) basis set (0.959).39 Antibacterial Activity of Silver(I) Complexes. Minimal Inhibitory Concentration (MIC). For the microdilution method,40 the serial dilutions of ligands and silver complexes were prepared in 96-well plates, starting from a concentration of 1 mg mL−1. All ligands and complexes were dissolved in DMSO. An equal concentration of bacterial inoculum (106 CFU mL−1), obtained by a direct colony suspension of an overnight culture in tryptone soya broth (TSB), was added to each well of the microtiter plate containing 0.1 mL of the serially diluted test samples. After incubation for 18−24 h at 37 °C for S. aureus and at 44 °C for E. coli, the optical density was measured at 600 nm (OD600) to determine the MIC values. The minimum inhibitory concentrations (MICs) were defined as the lowest concentration of the compound able to inhibit the growth of the microorganisms. All tests were done in triplicate.41 Growth Inhibition Assay. The OD600 of 96-well plates, prepared as described for microdilution method, was measured at predetermined time intervals to draw the growth curves of the bacteria.42 The growth curves of E. coli and S. aureus without antibacterial agents were also measured as blanks. Disc Diffusion Method. Antibiotic susceptibility testing was performed by the paper disk diffusion method.43 Growth media used for the antimicrobial assay were as indicated by the international guidelines of the CLSI.44,45 The target microorganism cultures (E. coli and S. aureus) was adjusted to match the tube of 0.5 McFarland C

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. (a) Perspective view of the asymmetric unit content of [Ag(Qfb)(EtimH)] (4) with atomic numbering scheme (ellipsoids at the 40% level); (b) crystal packing view of 4 showing the N−H---O hydrogen bonds and the formation of corrugated chains. turbidity standard using spectrophotometry at 600 nm, which equals 1.5 × 108 colony forming units/mL.46 Subsequently, the cultures were separately placed onto the surface of Mueller Hinton Agar (MHI) plates and spread over the surface by means of a sterile swab.47,48 Sterile filter papers disks (6 mm in diameter) were placed on the surface of inoculated plates and spotted with 20 μL of different concentrations of complexes dissolved in DMSO. The plates were incubated for 18 h at 37 °C for S. aureus and at 44 °C for E. coli. The diameters of the inhibition zones (including the 6 mm disk) were measured with calipers. A reading of more than 6 mm indicated the growth inhibition. No zone inhibition was observed using DMSO alone. Silver nitrate was used for comparison.

structurally characterized. While complexes 5−7 and 9 are soluble in methanol, acetonitrile, chloroform, and DMSO, 4 is soluble only in acetonitrile and DMSO and 8 only in chloroform and DMSO. Conductance values of 4−9 in DMSO indicate that they partially dissociate. The broad band at 2500−3200 cm−1 in the IR spectra of free neutral proligands, ascribed to the intramolecular O−H···O stretching mode, disappears upon coordination. Only for complexes 7−9 a clear shift of the carbonyl ν(CO) band to lower frequency has been observed, in accordance with coordination of silver through both oxygen atoms. By contrast, for the other complexes, a more complex pattern has been found in the region 1500−1650 cm−1, and no straightforward assignment of ν(CO) can be established in principle. In the far-IR region a number of bands appeared upon coordination, tentatively attributed to Ag−N, Ag−O, and Ag−P. In the 1H and 13C NMR spectra, carried out in DMSO-d6 or CDCl3, depending to the solubility of the complexes, the H and C ligand resonances are slightly displaced toward lower field upon coordination. The proton spectra of complexes 1−9 display one set of resonances with an integration of peaks in accordance with the formulation proposed. 31P NMR studies at 25 °C and at −55 °C have been carried out for 7−9 in in CDCl3. The 31P NMR spectra show one broad single resonance at room temperature, centered at ca. 10 ppm, while on cooling the samples this signal is splitted into a double doublet, from which it is possible to determine 1J(109Ag−31P) and 1J(107Ag−31P) coupling constants, which give information about the strength of the Ag−P bond and about the geometry at the silver as well. In complexes 7−9 the 1J(31P-107/109Ag) coupling constants are in the range 400− 500 Hz, typical of silver in (R3P)2Ag(O−O-chelating) compounds.14−17 Complex 7 also shows three 31P broad resonances at low temperature, suggesting that a partial dissociation of a phosphine occurs in chlorinated solution, as previously found for analogous compounds.14−17 The ESI-MS spectra of 4 and 7 have been recorded in acetonitrile, while those of the complexes 5, 6, 7, and 9 in

■

RESULTS AND DISCUSSION Syntheses and Characterization of Silver Complexes. The synthesis of silver complexes 1−9 and their molecular structure are reported in the scheme below. In detail, complexes 1−3 have been isolated as insoluble products from the reaction of AgNO3 with HQ in the presence of equimolar NaOMe. By suspending complexes 1−3 in methanol/chloroform or acetonitrile, in the presence of the appropriate imidazole of phosphine, the mononuclear complexes 4−9 were afforded (Scheme 1). They are all air-, moisture-, and light-stable, and 1−3 are quite high temperature melting solids, while 7−9 show lower melting points than the others. The polynuclear structure of 1− 3 has been hypothesized on the basis of their insolubility in most common solvents, apart DMSO where they partially dissociate, based on their conductance values, and the wellknown ability of acylpyrazolone ligands to act as ditopic spacers through the O2-chelating arm on one side and the nitrogen atom of the pyrazolone ring on the other side. Computational findings seem to further confirm our structural hypothesis (see below in Computational Results). Similarly, the composition and structures of the mononuclear complexes 4−9 have been established on the basis of their analytical and spectral data reported in the Experimental Section, and by analogy with previous Ag(I) acylpyrazolonates D

