Rationalizing the Effects of Amino Acid Side Chain, Pyridine, and

The SMART software was used for data acquisition and the SAINT software for data extraction. Structures were solved and refined using SHELX97.5 All ...
0 downloads 0 Views 306KB Size
Download PDF
Rationalizing the Effects of Amino Acid Side Chain, Pyridine, and Imidazole on the Assembly and Reversible Disassembly of a Octanuclear Cu(II) Complex Md. Akhtarul Alam,† Rik Rani Koner,† Amrita Das,† Munirathinum Nethaji,*,‡ and Manabendra Ray*,†

CRYSTAL GROWTH & DESIGN 2007 VOL. 7, NO. 9 1818-1824

Department of Chemistry, Indian Institute of Technology Guwahati, Guwahati -781039, India, and Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore - 560012, India ReceiVed March 14, 2007; ReVised Manuscript ReceiVed June 25, 2007

ABSTRACT: This paper reports the observation of a reversible disassembly process for a previously reported octanuclear Cu(II) complex with imidazole. To identify the factors responsible for the process, five Cu(II) complexes of different nuclearity with different amino acid-derived tetradentate ligands were structurally characterized. The results show that the coordination geometry preference of Cu(II), the tendency of imidazole to act as in-plane ligand, and H-bonding played important role in the formation and disassembly of the octanuclear complex. A general scheme describing the effect of different amino acid side arms, solvents, and exogenous ligands on the nuclearity of the Cu(II) complexes has been presented. The crystals of the complexes also showed formation of multifaceted networks in the resulting complexes. Introduction Supramolecular assembly in shapes resembling cups, capsules, or boxes have been receiving attention because of their potential applications in drug delivery, reaction container, and molecular recognition.1 Earlier, we communicated the formation of a capsule-shaped Cu(II) complex, [Cu8(S-his)8(pyridine)4] (1b), where H2S-his (1) is a tetradentate L-histidine-derived ligand, isolated from pyridine.2 In 1b H-bonds and ligand geometry had played a major role in stabilizing the capsule. Vittal and co-workers on the other hand reported isolation of mostly binuclear Cu(II) complexes from MeOH with similar amino acid-derived ligands.3 Thus, we felt the necessity to understand the role of the amino acid side chain, solvent, and other factors on the formation of 1b, as these might help in designing and controlling the functionality of metal ion-based multinuclear capsules. To probe we have characterized four Cu(II) complexes with methionine and tyrosine-derived ligand isolated from different solvents. In the process we have identified the factors responsible for the formation of species with different nuclearity as has been observed by us and others.2,3 To use the capsular complexes as an efficient carrier for guest molecules, the selective disassembly of the cage through a chemical event releasing the guest and reassembly of the host later would be important. While a great deal of effort is being put into assembling molecular cages or dendrimers, there are only a few literature sources on reversible carriers.4 Kutateladze and co-workers reported photoinduced cleavage of an organic cage,4a and recently Scheeren and co-workers reported “cascaderelease dendrimers” where the dendrimer gets completely disassembled by a single chemical trigger.4c Among coordination cages, a reversible disassembly process using a change of pH or solvent was observed in Cu(II) and a Ni (II)-based cages by Matsumoto et al. and Carbonaro et al., respectively.4b,c In this paper, we are able to reversibly disassemble the octanuclear complex (1b) to a mononuclear [Cu(S-his)‚(imidazole)] (1c) * To whom correspondence should be [email protected]. Fax: (91) 361 2690762. † Indian Institute of Technology Guwahati. ‡ Indian Institute of Science, Bangalore.

addressed.

E-mail:

using imidazole. The presence of pyridine or pyrazine did not affect the stability of the capsule. The present work differs from the pH- or solvent-dependent reversible assembly of coordination cages by Matsumoto et al. and Carbonaro et al., respectively, as the triggering event in the present case is molecular (addition of imidazole) rather than a bulk property of the medium such as pH or solvent.4b,c Experimental Section Materials and Methods. N,N-Dimethylformamide (DMF) and salicylaldehyde were purchased from Aldrich Chemical Co. L-Histidine monohydrochloride, L-methionine, DL-methionine, l-tyrosine, and imidazole were purchased from Sisco Research Laboratories Pvt. Ltd. (SRL), India, and used as received. The ligands H2S-his(1),2 H2S-tyr (4),3a and the complex [Cu8(S-his)8‚(pyridine)4]‚8H2O (1b)2 were synthesized as before. The IR spectra were recorded on either Nicolet Impact 410 or PerkinElmer Spectrum One FT-IR spectrophotometer with KBr discs in the range 4000-400 cm-1 and electronic spectra on either a Shimadzu U-2001 or Perkin-Elmer Lambda 25 UV-vis spectrophotometer. The thermogravimetric analysis (TGA) of the compounds were performed by using an Mettler Toledo SDTA 851e TGA thermal analyzer with a heating rate of 5 °C per min. under N2 atmosphere using a sample size of 5-10 mg per run. Solid-state magnetic susceptibility of the complexes at room temperature was recorded using Sherwood Scientific Magnetic balance MSB-1. Solution electrical conductivity measurements were made with a Systronics Conductivity Meter 306 by using 0.01 N KCl solution as calibrant. Elemental analyses were done using a Carlo Erba 1108 and also by using a Perkin-Elmer series II 2400 instrument. ESI-MS for ligands were recorded using a Micromass Quattro II mass spectrometer. Optical rotations were measured using a Perkin-Elmer 343 polarimeter. X-Band EPR spectra were recorded with a Jeol JES-FA series spectrometer fitted with a quartz dewar for measurements at liquid nitrogen temperature. The spectra were calibrated with DPPH (g ) 2.0037). Powder diffraction patterns were recorded using a Bruker Advance D8 with copper R source with 1.5406 Å wavelength. Caution! Perchlorate salts used as the starting materials are potentially dangerous as explosives and should only be handled in small quantities. S-2-(2-Hydroxybenzylamino)-4-methylsulfanylbutyric Acid. [H2Smet] (2). A mixture of l-methionine (1.00 g, 6.71 mmol) and LiOH‚H2O (0.284 g, 6.77 mmol) in methanol (dry, 30 mL) was stirred for 30 min, and salicylaldehyde (0.820 g, 6.72 mmol) in methanol was added