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

distances and angles around the metal ion are in good agreement with those reported in similar bis-triphenylphosphine Ag(I) complexes (Table 1).17

methanol. They show several clusters of signals, which correspond to species containing the ligands QR together with the silver ion, often in the presence of solvent molecules. Thermogravimetric studies showed that 1−9 are thermally stable compounds decomposing progressively within 300 °C, apart 5 and 6 which display decomposition steps due to loss of the ancillary methylimidazole ligands. X-ray Structural Analysis. The single crystal X-ray analysis of complex 4 has confirmed its neutral nature and therefore the general formula [Ag(Qfb)(EtimH)]. The HQfb ligand coordinates the silver(I) ion in a monodentate fashion, through the negatively charged nitrogen atom, where deprotonation arises from the original keto-form of the Qfb molecule, as already observed in similar HQR pyrazolones15 (Figure 1). The Ag(I) ion, being coordinated by the not charged nitrogen atom of one EtimH ligand, is found in a distorted linear geometry (N(1)−Ag−N(3) of 177.5(1)°). Bond distances and angles around the Ag(I) ion are in agreement with those found in similar complexes recently appearing in the literature.14−17 While the pyrazole ring of the Qfb coordinated ligand and the imidazole ring of EtimH are nearly coplanar, with a dihedral angle between the two mean planes of about 15°, the rotationally free phenyl ring of Qfb breaks the overall molecular planarity being almost perpendicular to the best mean plane passing through the pyrazole central ring. The fluorinated propyl group as acyl moiety on the Qfb ligand shows an all trans conformation with the carboxyl oxygen atom pointing on the opposite side with respect to the carbonyl oxygen atom of the pyrazole ring (C(3−C(2)−C(11)−O(1) 164.64)°). The presence of the N−H synthon on the EtimH coordinated ligand is able to give rise in the crystal packing to a strong linear hydrogen bond with the carbonyl oxygen atom of the pyrazole ring [N(4)---O(2)i 2.740(1) Å, N(4)−H(4)--O(2)i 169.7°, i = x + 1, −y + 1/2, z − 1/2] with the generation of corrugated chains almost in the bc plane. With the Qbe ligand, containing the 2-methyl-but-2-enoyl substituent, a completely different binding mode is observed. Indeed, the crystal structure solution of complex [Ag(Qbe)(PPh3)2], 9, proves the monoanionic O2-chelating mode of the Qbe ligand, with the Ag(I) coordination sphere completed by two triphenylphosphine ligands (Figure 2). The geometry around the central metal ion could be defined as very distorted trigonal-planar, being the O(1)−Ag−O(2) “bite” angle of 76.1(2)° and the angle P(1)−Ag−P(2) of 127.4(1)°. Bond

Table 1. Relevant Bond Lengths (Å) and Angles (deg) for Complexes 4 and 9a experimentalb Ag−N(1) Ag−N(3) C−O(2) C−O(1) N(1)−Ag−N(3) N(1)-σ[N(2);C(1);Ag]d Ag−O(1) Ag−O(2) Ag−P(1) Ag−P(2) O(1)−Ag−O(2) P(1)−Ag−P(2) O(1)−Ag−P(1) O(1)−Ag−P(2) O(2)−Ag−P(1) O(2)−Ag−P(2)

Complex 4 2.094(3) 2.082(3) 1.248(4) 1.223(4) 177.5(1) 0.06 Complex 9 2.354(4) 2.438(5) 2.481(2) 2.465(1) 76.1(2) 127.4(1) 112.5(1) 117.8(1) 106.5(1) 99.5(1)

computedc 2.097 2.118 1.211 1.216 175.3 0.25

(+0.1%) (+1.7%) (−3.0%) (−0.6%) (−1.3%)

2.329 (−1.1%) 2.417 (−0.9%) 2.488 (+0.3%) 2.461 (−0.2%) 77.65 (+2.0%) 125.48 (−1.5%) 102.26 (−10.0%) 129.25 (+8.9%) 117.32 (+9.2%) 92.85 (−7.2%)

a Both experimental and computational results are included. bFrom the crystallographic analysis in this paper. cFully optimized computational structures. For 4, it corresponds to the 4-κN computed structure in Figure 3. For 9, it corresponds to 9-κO2. Percentage errors (with sign) respect to the experimental structure are reported in parentheses. d N(1) distance from the plane defined by N(2), C(1), and Ag.

Computational Results. The structures of complexes 4 and 9 were optimized starting from the crystallographic ones. In our computations about 4, the perfluorinated chain was replaced by a single CF3 group for computational convenience. In Table 1, relevant computated structural parameters have been collected. It is possible to note that the general characteristics of the experimental structure are fairly well reproduced, apart from the C−O(2) bond distance and the N(1) distance from the plane defined by N(2), C(1), and Ag, which has been chosen to evaluate the pyramidalization of the N(1) donor atom. Such relatively large deviations have been discussed in more detail in the Supporting Information, section 3. Within it, evidence is reported that the C−O(2) bond distance is affected by the hydrogen bond between adjacent molecules which has been discussed above. Moreover, the computed higher pyramidalization is related to a small force constant for its modification, so that it appears easy to adjust this structural parameter to external perturbations like packing forces within the crystal. This hypothesis is corroborated by the finding of closer to the X-ray value of the CO internal parameters once the local interactions due to hydrogen bonding are mimicked, suggesting that the torsional value related to the pyramidalization is dependent on the overall environment influence and not an intrinsic molecular property. In the case of 9, a more satisfying match was observed (Table 1) between X-ray and computational bond distances obtained for a single molecule in a vacuum. This could be traced back to a predictable lower interaction of 9 with the environment, due to the nature of its ligands and the way they bind Ag.

Figure 2. Perspective view of the asymmetric unit content of [Ag(Qbe)(PPh3)2] (9) with atomic numbering scheme (ellipsoids at the 40% level). E

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry An interesting finding in the studied complexes is the linkage isomerism of the acylpyrazolonato ligand observed in the linear two-coordinated 4 and 6 compared to the four-coordinated 7− 9, as found in the performed crystallographic analysis on 4 and 9. Figure 3 shows some possible linkage isomers of 4 and 9 as