10.1021/cg070250s CCC: $37.00 © 2007 American Chemical Society Published on Web 08/21/2007

Reversible Disassembly of a Cu(II) Capsule Scheme 1. Amino Acid-Derived Ligands Used in This Work

slowly drop by drop. After the mixture was stirred for 20 min, a clear yellow color solution was obtained. The solution was treated with sodium borohydride (0.248 g, 6.71 mmol) with constant stirring upon which the solution became colorless. The solvent was evaporated using a rotary evaporator. The resulting sticky mass was dissolved in water. Clear solution was obtained, which was then acidified with dilute HCl and solution pH was maintained between 5 and 7. The ligand as a white solid was precipitated out. The solid was filtered and thoroughly washed with water. The solid was dried under reduced pressure inside a desiccator. Yield 1.45 g (80%). Anal. Calcd for C12H17NO3S: C, 56.44; H, 6.71; N 5.48. found: C, 55.92; H, 6.74; N, 5.37. 1H NMR Li2S-met (CD3OD, 300 MHZ. ppm): 2.06 (s, 3H, CH3S), 2.57 (m, 2H, CH2-S), 1.97 (m, 1H, CHCHH), 1.86 (m, 1H, CHCHH), 3.19 (dd, 1H, CH), 3.83 (d, 1H, CHH-phenolate), 3.44 (d, 1H, CHH-phenolate), 6.44 (t, 1H, p-phenolate), 6.97 (m, 2H, m-phenolate), 6.65 (d, 1H, ophenolate). IR (KBr, cm-1) υ(COO)asym 1615(s); υ(COO) sym 1430(m). [R]D25 ) -42° in MeOH, c ) 1.00 g/100 mL, in presence of 2 equiv of LiOH‚H2O. rac-2-(2-Hydroxybenzylamino)-4-methylsulfanylbutyric Acid. [H2rac-met] (3). This was synthesized following the same procedure as described for H2S-met (2) starting with racemic methionine (2.00 g, 13.40 mmol). Yield 2.85 g (83%). Anal. calcd (%) for C12H17NO3S (4): C, 56.44; H, 6.71; N 5.48. found: C, 55.92; H, 6.98; N, 5.30. 1H NMR Li2Rec-met (D2O, 300 MHZ. ppm): 2.08 (s, 3H, CH3-S), 2.27 (q, 2H, CHCHH), 2.658 (m, 1H, CH2S), 4.142 (t, 1H, CH), 4.358 (s, 2H, CHH-phenolate), 6.995 (m, 2H, m-phenolate), 7.340-7.409 (m, 2H, o,p-phenolate). IR (KBr, cm-1) υ(COO)asym 1615(sh), 1594; υ(COO) sym 1387(m). [R]D25 ) 0° in MeOH, c ) 1.00 g/100 mL, in presence of 2 equiv of LiOH‚H2O. [Cu(S-his)‚(imidazole)]‚2H2O (1c). Complex [Cu8(S-his)8‚(pyridine)4]‚ 8H2O (1b)2 (0.200 g, 0.066 mmol or 0.528 mmol based on copper) was taken up in 20 mL of dry methanol followed by addition of imidazole (0.100 g, 1.47 mmol) to the methanolic solution. The reaction mixture was refluxed with stirring for 1 h. Color of the solution changed from light green to deep green. Rotary evaporation was used to reduce the volume of the solution to ∼5 mL and then slowly allowed to evaporate at room temperature. After 2 days, green crystals formed. The crystals were filtered, washed with methanol, and dried in vacuum, in a desiccator. Yield: 0.120 g, (60%). IR (KBr, cm-1) υ(COO)asym 1601, υ(COO)sym 1396. Alternatively, complex 1c can be prepared directly from the ligand by the procedure as follows: Ligand H2S-his (0.202 g, 0.771 mmol) was deprotonated with KOH (0.087 g, 1.54 mmol) in 10 mL of dry MeOH. Cu(NO3)2‚3H2O (0.186 g, 0.771 mmol) was dissolved in 3 mL of dry MeOH and was added dropwise into the deprotonated ligand solution. The color of the reaction mixture became green with white particles of KNO3 precipitated out from the solution. To this reaction mixture was added excess imidazole 0.1050 g (1.54 mmol), and the color of the solution turned deep green. The reaction mixture was filtered and kept in an open atmosphere for crystallization. After 3-4 days, green crystals formed. Crystals were filtered and washed with cold MeOH. Yield 0.170 g (57%). The IR of 1c prepared this way was identical with that of prepared from 1b (above). Anal. calcd (%) for Cu(C13H13N3O3)‚(C3H4N2)‚2H2O (1c): C, 45.01; H, 4.96; N 16.40. found: C, 45.21; H, 5.20; N, 15.93. IR (KBr, cm-1) ν(COO)asym 1600, ν(COO)sym 1396. ΛM (MeOH): 2 S cm-2 mol-1. µeff(solid, 298 K); 1.63 µB/Cu. EPR: MeOH, 77 K, g| ) 2.253, g⊥)2.060, A| ) 175 G. [Cu2(S-met)2](2a). A methanolic solution of Cu(ClO4)2‚6H2O (0.294 g, 1.19 mmol) was added dropwise to a clear solution of H2S-met (0.203 g, 1.19 mmol) and KOH (0.090 g, 2.39 mmol) in 25 mL of dry methanol. The resulting dark green solution along with undissolved white particles was filtered immediately through a medium-porosity frit. Then the volume of the filtrate was reduced to ∼10 mL by rotary

Crystal Growth & Design, Vol. 7, No. 9, 2007 1819 evaporation and the complex was precipitated by the addition of diethyl ether. The resulting light green powder was filtered off and washed with diethyl ether prior to drying under vacuum in a desiccator (yield is high because KClO4 is present as a impurity). IR (KBr, cm-1) ν(COO)asym 1630; ν(COO)sym 1383, ν(ClO4) 1100. Purification and recrystallization: Diffusion of diethyl ether into the solution of the complex from dimethylformamide afforded dark green crystals with the chemical formula [Cu2(C12H17NO3S)2]. Yield: 51%. Anal. calcd for [Cu2(C12H17NO3S)2] (2a): C, 45.71; H, 4.47; N 4.44. found: C, 45.18; H, 4.91; N, 4.55. IR (KBr, cm-1) ν(COO)asym 1633(sh), 1600, ν(COO)sym 1383(s). ΛM (MeOH): 2.5 S cm-2 mol-1. µeff(solid, 298 K); 1.38 µB/Cu. UV-vis (λ, nm; , M-1 cm-1/Cu): (DMF) 362 (1560), 660 (180). EPR: DMF, 77 K, g| ) 2.271, g⊥)2.061, A| ) 174 G, AN )17 G. [Cu2(rac-met)2](3a). This compound has been prepared by following the same procedure like (2a) using 3 as ligand. Diffusion of diethyl ether into the solution of the complex in dimethylformamide afforded dark green microcrystals. Yield: 50%. Anal. calcd (%) for [Cu2(C12H17NO3S)2] (3a): C, 45.71; H, 4.47; N 4.44. found: C, 45.26; H, 4.80; N, 4.28. IR (KBr, cm-1) υ(COO)asym 1626(sh), 1599, υ(COO)sym 1385(s). ΛM (DMF): 2.0 S cm-2 mol-1. µeff (solid, 298 K); 1.15 µB/Cu. UVvis (λ, nm; , M-1 cm-1/Cu): (DMF) 358 (1300), 658 (180). EPR: DMF, 77 K, g| ) 2.271, g⊥)2.061, A| ) 174 G, AN )17 G. [Cu(rac-met)‚(pyridine)2](3b). This compound has been prepared by recrystallizing 3a from pyridine. Diffusion of diethyl ether into the solution of the complex in pyridine afforded dark green crystals. Yield: 50%. Anal. calcd for [Cu(C12H17NO3S)(C5H5N)2] (3b): C, 55.45; H, 5.71; N 8.82. found: C, 55.30; H, 4.95; N, 8.61. IR (KBr, cm-1) υ(COO)asym 1645(sh), 1592, υ(COO)sym 1380(s). ΛM (DMF): 3.0 S cm-2 mol-1. UV-vis (λ, nm; , M-1 cm-1/Cu): (pyridine) 423 (770), 678 (260). µeff(solid, 298 K); 1.73 µB/Cu. EPR: pyridine, 77 K, g| ) 2.253, g⊥)2.071, A| ) 174 G. [Cu(S-tyr)‚(pyridine)2](4b). A methanolic solution of Cu(ClO4)2‚ 6H2O (0.168 g, 0.697 mmol) was added dropwise to a clear solution of H2S-tyr (0.200 g, 0.697 mmol) and Et3N (0.141g, 1.39 mmol) in 25 mL of dry methanol. The resulting dark green solution along with undissolved white particles was filtered immediately through a mediumporosity frit. Then the volume of the filtrate was reduced to ∼10 mL by rotary evaporation, and the complex was precipitated by the addition of diethyl ether. The resulting light green powder was filtered off and washed with diethyl ether prior to drying under vacuum in a desiccator (KClO4 present as a impurity). IR (KBr, cm-1) ν(COO)asym 1630; ν(COO)sym 1394, ν(ClO4) 1100. Diffusion of diethyl ether into the solution of the complex from pyridine afforded dark green crystals of pure 4b. Yield: 50%. Anal. calcd (%) for [Cu(C16H15NO4)‚(C5H5N)2] (4b): C, 61.65; H, 4.98; N 8.30. found: C, 61.30; H, 4.70; N, 8.30. IR (KBr, cm-1) ν(COO)asym 1623(s), 1606(sh), ν(COO)sym 1394(s). ΛM (MeOH): 3.0 S cm-2 mol-1. UV-vis (λ, nm; , M-1 cm-1/Cu): (pyridine) 417 (592), 667 (177). µeff(solid, 298 K); 1.84 µB/Cu. EPR: MeOH, 77 K, g| ) 2.230, g⊥)2.011, A| ) 174 G. [Cu(S-met)‚(imidazole)](2c). This compound has been prepared by following the same procedure like (1c) using 2 as ligand. Slow evaporation of methanol afforded green crystal. Yield: 55%. Anal. calcd (%) for [Cu (C12H17NO3S). (C3H4N2)] (2c): C, 46.87; H, 4.99; N 10.94. found: C, 46.70; H, 3.99; N, 10.80. IR (KBr, cm-1) ν(COO)asym 1618, 1599(sh), ν(COO)sym 1384(s). ∆M (MeOH): 2 S cm-2 mol-1. µeff(solid, 298 K); 2.05 µB/Cu. UV-vis (λ, nm; , M-1 cm-1/Cu): (MeOH) 389 (660), 632 (142). EPR: MeOH, 77 K, g| ) 2.192, g⊥)2.00, A| ) 182 G. X-ray Crystallography. A single crystal of 1c prepared from 1b was grown by slow evaporation of the methanolic solution of the complex. The single crystals of 2a and 3b were grown from DMF and pyridine by slow diffusion of diethyl ether, respectively. The crystals of 1c, 2a, 3b, 4b, and 2c were mounted on glass fiber. All geometric and intensity data for the crystals were collected at room temperature using a Bruker SMART APEX CCD diffractometer equipped with a fine focus 1.75 kW sealed tube Mo KR (λ ) 0.71073 Å) X-ray source, with increasing ω (width of 0.3° per frame) at a scan speed of 3 s/frame. The SMART software was used for data acquisition and the SAINT software for data extraction. Structures were solved and refined using SHELX97.5 All non-hydrogen atoms were refined anisotropically. The hydrogen atoms were located from the difference Fourier maps and