possibility is that 4-κN is the most stable isomer in the solid phase as well. This could result from the larger dipole moment and consequent potentially higher crystallization energy. Furthermore, a hydrogen bond was observed between adjacent molecules in the crystal, where the 4-κN CO group in the Qfb ring binds an NH group on an adjacent molecule (Figure 2). A computational analysis was performed for a better assessment of the importance of such hydrogen bonds. The reader is referred to the Supporting Information, section 4 for details about the performed analysis. Here we only summarize that 4-κN is more suitable to form a stable hydrogen bond and that this factor alone is able to counterbalance a good percentage of the 4-κN gas-phase lower stability (which amounts to 7.10 kcal/mol; see Table 2). The higher 4-κN dipole moment could definitely stabilize this isomer in the solid phase. In the case of 9, the situation is different. In analogy to 4, the three linkage isomers were computed (Figure 3), and their energies and dipole moments are reported in Table 2. 9-κO2 (Figure 3 and Table 2) is the structure described by the crystallographic study, and it is more stable in comparison to 9κN, both in a vacuum and in the considered modeled solutions. The gap in dipole moment between 9-κO2 and 9-κN is not as large as in the analogue isomers of 4. Furthermore, it seems unlikely that strong hydrogen-bonded adducts can be formed in the solid phase. Thus, the fact that 9-κO2 was isolated as crystal is in line with the reported computations. A second point of difference between 9 and 4 is the high stability of 9-κO, which is not observed in the 4-κO analogue (Table 2). In fact, in solution, the trigonal planar 9-κO shows a comparable stability to 9-κO2. It is likely that the higher 9-κO dipole moment and the acylic group interaction with the environment allow the deletion of the stability gap computed in a vacuum (2.28 kcal/ mol in favor of 9-κO2). In solid phase, however, only the 9-κO2 isomer has been found, possibly due to the relatively apolar environment surrounding the acylpyrazolonato ligand (surrounded by phenyl groups). From the above discussions, 9 shows a relatively marked preference for the acylpyrazolonato oxygen coordination, which is not observed in 4. Aiming to explain this finding, the bonding energy between ligands and Ag+ was decomposed according to the following sequence of ideal chemical processes for 9-κO2:

Figure 3. Computed linkage isomers of 4 and 9.

computed in this work. With 4-κN, we label the linkage isomer observed in the crystallographic analysis (N-coordinated Qfb), whereas, with 4-κO and 4-κO2, we labeled the two computed linkage isomers, characterized by an O-monocoordinating and an O2-chelanting Qfb, respectively (both not observed in the crystallographic studies). Table 2 collects some computed energetic parameters of the three isomers. It is interesting to note that 4-κO2 has a lower Table 2. Relative Electronic Energies (kcal/mol) and Computed in-Vacuum Dipole Moments (Debye) of 4 and 9 Linkage Isomersa 4-κN (crystal) 4-κO 4-κO2 9-κO2 (crystal) 9-κO 9-κN a

in a vacuum

dipole moment

CH2Cl2

methanol

0.00 +1.17 −7.10 0.00 +2.28 +10.04

19.96 16.15 10.86 8.32 12.11 14.13

0.00 +3.65 +1.82 0.00 −0.12 +4.03

0.00 +3.17 +2.52 0.00 +0.03 +3.20

The crystallographic isomer is indicated in parentheses.

molecular energy than 4-κN in our in-vacuum computations. However, inclusion of solvation effects by the means of IEFPCM (polarizable continuum) changes the stability order so that 4-κN becomes more stable, in agreement with the experimental studies. In this case, both methanol and the less polar dichloromethane have been chosen as solvents and both produce the same stability inversion. This is in line with the computed dipole moments that are 10.86 and 19.96 D (in a vacuum) for 4-κO2 and 4-κN, respectively, which clearly favors 4-κN in the presence of solvent polarization. The 4-κO isomer is the less stable in all the performed computations, possibly due to the loss in metal−ligand interaction passing from Qfb chelation to Qfb monodentate coordination. These results suggest that environmental effects can be essential in determining the relative stability of the linkage isomers in 4. In this light, the obtainment of 4-κN crystals could be explained as a consequence of the 4-κN presence in solution before the crystallization process, but a further

Q be(relaxed) → Q be(deformed)

(R1)

2PPh3(relaxed) → (PPh3)2(deformed)

(R2)

Q be(deformed) + (PPh3)2(deformed) → [(Q be‐κ O2 )(PPh3)2 ](deformed)

(R3)

Ag + + [(Q be‐κ O2 )(PPh3)2 ](deformed) → [Ag(Q be‐κ O2 )(PPh3)2 ](relaxed)

(R4)

Steps R1 and R2 represent the structural deformation of the ligands from their relaxed (and isolated) structure to the structure assumed in 9-κO2 (Qbe is computed as an anion). In R2, additionally, the two PPh3 moieties are placed exactly as in the complex. Thus, the whole energy change of steps R1 and R2 quantifies the energy spent by the ligands for changing their structure as required by the complex formation, including some (likely small) interaction energy between the two PPh3 ligands in step R2. Step R3 evaluates the necessary energy for passing F

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry from isolated and deformed Qbe and (PPh3)2(deformed) to their configuration in the complex. The R3 energy change can be considered an estimate of the interactions among ligands, and thus it can be associated with the steric repulsion among ligands in the complex. Step R4 measures the metal−ligand (bonding) interaction energy. The same ideal sequence of elemental steps can be obviously written for the 9-κN linkage isomer. Table 3

[Ag(Q be‐κ N)](deformed) → [Ag(Q be‐κ O2 )](deformed)