1820 Crystal Growth & Design, Vol. 7, No. 9, 2007

Alam et al.

Table 1. Crystal Data and Structure Refinement for 1c, 2a, 3b, 4b, and 2c 1c

2a

3b

4b

2c

empirical formula fw crystal system space group a, Å b, Å c, Å R, β, γ, deg

C128H136Cu8 N40 O36 3319.17 monoclinic P21 19.625(12) 9.617(6) 41.56(03) β ) 92.260(11)

C48H55Cu4N4O1 2S4 1262.44 monoclinic P21 13.337(8) 8.622 (5) 23.425(16) β ) 106.032(9)

C26H25CuN3 O4 507.04 orthorhombic P212121 9.2596(4) 9.7379(4) 26.6724(14)

C15H19CuN3O3S 384.93 monoclinic P21 6.9706(7) 21.827(2) 11.0363(12) β ) 100.638(7)

V, Å3 Z/r/µ collected/indep reflns FLACK parameter GOF final R indices [I > 2σ(I)] R indices (all data)

7837(9) 2/1.406/1.147 36028/26886 0.06(2) 0.955 R1 ) 0.1044 wR2 ) 0.2324 R1 ) 0.2080 wR2 ) 0.2932

2589(3) 2/1.619/1.847 20081/10275 0.041(16) 1.174 R1 ) 0.0550 wR2 ) 0.1625 R1 ) 0.0586 wR2 ) 0.1664

C22H25CuN3O 3S 475.07 triclinic P1h 8.988(3) 10.602(3) 13.550(4) 69.213(4), 84.875(5), 67.454(4) 1113.3(6) 2/1.417/1.103 12969/9874 achiral 1.466 R1 ) 0.0733 wR2 ) 0.1498 R1 ) 0.1137 wR2 ) 0.1603

2405.03(19) 4/1.400/0.945 23040/3979 0.000(10) 1.184 R1 ) 0.0314 wR2 ) 0.0497 R1 ) 0.0456 wR2 ) 0.0513

1650.3(3) 4/1.549/1.467 15588/6972 0.017(14) 1.248 R1 ) 0.0509 wR2 ) 0.1064 R1 ) 0.0806 wR2 ) 0.1149

were refined isotropically. Thus, some of the C-H bonds will not be ideal and may vary. The hydrogen atoms attached to the solvent water molecules in 1c could not be located or fixed, so the molecular weight may not match. Selected crystallographic data are summarized in Table 1. Perspective views of the complexes were obtained by ORTEP.6

Scheme 2. Reversible Disassembly Process and Identity of the Species in Different Solvents

Results and Discussion Synthesis and Characterization. The complexes were characterized using IR, elemental analysis and conductance measurement. The complexes 2a, 3a, 3b, and 4b show carboxylate stretches at ∼1630 and at ∼1380 cm-1 for νasym and νsym, respectively (Experimental Section).7 The elemental analyses support the formulation of the complexes as [Cu(Shis)(imidazole)]‚2H2O (1c), [Cu2(S-met)2] (2a), [Cu2(rac-met)2] (3a), [Cu(rac-met)‚(pyridine)2](3b), [Cu(S-tyr)(pyridine)2](4b), and [Cu(S-met)(imidazole)](2c). The very low molar conductance supports the nonelectrolytic nature of the complexes.8 The solid-state room-temperature magnetic moments of mononuclear 1c, 3b, 4b, and 2c are 1.65, 1.73, 1.84, and 2.05 µB, respectively, closer to the spin-only value of 1.73 for Cu(II) expected for mononuclear complexes.9 The room-temperature magnetic moments of binuclear 2a and 3a, 1.38 and 1.15 µB/copper, respectively, are considerably lower than the spin-only value because of the antiferromagnetic coupling expected for bridged binuclear complexes. This has been observed by others for similar phenoxo-bridged binuclear Cu(II) complexes.3a Thus, bulk property of the complexes are consistent with the structural observations described below. Reversible Disassembly. Addition of 2 equiv of imidazole per Cu(II) to the octanuclear 1b2 resulted the isolation of 1c in MeOH. The isolation of mononuclear 1c from 1b using imidazole demonstrates that it is possible to disassemble the multinuclear cage structure using a simple molecular trigger. The fact that the capsule itself was prepared from pyridine, a strong enough ligand, showed that the chemical trigger has to be stronger donor than pyridine. No disassembled product was isolated upon addition of pyrazine, pyrazole, or ammonia to the 1b in methanol. This is significantly different from the work by Matsumoto et al. and Carbonaro et al., as the present disassembly process is triggered by a molecule rather than change in bulk properties such as pH or solvent.4b This is perhaps the result of competition between imidazole of the ligand and external imidazole in excess (∼2 equiv) for a binding site on Cu(II) which resulted in a disassembled capsule. Thus, in this case the chemical event is responsible for breaking the Cu-imidazole coordination bond in 1b using a equally strong

donor in excess which replaces the histidine imidazole from the coordination sphere of Cu(II). Imidazole (pKa ∼7) being a strong base, the tetrameric cage structure is protected from weaker ligands in solution. Dissolving the crystals of 1c in pyridine yields the crystals of 1b, making the process reversible. The presence of a large excess of pyridine shifts the equilibrium toward 1b because of the labile nature of Cu(II) complexes. The overall process is shown schematically (Scheme 2). Structure of 1c. The mononuclear nature of 1c was confirmed by structural characterization. Because of weak diffraction and large number of non hydrogen atoms, 212 in total, refinement was difficult and final R indices were high (0.1044, Table 1). The amount of water in 1c was confirmed by thermogravimetric analysis (TGA). The TGA between 50-140 °C shows weight loss of 8.5% against calculated weight loss of 8.4% for 2 mol of water for 1c. Powder diffraction data (Supporting Information) also support the structure. The ORTEP diagrams of two units are shown in Figure 1. Data for all the units are given in Supporting Information. The coordination geometry at the Cu(II) is best described as square-pyramidal with varying amount of trigonal bipyramidal

Reversible Disassembly of a Cu(II) Capsule

Figure 1. Molecular structure of two independent units of 1c with (a) lowest TBP distortion and (b) highest TBP distortion. Thermal ellipsoids set to 50% probability level. Selected distances (Å) in part a: Cu2O3b 1.978(11), Cu2-N4b 1.891(13), Cu2-N2b 2.008(12), Cu2-N3b 2.014(11), Cu2-O1b 2.219(10), τ ) 0.11; in part b: Cu4-O3d 1.938(11), Cu4-N4d 1.986(12), Cu4-N2d 2.014(11), Cu4-N3d 1.998(12), Cu4-O1d 2.344(13), τ ) 0.37.

(TBP) distortion across the mononuclear units. The amount of trigonal distortion τ was calculated from structural data where the value of τ should be 0 for perfect square-pyramidal geometry and 1 for perfect TBP structure.10 In 1c, the τ values for eight units vary from 0.11 to 0.37, showing significant trigonal distortion in some of the units (Figure 1b). The in-plane Cu-N and Cu-O bond lengths are similar to that found for 1b and are typical for Cu(II) complexes.11 The conformation at the chiral carbon for all eight units is S, as the ligand was synthesized from S-isomer of the histidine. One significant change in the mode of coordination that occurred in 1c is the axial coordination of carboxylate. Octanuclear 1b and all other complexes reported here have the carboxylate always coordinated as an in-plane ligand. Complexation in MeOH with Different Amino Acids. Discussion in the previous sections highlight that imidazole coordination plays important role in the formation and disassembly of the capsule and extra imidazole in 1c influences carboxylate group of the ligand to bind axially. To gain insight, we chose to synthesize Cu(II) complexes of the methionine and tyrosine derivative of the ligand, H2S-met(2) and H2S-tyr(4), respectively. These choices were based on having amino acid with donor weaker than histidine in the side arm. Addition of S-Met(2) or rac-Met(4) with Cu(II) in MeOH resulted in green crystalline complexes 2a and 3a, respectively, with identical formulation of [Cu2(2 or 3)2]. A single crystal of 2a was grown from DMF. The structural identity of 2a was found to be a bis-phenoxo-bridged binuclear complex using single-crystal X-ray diffraction. The complex 2a is a phenoxo-bridged binuclear Cu(II) complex with thioether arm of the ligand remains uncoordinated (Figure 2). The axial coordination is provided by the carboxylate oxygen from another binuclear unit, forming a planar coordination polymeric network. The geometry around the two copper centers in each binuclear unit is different. While one is closer to the ideal square pyramidal geometry with Cu(II) 0.21 Å above