Hence, the Qbe oxygen coordination of Ag+ seems to be more effective and is the real origin of the difference in stability between 9-κO2 and 9-κN. The computed Mulliken charge of Ag+ in [Ag(Qbe-κO2)](deformed) (reaction R5) is 0.58e, the analogue one in [Ag(Qbe-κN)](deformed) is 0.38e. Not surprisingly, this fact seems to point to a higher electrostatic character of the O−Ag interaction with respect to the N−Ag one. In conclusion, the higher stability of 9-κO2 is assigned to the higher electrostatic stabilization associated with the interaction between Qbe and Ag+ through the Qbe oxygen atoms as donors. At this point, a direct link between the discussions above and the observed 4 linkage isomer can be seen. Also in 4, a higher stability was computed, in a vacuum, for 4-κO2 compared to 4κN. It seems reasonable to explain this finding by the same electrostatic stabilization found in 9. However, such in-vacuum preference for 4-κO2 with respect to 4-κN is not as large as the one of 9-κO2 with respect to 9-κN, so that environmental phenomena are able to invert this stability trend in 4 but not in 9 (see Table 2). We already mentioned the impossibility of 9 to form hydrogen bonds in the solid phase, differently from 4, and thus, the 9 solid phases are produced by 9-κO2 and 9-κO. A final computational study was performed on one of the polymeric compounds studied in this paper in order to get insight into the behavior of Qfb as a monocoordinating ligand or chelant. In this light, we designed an ideal small chain consisting of three coordinated Ag metals (Figure 4) where the central Ag is coordinated to two Qfb ligands as supposed in the real polymer (note that the perfluorinated chains are replaced by a single CF3 as in the previous study about 4). The peripheral metals (further coordinated to an ammonia and methoxide ligands) account for the rest of the real polymeric chain. As evident from Figure 4, in structures (a) and (b), a Qfb ligand operates as a chelant, in (c) and (d) as a monodentate ligand. This is the ligand of interest in the following discussion. The two chelated and the two nonchelated structures differ within each couple in the relative orientation of the two ligands around the central metal. Table 4 collects the computed relative energies. It is possible to note that the computed chelated structures (a) and (b) are energetically favored in a vacuum. In solution, however, the nonchelated structure (d) is significantly more stable. This fact can be traced back to the acyl CO interaction with the environment. It is not easy to predict the most stable form in the solid phase. It seems reasonable to exclude strong direct intermolecular interactions (like hydrogen bonds) in 1 solid phases. In this light, like in 9, in-vacuum stability order could be retained. However, it is hard to take into account the polarization effects with surrounding molecules in the solid phase. More insights can be achieved from the recorded IR spectra. Figure 5 shows the transmittance IR spectrum of the collected solid phases of 1 in the 1400−1800 cm−1 range. The computed IR spectra are reported for structures (a) and (d). Red squares are used for (a), and yellow triangles for (d). The computed features reported in Figure 5 are the normal modes mainly localized on the Qfb ligand of interest, “of interest” in the sense we stated above. In a real polymeric chain, it is reasonable that the chelant probes close (although different) chemical environments, so that they produce similar features. Furthermore, it is reasonable that a particular vibration mode on a ligand is relatively uncoupled to the analogue modes on the

Table 3. Electronic Energy Changes Associated to the Ideal Chemical Steps R1−R5 (kcal/mol) in the Formation of 9κO2 and 9-κN R1 R2 R3 total preparation energy (R1−R3) R4 total formation energy (R1−R4) R5

9-κO2

9-κN

+4.07 −2.38 −19.69 −18.00 −208.50 −226.50 −71.96

+3.57 −1.07 −21.55 −19.05 −197.40 −216.46 −78.38

collects the energies associated with the ideal elemental steps described above for both 9-κO2 and 9-κN. As expected, the total formation energy of 9-κO2 is 10.04 kcal/mol lower than the 9-κN one (as in Table 2). As evident, this is due almost entirely to the more negative R4 energy (−11.09 kcal/mol in favor of 9-κO2). On the other side, steps R1−R3 are much more similar in the two linkage isomers. These findings seem to rule out steric effects (or interactions in general) among ligands as a discriminating factor. On the contrary, a maximization of the whole metal−ligands interactions in 9-κO2 is likely the reason in favor of the observed isomery. An additional step was considered for further insight: [Ag(Q be‐κ O2 )](deformed) + (PPh3)2(deformed) → [Ag(Q be‐κ O2 )(PPh3)2 ](relaxed) be

(R5) +

In R5, the complex of Q -κO2 and Ag (in the structure of the complex) interacts with the couple of PPh3 ligands (deformed as in the complex) to form 9-κO2. A similar step has been analyzed for the κN linkage isomer. The associated reaction energy amounts to −71.96 kcal/mol (Table 4). In 9Table 4. Relative Electronic Energies, Standard Enthalpies, and Gibbs’ Energies (kcal/mol) of the Structures Shown in Figure 4

(a) (b) (c) (d)

Eel (vacuum)

H0 (vacuum)

G0 (vacuum)

Eel (CH2Cl2)

Eel (MeOH)

+0.01 0.00 +5.58 +3.11

0.00 +0.24 +5.45 +3.02

0.00 +0.41 +5.47 +1.79

1.26 1.18 3.28 0.00

1.72 1.78 3.31 0.00

(R6)

κN, an analogue evaluation led to an energy of −78.38 kcal/ mol. From above, there seems to be a larger energy gain (in terms of metal−ligand interaction) when the couple of PPh3 ligands bind Ag+ coordinated to the nitrogen donor atom of Qbe. This fact would favor 9-κN by 6.42 kcal/mol, but the computed reaction R6, shown below, is associated with an energy change of −15.15 kcal/mol, which is in favor of 9-κO2: G

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 4. Computed ideal structures used to model the polymers 1.

Figure 5. IR transmittance spectrum of 1 (T) recorded from the produced solid phases (top); computed Qfb vibration frequencies on structure (b) of Figure 4 (red squares) and structure (d) (yellow triangles). Computed intensities (I) are km/mol.

other ligands in the polymer. In this light, the use of computed features on a single ligand can be reasonable. From Figure 5, it is possible to note that a good correspondence can be found between experimental and

computed features of (b). All the experimental peaks can be easily associated with computed features. Furthermore, the trend in peak intensity is reasonably reproduced, so that we have used the same letters for labeling experimental and H

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 6. MIC values of HQR proligands and of complexes 1−9 against Escherichia coli, where AgNO3 is the positive control and blank is the negative control.

Figure 7. MIC values of HQR proligands and of complexes 1−9 against Staphylococcus aureus, where AgNO3 is the positive control and blank is the negative control.

Figure 8. Growth trends of E. coli in 62.5 ppm solutions of complexes 1−9.

in-vacuum energies and the IR study seem to suggest that chelation takes place in the polymer 1. Antibacterial Activity. The antibacterial activity of complexes 1−9 has been investigated against E. coli and S. aureus by several tests. The minimal inhibition concentration (MIC) values of complexes 1−9 against E. coli and S. aureus were determined by using an optical density method as previously described42 and are shown in Figures 6 and 7, respectively, where their values are compared with those of free neutral HQR proligands and with that of AgNO3 as positive

corresponding computed peaks. Peaks a and d, in particular, correspond to the symmetric and asymmetric CO stretching modes of the chelating Qfb ligand, respectively. The match between experimental and computed features of (d) (yellow triangles) is clearly less satisfying, in terms of both peaks positions and relative intensities. Computed peaks a and b correspond to the stretching modes of the acyl CO (not coordinated) and the coordinated (Qfb ring) CO. There is not a clear way to match the experimental ones. Although it appears tricky to consider conclusive these findings, both the computed I