Crystal Growth & Design, Vol. 7, No. 9, 2007 1821

Figure 2. Formation of dinuclear species in methanol (top) and molecular structure of one independent unit of 2a with thermal ellipsoids set to 50% probability level (bottom). Selected distances (Å) and angles (deg): Cu1‚‚‚Cu2 3.0009(17), Cu1-O1 1.938(4), Cu1-O3a 1.960(4), Cu1-O3 1.952(4), Cu1-N1 1.976(5), Cu1-O2 2.378(5), O3Cu1-O3a 79.74(16), O3a-Cu1-O1 100.06(18). τ at Cu1 0.04 and at Cu2 0.38.

the mean plane formed by N1, O1, O3a, O3, the geometry at the other copper center is distorted toward TBP with a τ value of 0.38, higher than that observed in any of the units of 1c. Vittal and co-workers have structurally characterized a number of Cu(II) complexes of related ligands with alanine, valine, glycine, tryptophan, and tyrosine.3 All the reported complexes isolated either from water or alcohols have a phenoxo-bridged binuclear structure like that of 2a. The phenolic arm of the tyrosine-derived ligand remains uncoordinated to Cu(II) similar to that of the methionine arm in 2a.3a Thus, all the Cu(II) complexes of reduced Schiff bases of amino acidsalicylaldehyde ligands isolated from alcohol or water irrespective of amino acid used are phenoxo-bridged binuclear complexes. Although crystals of 2a were isolated from DMF, it is unlikely to have an effect on binuclear structure, as DMF is not a strong ligand. We conclude that in the absence of a strong ligand in the form of solvent and in the amino acid side arm, the formation of a phenoxo-bridged binuclear complex is favored irrespective of amino acid used. The ligands 1-4 being nonplanar ligands, phenoxo-bridged binuclear complex formation satisfies the coordination requirements on the Cu(II). This indicates that the solvent such as pyridine which can act as a ligand to Cu(II) might be able to break the phenoxo bridge, forming a mononuclear complex. Pyridine Complexes. The pyridine-coordinated mononuclear 3b and 4b were isolated by crystallizing corresponding dinuclear species from pyridine (Figure 3). The structures of 3b and 4b show that the pyridine, being the preferred ligand for Cu(II) compared to phenolate, is capable of breaking the phenoxo bridge of the binuclear complex to form mononuclear complexes

1822 Crystal Growth & Design, Vol. 7, No. 9, 2007

Alam et al.

Figure 4. Molecular structure of one independent unit of 2c with thermal ellipsoids set to 50% probability level. Selected distances (Å) and angles (deg): Cu1-N1 1.984(4), Cu1-O1 1.925(3), Cu1-N2 1.941(5), Cu1-O2 1.958 (4), N1-Cu1-O1 93.95(15), O1-Cu1-N2 91.17(16), N2-Cu1-O2 91.29(16), O2-Cu1-N1 84.02(16), N1Cu1-N2 170.23(16), O1-Cu1-O2 176.31(16), τ ) 0.017.

Figure 3. Molecular structure of 3b (top) and 4b (bottom) with thermal ellipsoids set to 50% probability level. Selected distances (Å)in 3b: Cu1-O3a 1.910(3), Cu1-N2a 2.018(3), Cu1-O1a 1.965(3), Cu1N1a 1.997(3), Cu1-N3a 2.429(4), τ ) 0.00. In 4b: Cu1-N1 2.012(2), Cu1-O1 1.9101(18), Cu1-N3 2.024(2), Cu1-O2 1.9516(18), Cu1-N2 2.280(2), τ ) 0.32.

where pyridines occupy both axial and equatorial positions (Figure 3). Thus, pyridine breaks the phenoxo-bridged binuclear complexes into mononuclear bis-pyridine adducts if a stronger ligand such as imidazole is absent in the amino acid. The presence of histidine, the imidazole in 1 probably displaces the in-plane pyridine ligand to form the cyclic capsular complex in 1a. Role of Imidazole. Above we observed that formation of octanuclear 1b occurred most likely because of displacement of in-plane pyridine by the imidazole arm of histidine from another ligand. Also, addition of excess imidazole influences carboxylate to coordinate axially in 1c. One possibility is that imidazole, being a stronger ligand, tends to occupy the equatorial position rather than the axial position even at the expense of shifting other coligands to the axial position.12 This stronger preference of imidazole for the equatorial position is driving the ligand coordination mode switching in 1c. This has been further substantiated by the synthesis and structural characterization of square planar 2c from the corresponding binuclear 2a upon addition of 2 equiv of imidazole. In 2c, the imidazole coordinates as in-plane ligand only (Figure 4). Addition of up to 4 equiv of imidazole did not yield any five-coordinated complex with axially coordinated imidazole. The isolation and structural characterization of this complex substantiate that the tendency of imidazole to coordinate in-plane was responsible for (a) capsule formation by displacing a pyridine in 1b and (b) axial carboxylate coordination in 1c where both externally added imidazole and histidine imidazole occupied two in-plane positions.

Overall, these complexes preferably form bis-phenoxo binuclear complexes in weakly coordinating solvent. Solvents or reagents with better binding ability to Cu(II) compared to phenolate such as pyridine will break the phenoxo bridge and form mononuclear complexes. If the amino acid used is histidine, then replacement of the in-plane ligand can lead to cyclic multinuclear complex (tetramer in 1b) formation. The generalized process is shown in Scheme 3. Absorption and EPR Spectral Characteristics. The UVvisible spectral characteristic of all the complexes are given in the Experimental Section. The absorption maxima ∼400 nm (, ∼500 M-1 cm-1) is of LMCT origin, and the absorption maxima between 600 and 680 nm (, ∼150 M-1 cm-1) are of ligand field origin. Several other square-pyramidal Cu(II) complexes with an N/O donor environment have similar spectral characteristics.3,13 The similarity of spectra of 1b and 1c in MeOH reflects their N3O2 coordination environment. Similarly, the spectral characteristics of 2b, 3b, and 4b are almost identical. The complex 2c, being the only complex with four coordination, has a ligand-field transition at 632 nm, quite different from the others at 650 nm or above. This suggests retention of squareplanar geometry in solution. Similar to the observations of absorption spectral characteristics, EPR spectral data at 77 K (Experimental Section) of all the complexes except 2c are consistent with their square pyramidal structure as evident from their AII values ∼175 G and g values.14 The complex 2c, being a square planar complex, shows higher AII and lower gII values.15 The solution spectral properties of the complexes are consistent with their solid-state structure. Intermolecular Network. Amino acid-derived ligands such as ours have shown a variety of network formation because of the presence of multiple H-bond donor and acceptor in the ligand.16 The multiple possibilities make it difficult to form a particular type of network. In the present set, the phenoxo binuclear complex has all inplane sites around Cu(II) filled, leaving only the weak axial sites (Jahn-Teller distortion) for coordination from water or carboxylate from another molecule if water is scarce. Thus, 2a crystallized from dry DMF has carboxylate coordination, forming a coordination polymer (Figure 5a), but other dinuclear complexes have water bound axially.3

Reversible Disassembly of a Cu(II) Capsule Scheme 3.

Crystal Growth & Design, Vol. 7, No. 9, 2007 1823

Effect of Amino Acid, Solvent, and Imidazole on the Formation of Mononuclear to Octanuclear Species

In pyridine adducts 3b and 4b, the presence of two pyridines generates several C-H‚‚‚O/N weak interactions but limits H-bonding to one between phenol and carboxylate in 4b. Because of the planarity of 2c and the presence of imidazole, complementary H-bonding of carboxylate oxygen and imidazole NH (N3a‚‚‚O3 2.741 Å, N3‚‚‚O3a 2.761 Å) between the adjacent molecules and vertical H-bonding of amine and phenolate (N1‚‚‚O1a 3.061, N1a‚‚‚O1 3.153 Å) forms a 2D honeycomb network (Figure 5b). Increase in the number of imidazoles and axial coordination of carboxylate increases the complexity of the H-bond in 1c and a resulting in a network which has the ability of supporting water chains stabilized through H-bonding with carboxylate, phenolate, and imidazole (Figure 5c).17 Conclusions Reduced Schiff base ligands from amino acid and salicylaldehyde are chiral, multidentate, and easy to synthesize ligands which showed their versatility forming molecules with different nuclearity as well as H-bonded networks in the presence of metal ion. However, the effect of amino acid side chain, coordinating solvent, and coordination chemistry of Cu(II) on the nuclearity and the networking ability of these complexes was never studied

thoroughly. In this paper, using a set of Cu(II) complexes with different amino acid and solvents we attempted to rationalize the various factors responsible for generating such diverse structures. A unified scheme describing how these factors led to different structures has been presented (Scheme 3). In the process we were able to demonstrate that the capsular cavity in 1b can be reversibly disassembled using the competition between ligand imidazole and external imidazole (pKa 6.92). The fact that the 1b was isolated intact from pyridine (pKa 5.19) showed that the tetrameric units of the capsule is immune to weaker ligands. Addition of weaker pyrazine (pKa 0.78) did not yield any disassembled product. Thus, the disassembly process was triggered with a specific molecule within the range of N donors used in this study. This is unlike the pH (a bulk property of the solution)-dependent disassembly process observed previously.4b,c The structural analyses of the complexes showed formation of a two-dimensional grid in 2a and two-dimensional honeycomb network in 2c (Figure 5). Identification of the factors governing the nuclearity and geometry of the Cu(II) complexes along with choice of amino acid might help designing metal complexes and networks with this type of ligand in a more rational way (Figure S1, Supporting Information).

Figure 5. Coordination polymeric network (a) in 2a and H-bonded honeycomb network (b) in 2c and chain of water (c) in 1c.

1824 Crystal Growth & Design, Vol. 7, No. 9, 2007

Acknowledgment. We thank the Department of Science and Technology (DST), New Delhi (Grant SR/S1/IC-49/2003 to M.R.), and CSIR, New Delhi (Grant no. 01/(1669)/00/EMR-II to M.R. and 9/731(41)/2005-EMR-I to R.R.K.) for funding. We thank S. Supriya, Dr. Samar K. Das of the Department of Chemistry of Hyderabad, IIT Kanpur, Central Instrument Facility of IIT Guwahati, and CDRI Lucknow for EPR and the analytical facility. Some of the structural data were collected using a X-ray diffractometer provided by FIST-DST, New Delhi, at the Department of Chemistry, IIT Guwahati. Supporting Information Available: X-ray crystallographic file of the structures in CIF format; bond parameters, network, TGA, powder diffraction, and UV-vis spectra. This material is available free of charge via the Internet at http://pubs.acs.org.