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 9. Growth trends of S. aureus in 125 ppm solutions of complexes 1−9.

control and those of E. coli or S. aureus cultures as negative control. While HQR proligands only slightly inhibit the growth of both E. coli and S. aureus, the lowest MICs against E. coli and S. aureus were observed for complexes 1−6, values being in the range of 31.3−62.5 ppm against E. coli and 62.5−125 ppm against S. aureus. Worthy of mention are the MICs of complexes 3 and 5 against E. coli, being 15.6 and 31.3 ppm, respectively, thus revealing an antibacterial activity comparable to the activity of commonly used chemical disinfectants.47,48 Moreover, the MIC values (in the range from 15.6 to 62.5 ppm) fall in the range from 8 to 80 μg/mL explored in other studies about the antimicrobial activity of AgNO3.49 By contrast, complexes 7−9 showed no significant MICs either against E. coli and S. aureus (Figures 6b and 7b). An additional evidence is that silver(I) complexes generally display MIC values against S. aureus (Figure 7) higher than those against E. coli (Figure 7). As showed in the histograms in Figures 6 and 7, the MIC values of complexes 1−6 range from 62.5 to 125 ppm, whereas those of 7−9 fall in the range from 500 to 1000 ppm. Probably, this different behavior against Gram-positive and Gram-negative bacteria strains can be explained by the structure of their cell wall. In fact, the cell wall of Gram-positive bacteria is thicker than that of Gramnegative ones, and such higher protection against the attack from chemical agents is likely the reason for the lower activity of our Ag(I) complexes toward S. aureus (Gram-positive) than toward E. coli (Gram-negative).47,48 The time evolution of the antibacterial activity of Ag(I) complexes in MICs against E. coli and S. aureus was also evaluated, through the growth inhibition assay, and results are reported in Figures 8 and 9. The used MIC concentrations of silver complexes 1−9 were 62.5 ppm against E. coli and 125 ppm against S. aureus, with an exposition time of 96 h. From Figure 8 it can be appreciated that complexes 1−6 display a high inhibitory capability against E. coli, comparable to that of AgNO3, for all exposition time (96 h). Complexes 7−9 show instead a lower inhibitory capability, at the same concentration. Similarly, complexes 1−6 display higher inhibitory capability, with respect to 7−9, to reduce the bacterial charge within 96 h of exposition also against S. aureus (Figure 9). This experimental evidence is consistent with many other results from the literature. For example, Nomiya et al. reported a number of polynuclear silver(I) complexes containing Ndonor ligands, such as [Ag(im)]n,50 [Ag(1,2,3-triz)]n, [Ag(1,2,4-triz)]n,51 and [Ag(tetz)]n52 (imH = imidazole, 1,2,3-trizH = 1,2,3-triazole; 1,2,4-trizH = 1,2,4-triazole; tetzH = tetrazole),

which display effective antimicrobial activities, while their PPh3 containing complexes such as [Ag(im)(PPh3)3], [Ag(1,2,3triz)(PPh3)2]n, [Ag(1,2,4-triz)(PPh3)2]n,53 and [Ag(tetz)(PPh3)2]n51 show poor or no activity (Table 5). The authors attributed this drastic reduction of antimicrobial activity after incorporation of PPh3 to the restricted ligand exchange ability of the Ag−P containing complexes.50−53 Table 5. Antibacterial Activities of 1−9 Evaluated by MICb, Compared with Those of Previously Reported Silver Compoundsa compound

E. coli

S. Aureus

1 2 3 4 5 6 7 8 9 Ag(im)]nc [Ag(1,2,3-triz)]nd [Ag(1,2,4-triz)]nd [Ag(tetz)]ne [Ag(im)(PPh3)3]c [Ag(1,2,3-triz)(PPh3)2]ne [Ag(1,2,4-triz)(PPh3)2]ne [Ag(tetz)(PPh3)2]ne

62.5 62.5 31.3 31.3 31.3 62.5 >1000 >1000 >1000 6.3 250 4.0 4.0 >1000 >1000 >1000 >1000

125 250 125 125 250 250 >1000 >1000 >1000 50 >1000 125 15.7 >1000 >1000 500 >1000

a c

See refs 50−53. bMinimum inhibitory concentration (μg/mL). Reference 50. dReference 51. eReference 52.

In order to further investigate the antimicrobial efficacy of minimal inhibitory concentrations of complex 3, the suspension of each well was withdrawn and included in plate count agar (PCA) and then incubated overnight at 44 °C (E. coli) and 37 °C (S. aureus). In Supporting Information Figure S2 displays, as an example, the microbial growth of E. coli and S. aureus in the absence and presence of complex 3, which showed the best inhibitory performances together with 4. To check the persistence within time of the antibacterial activity of 1−9, a further bacterial inoculation (50 μL) has been carried out on wells containing complexes 1−9 previously exposed for 1 day to a first bacterial culture. After 24 h from the second inoculation, the optical density at 600 nm was estimated, and values are shown in Figure 10: complexes 1− J

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 10. Optical density of 1−9 revealed at 600 nm after additional inoculum of (a) Escherichia coli and (b) Staphylococcus aureus.

Figure 11. Inhibition zone diameters for complexes 1−9 against E. coli and S. aureus..

any inhibition against S. aureus. The superior antimicrobial activity of complexes 1−6 with respect to 7−9 can be tentatively explained by the less crowded coordination environments in 1−6 (with silver coordination numbers ranging from 2 to 3) with respect to the 4-coordinate tetrahedral Ag(I) centers in 7−9, which in turn afford a different thermodynamic stability and different ability to deliver Ag+ ions, which are known to carry out the antibacterial activity. To further confirm the different ability in delivering Ag+ ions, we have performed conductance measurements on compounds 1−9 in conditions identical to those used in the antibacterial tests (see experimental section): the ΛM range from 4.2 to 27.4

6 generally maintain a good antibacterial activity, while 7−9 seem to lose their inhibitory activity with time; in fact the optical density showed an increase against S. aureus, with values higher than the control (Figure 10b). The antibacterial activity of 1−9 was also analyzed by the inhibition zones obtained from agar diffusion tests with concentrations for 1−9 of 125 ppm, and the results are shown in Figure 11. Complexes 1−6 possess a similar behavior, with diameters inhibition zones ranging from 10 to 18 mm (±1) against E. coli and 7−9 mm (±2) against S. aureus, the most significant diameter being that of complex 1 against E. coli (18 mm ±2). Finally, complexes 7−9 seem not able to exhibit K

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Table 6. Computed Reaction Energies Associated to the Ideal Chemical Steps R7−R15 (kcal/mol) 4-κN R7 R8 total (R7 + R8) R9 R10 total (R9 + R10)

9-κO2

vacuum

DMSO

MeOH

+113.07 +64.26 +177.33 +44.35 +132.98 +177.33

+30.30 +22.04 +52.34 +30.97 +21.37 +52.34

+31.05 +22.42 +53.47 +31.25 +22.22 +53.47

R11 R12 R13 total (R11 + R12 + R13) R14 R15

ohm−1 cm2 mol−1, indicative that in DMSO at 37 °C all complexes are able to dissociate, even if to a different extent and regardless of the acylpyrazolone QR present. However, no clear relationship has been found with their antibacterial activity, some of the more active complexes being the less dissociated, as for 3. On the other side, the existence of cationic fragments in DMSO containing imidazoles or phosphines bonded to silver ion cannot be excluded: ESI-MS investigations on 7−9 suggest that the {Ag(PPh3)2}+ fragment is very stable and its bulkiness can reduce the penetration in the bacteria cell membrane and inhibit the ability to stimulate a cis−trans isomerization of unsaturated membrane fatty acids, which seems the first step at the basis of the antibacterial mechanism for silver salts,54 thus affording a lower antibacterial activity with respect to that of complexes 1−6. A recent work by Smoleński and co-workers54 has shown that some silver coordination polymers, assembled from 1,3,5triaza-7-phosphaadamantane (PTA) and flexible cyclohexanecarboxylate blocks, are soluble in water, where they release mainly {Ag(PTA)2}+ fragments, which are thought responsible for the good antibacterial activity of such substances. Hence, there are some structural analogies in their behavior in solution with respect to our complexes 7−9, which however are not good antibacterial agents. Our hypothesis on the role of the {Ag(PPh3)2}+ fragment in the reduction of antibacterial activity of 7−9 with respect to 1−6 seems apparently in contrast with results by Smoleński, but the differences between PPh3 and PTA, mainly the superior steric hindrance offered by PPh3,55 could be responsible for such different antibacterial activity. Further computational studies allow the evaluation of the energetics of different dissociation paths and of the presence in solution of some of its charged complexes that can contribute to the conductivity but unrelated to the “naked” Ag(I). In the case of 4, we have computed the reaction energy of the following steps: 4‐κ N → [Ag(EtimH)]+ + Qfb

(R7)

[Ag(EtimH)]+ → Ag + + EtimH

(R8)

4‐κ N → [Ag(Qfb‐κ N)] + EtimH

(R9)

[Ag(Qfb‐κ N)] → Ag + + Qfb

vacuum

DMSO

MeOH

+104.34 +49.78 +72.24 +226.36 +65.44 +160.91

+32.47 +27.14 +25.04 +84.65 +55.27 +29.38

+33.22 +27.45 +25.34 +86.00 +55.36 +30.64

In the case of 9, the following reaction steps were taken into account: 9‐κ O2 → [Ag(PPh3)2 ]+ + Q be

(R11)

[Ag(PPh3)2 ]+ → [Ag(PPh3)]+ + PPh3

(R12)

[Ag(PPh3)]+ → Ag + + PPh3

(R13)

9‐κ O2 → [Ag(Q be‐κ O2 )] + 2PPh3

(R14)

[Ag(Q be‐κ O2 )] → Ag + + Q be

(R15)

the sum of steps R11 + R12 + R13 corresponds to the total coordination energy. From Table 6, the whole coordination energy of 9 (R11 + R12 + R13) is significantly higher than the corresponding one of 4 (R7 + R8). This finding is in line with the higher antibacterial activity of 4. R7 and R11 describe the loss of the negatively charged acylpyrazolato ligand and a positive Ag+-ligand fragment as first dissociating step. This is the energetically easier route for producing charged species. In this respect, 4 seems to be more suitable to free ionic species, R7 being less demanding in terms of energy with respect to R11. This finding is evident in vacuum, but, in methanol, the two reaction energies are not so different, so that it is not possible to exclude that other effects could invert the predicted trend. In this light, in a polarizable medium, the release of charged fragment can be more similar in 4 and 9 in comparison to the release of Ag+. This Ag+ release in 9 is hindered by the strongly coordinating PPh3 ligands (steps R12 and R13). Hence, conductivity measures and antibacterial activities could not be easily correlated, a fact that was observed in our conductivity experiments. Moreover, as R7 and R11 are the energetically easier ways to produce the positively charged Agligand fragments, it is not surprising that the performed ESI measures always show the presence of such positive fragments. R9 and R14 describe the loss of the neutral ligands as a first dissociating step to produce a neutral Ag-acylpyrazolato complex. In 4, this process (R9) appears easier than the loss of the acylpyrazolato one (R7). On the contrary, in 9, the loss of the two PPh3 ligands (R14) is much more demanding. Again and not surprisingly, the two PPh3 ligands appears strongly coordinated to Ag+. In summary, the experimental and computational evidences allow us to conclude that, in complexes 7−9, the two phosphorus-donor PPh3 ligands strongly bind silver, likely reducing the release of Ag+ ions in the culture medium, which is the generally accepted pathway expected for silver species to exert their antimicrobial action. For the same reason, conductibility measures could not be indicative in this respect, the release of charged Ag-ligand complexes and acylpyrazolato being much easier and unrelated to the Ag+ release process.

(R10)

fb

where Q is an anion. Obviously, the couple of reactions R7−R8 and the couple R9−R10 correspond to the same complete complex dissociation. Table 6 collects the reaction energies (electronic energies) of such steps computed in a vacuum, DMSO, and methanol solutions after structure optimization of all the chemical species. In the case of different minima (different conformers), the most stable one was used for the energy evaluations. L

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

■

■

CONCLUSIONS Novel silver(I) complexes have been synthesized by employing three different acylpyrazolone proligands, and also imidazoles and triphenylphosphine as ancillary donors. The coordination of acylpyrazolonates to silver occurs through the O2-chelating face and the N2 of the pyrazole ring in the [Ag(QR)]n coordination polymers, while the N2 is preferred in the presence of imidazoles in complexes [Ag(QR)(imidazole)] and [Ag(QR)(imidazole)2]. By contrast, the O2-chelating face is found in complexes [Ag(QR)(PPh3)2]. According to our computational study, the origin of this linkage isomerism is to be searched in the stabilizing effect from the environment. Computational studies were performed on compounds 4 and 9, representative of the N-coordinated [Ag(QR)(imidazole)] and the O2-coordinated [Ag(QR)(PPh3)2] structures, respectively. In both 4 and 9, the O2chelated structures resulted the most stable linkage isomers in a vacuum. However, inclusion of solvation effects produces changes in the stability order of 4, making the N-coordinated isomer considerably more stable. The higher dipole moment of the N-coordinated isomer and the possibility to form stabilizing hydrogen bonds in solid phases can be indicated as the principal reasons for the higher stabilization from the environment (being the environment a liquid solution or the crystal packing). In 9, similar effects are computed, but they are not sufficiently extended to invert the in-vacuum stability order. As a consequence, the O2-coordinated complex is the most stable both in a vacuum and in condensed phases. The antibacterial activity of [Ag(Q R )] n , [Ag(Q R )(imidazole)], and [Ag(QR)(imidazole)2] is very high against both Gram-positive and Gram-negative cultures, differently from [Ag(QR)(PPh3)2] complexes, which display a very low activity. According to our analysis, it is likely due to the presence of triphenylphosphine ligands, which seem able to stabilize the {Ag(PPh3)2} fragment, even after dissociation in DMSO, thus reducing the Ag+ delivery.

■

REFERENCES

(1) Gimeno, M. C.; Laguna, A. Silver and Gold. In Comprehensive Coordination Chemistry II; Elsevier Ltd: Amsterdam, 2004; Vol. 6, pp 911−1145. (2) Khlobystov, A. N.; Blake, A. J.; Champness, N. R.; Lemenovskii, D. A.; Majouga, A. G.; Zyk, N. V.; Schröder, M. Coord. Chem. Rev. 2001, 222, 155−192. (3) Lansdown, A. B. G. Silver in health care: Antimicrobial effects and safety in use. In Current Problems in Dermatology; Hipler, U. C.; Elsner, P., Eds.; Karger Publishers: Basel, Switzerland, 2006; Vol. 33, pp 17− 34. (4) Farrell, N. Metal Complexes as Drugs and Chemotherapeutic Agents. In Comprehensive Coordination Chemistry II; Elsevier Ltd: Amsterdam, 2004; Vol. 9, pp 809−840. (5) (a) Fromm, K. M. Nat. Chem. 2011, 3, 178. (b) Fromm, K. M. Appl. Organomet. Chem. 2013, 27, 683. (6) Batarseh, K. I. J. Antimicrob. Chemother. 2004, 54, 546−548. (7) Cuin, A.; Massabni, A. C.; Leite, C. Q. F.; Sato, D. N.; Neves, A.; Szpoganicz, B.; Silva, M. S.; Bortoluzzi, A. J. J. Inorg. Biochem. 2007, 101, 291−296. (8) Pettinari, C.; Marchetti, F.; Lupidi, G.; Quassinti, L.; Bramucci, M.; Petrelli, D.; Vitali, L. A.; Guedes Da Silva, M. F. C.; Martins, L. M. D. R. S.; Smoleński, P.; Pombeiro, A. J. L. Inorg. Chem. 2011, 50, 11173−11183. (9) Smoleński, P.; Jaros, S. W.; Pettinari, C.; Lupidi, G.; Quassinti, L.; Bramucci, M.; Vitali, L. A.; Petrelli, D.; Kochel, A.; Kirillov, A. M. Dalton Trans. 2013, 42, 6572−6581. (10) Smoleński, P.; Pettinari, C.; Marchetti, F.; Guedes Da Silva, M. F. C.; Lupidi, G.; Badillo Patzmay, G. V.; Petrelli, D.; Vitali, L. A.; Pombeiro, A. J. L. Inorg. Chem. 2015, 54, 434−440. (11) Pettinari, C.; Marchetti, F.; Drozdov, A. Comprehensive Coordination Chemistry II 2004, 1, 97−115. (12) Marchetti, F.; Pettinari, C.; Pettinari, R. Coord. Chem. Rev. 2005, 249, 2909−2945. (13) Marchetti, F.; Pettinari, R.; Pettinari, C. Coord. Chem. Rev. 2015, 303, 1−31. (14) Cingolani, A.; Effendy; Marchetti, F.; Pettinari, C.; Pettinari, R.; Skelton, B. W.; White, A. H. Inorg. Chem. 2002, 41, 1151−1161. (15) Cingolani, A.; Effendy; Marchetti, F.; Pettinari, C.; Pettinari, R.; Skelton, B. W.; White, A. H. Inorg. Chim. Acta 2002, 329, 100−112. (16) Cingolani, A.; Effendy; Marchetti, F.; Pettinari, C.; Pettinari, R.; Skelton, B. W.; White, A. H. Inorg. Chem. 2004, 43, 4387−4399. (17) Lorenzotti, A.; Marchetti, F.; Pettinari, C.; Pettinari, R.; Skelton, B. W.; White, A. H. Inorg. Chim. Acta 2005, 358, 3190−3200. (18) Marchetti, F.; Palmucci, J.; Pettinari, C.; Pettinari, R.; Condello, F.; Ferraro, S.; Marangoni, M.; Crispini, A.; Scuri, S.; Grappasonni, I.; Cocchioni, M.; Nabissi, M.; Chierotti, M. R.; Gobetto, R. Chem. - Eur. J. 2015, 21, 836−850. (19) Marchetti, F.; Pettinari, C.; Pettinari, R.; Cingolani, A.; Camalli, M.; Spagna, R. Inorg. Chim. Acta 2000, 299, 65−79. (20) Pettinari, C.; Marchetti, F.; Cingolani, A.; Pettinari, R.; Troyanov, S. I.; Drozdov, A. J. Chem. Soc., Dalton Trans. 2000, 831−836. (21) Caruso, F.; Massa, L.; Gindulyte, A.; Pettinari, C.; Marchetti, F.; Pettinari, R.; Ricciutelli, M.; Costamagna, J.; Canales, J. C.; Tanski, J.; Rossi, M. Eur. J. Inorg. Chem. 2003, 2003, 3221−3232. (22) Pettinari, C.; Marchetti, F.; Pettinari, R.; Cingolani, A.; Rivarola, E.; Phillips, C.; Tanski, J.; Rossi, M.; Caruso, F. Eur. J. Inorg. Chem. 2004, 2004, 3484−3497. (23) Pettinari, C.; Marchetti, F.; Pettinari, R.; Drozdov, A.; Semenov, S.; Troyanov, S. I.; Zolin, V. Inorg. Chem. Commun. 2006, 9, 634−637. (24) Caruso, F.; Chan, E. J.; Hanna, J. V.; Marchetti, F.; Pettinari, C.; Di Nicola, C.; Pettinari, R.; Pizzabiocca, A.; Rees, G. J.; Quigley, D.; Rossi, M.; Skelton, B. W.; Sobolev, A. N.; White, A. H. Eur. J. Inorg. Chem. 2012, 2012, 1369−1379. (25) Hills, L.; Moyano, R.; Montilla, F.; Pastor, A.; Galindo, A.; Á lvarez, E.; Marchetti, F.; Pettinari, C. Eur. J. Inorg. Chem. 2013, 2013, 3352−3361.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b00495. Syntheses and experimental data of complexes 2, 3, and 5−8. Comparison between experimental and computed structures. Dimerization energies in complex 4. Computed structures and their energy in hartree. Photos of microbial growth on plate count agar (PCA) in the absence and presence of complex 3 as a representative example (PDF)

■

Article

AUTHOR INFORMATION

Corresponding Authors

*(F.M.) E-mail: [email protected]. *(M.A.) E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the University of Camerino (Fondo di Ateneo per la Ricerca 2011-2012) and Nuova Simonelli company. M

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

(54) Jaros, S. W.; Guedes da Silva, M. F. C.; Król, J.; Conceiçaõ Oliveira, M. C.; Smoleński, P.; Pombeiro, A. J. L.; Kirillov, A. M. Inorg. Chem. 2016, 55, 1486−1496. (55) Britvin, S. N.; Lotnyk, A. J. Am. Chem. Soc. 2015, 137, 5526− 5535.

(26) Senko, M. Iso.Pro 3.1, MS/MS software, Software tools for Mass Spectrometry; http://www.ionsource.com, 2009. (27) SMART, SAINT, and SADABS; Bruker AXS, Inc.: Madison, WI, 1997. (28) (a) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A. J. Appl. Crystallogr. 1994, 27, 435−436. (b) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 112−122. (29) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision D.01; Gaussian, Inc.: Wallingford, CT, 2013. (30) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241. (31) Andrae, D.; Häußermann, U.; Dolg, M.; Stoll, H.; Preuß, H. Theor. Chim. Acta 1990, 77, 123−141. (32) Martin, J. M. L.; Sundermann, A. J. Chem. Phys. 2001, 114, 3408−3420. (33) Wheeler, S. E.; Houk, K. N. J. Chem. Theory Comput. 2010, 6, 395−404. (34) Tomasi, J.; Mennucci, B.; Cancès, E. J. Mol. Struct.: THEOCHEM 1999, 464, 211−226. (35) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105, 2999−3093. (36) Scalmani, G.; Frisch, M. J. J. Chem. Phys. 2010, 132, 114110. (37) Alecu, I. M.; Zheng, J.; Zhao, Y.; Truhlar, D. G. J. Chem. Theory Comput. 2010, 6, 2872−2887. (38) Laury, M. L.; Carlson, M. J.; Wilson, A. K. J. Comput. Chem. 2012, 33, 2380−2387. (39) Kesharwani, M. K.; Brauer, B.; Martin, J. M. L. J. Phys. Chem. A 2015, 119, 1701−1714. (40) Wiegand, I.; Hilpert, K.; Hancock, R. E. W. Nat. Protoc. 2008, 3, 163−175. (41) Bauer, A. W.; Kirby, W. M.; Sherris, J. C.; Turck, M. Am. J. Clin. Pathol. 1966, 45, 493−496. (42) Banjara, R. A.; Jadhav, S. K.; Bhoite, S. A. J. Appl. Pharm. Sci. 2012, 2, 230−233. (43) Reller, L. B.; Weinstein, M.; Jorgensen, J. H.; Ferraro, M. J. Clin. Infect. Dis. 2009, 49, 1749−1755. (44) Petersen, P. J.; Jones, C. H.; Bradford, P. A. Diagn. Microbiol. Infect. Dis. 2007, 59, 347−349. (45) Pankey, G. A.; Sabath, L. D. Clin. Infect. Dis. 2004, 38, 864−870. (46) Bin Ahmad, M. B.; Lim, J. J.; Shameli, K.; Ibrahim, N. A.; Tay, M. Y.; Chieng, B. W. Chem. Cent. J. 2012, 6, 101. (47) Lu, X.; Ye, J.; Sun, Y.; Bogale, R. F.; Zhao, L.; Tian, P.; Ning, G. Dalton Trans. 2014, 43, 10104−10113. (48) Chamakura, K.; Perez-Ballestero, R.; Luo, Z.; Bashir, S.; Liu, J. Colloids Surf., B 2011, 84, 88−96. (49) Chopra, I. J. Antimicrob. Chemother. 2007, 59, 587−590. (50) Nomiya, K.; Tsuda, K.; Sudoh, T.; Oda, M. J. Inorg. Biochem. 1997, 68, 39−44. (51) Nomiya, K.; Tsuda, K.; Kasuga, N. C. J. Chem. Soc., Dalton Trans. 1998, 1653−1659. (52) Nomiya, K.; Noguchi, R.; Oda, M. Inorg. Chim. Acta 2000, 298, 24−32. (53) Nomiya, K.; Tsuda, K.; Tanabe, Y.; Nagano, H. J. Inorg. Biochem. 1998, 69, 9−14. N

DOI: 10.1021/acs.inorgchem.6b00495 Inorg. Chem. XXXX, XXX, XXX−XXX