Alam et al.

(5) (6) (7) (8) (9)

(10) (11)

References (1) (a) Stang, P. J.; Olenyuk, B. Acc. Chem. Res. 1997, 30, 502. (b) Rebek, J., Jr. Acc. Chem. Res. 1999, 32, 278. (c) Jasat, A.; Sherman, J. C. Chem. ReV. 1999, 99, 931. (d) Seidel, S. R.; Stang, P. J. Acc. Chem. Res. 2002, 35, 972. (e) Hof, F.; Craig, S. L.; Nuckolls, C.; Rebek, J., Jr. Angew. Chem., Int. Ed. 2002, 41, 1489. (f) Kang, J.; Hilmersson, G.; Santamaria, J.; Rebek, J., Jr. J. Am. Chem. Soc. 1998, 120, 3650. (g) Kang, J.; Santamaria, J.; Hilmersson, G.; Rebek, J., Jr. J. Am. Chem. Soc. 1998, 120, 7389. (h) McKinlay, R. M.; Thallapally, P. K.; Cave, G. W. V.; Atwood, J. L. Angew. Chem., Int. Ed. 2005, 44, 5733. (i) Heaven, M. W.; Cave, G. W. V.; Mckinlay, R. M.; Antesberger, J.; Dalgarno, S. J.; Thallapally, P. K.; Atwood, J. L. Angew. Chem., Int. Ed. 2006, 45, 6224. (j) McKinlay, R. M.; Thallapally, P. K.; Atwood, J. L. Chem. Commun. 2006, 2956 (2) Alam, M. A.; Nethaji, M.; Ray, M. Angew. Chem., Int. Ed. 2003, 42, 1940. (3) (a)Yang, C. T.; Vetrichelvan, M.; Yang, X; Keith, B. M.; Murray, S.; Vittal, J. J. J. Chem. Soc., Dalton Trans. 2004, 113-121. (b) Yang, C. T.; Moubaraki, B.; Murray, K. S.; Ranford, J. D.; Vittal, J. J. Inorg. Chem. 2001, 40, 5934. (c) Xiandong, Y.; Daqing, W.; Ranford, J. D.; Vittal, J. J. Cryst. Growth Des. 2005, 5, 41. (d) Xiandong, Y.; Ranford, J. D.; Vittal, J. J. Cryst. Growth Des. 2004, 4, 781. (e) Ranford, J. D.; Vittal, J. J.; Wu, D. Angew. Chem., Int. Ed. 1998, 37, 1114. (f) So, K. W.; Yang, C. T.; Vittal, J. J.; Ranford, J. D. Inorg. Chim. Acta 2003, 349, 135. (g) Wang, X.; Ding, J.; Vittal, J. J. Inorg. Chim. Acta 2006, 359, 3481. (4) (a) Wan, Y.; Mitkin, O.; Barnhurst, L.; Kurchan, A.; Kutateladze, A. Org. Lett. 2000, 2, 3817. (b) Sunatsuki, Y.; Motoda, Y.; Matsumoto, N. Coord. Chem. ReV. 2002, 226, 199. (c) Carbonaro,

(12)

(13) (14) (15) (16)

(17)

L.; Isola, M.; Liuzzo, V.; Marchetti, F.; Balzano, F.; Pomelli, C. S.; Raffaelli, A. Eur. J. Inorg. Chem. 2001, 353. (d) de Groot, F. M. H.; Albrecht, C.; Koekkoek, R.; Beusker, P. H.; Scheeren, H. W. Angew. Chem., Int. Ed. 2003, 43, 4490. Sheldrick, G. M. SHELXL-97, Program for the Solution of Crystal Structures; University of Go¨ttingen: Go¨ttingen, Germany. Johnson, C. K. ORTEP, Report ORNL-3794; Oak Ridge National Laboratory: Oak Ridge, TN, 1976. Nakamoto, K. Infrared and Raman Spectra of Inorganic Compounds, 5th ed.; Wiley-Interscience: New York, 1997; Part B. Geary, W. J. Coord. Chem. ReV. 1971, 7, 81. (a) Earnshaw, A. Introduction to Magnetochemistry; Academic Press: London, 1968. (b) Figgis, B. N.; Lewis, J. Prog. Inorg. Chem. 1964, 6, 37. Addision, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G. C. J. Chem. Soc., Dalton Trans. 1984, 1349. (a) Anderson, O. P.; Packard, A. B. Inorg. Chem. 1980, 19, 2123. (b) Barclay, G. A.; Kennard, C. H. L. J. Chem. Soc. 1961, 5244. (c) Tables of Interatomic Distances and Configuration in molecules and Ions. Spec. Publ. - Chem. Soc. 11; Sutton, L. E., Ed.; 1958. (a) Morehouse, S. M.; Suliman, H.; Haff, J.; Nguyen, D. Inorg. Chim. Acta 2000, 297, 411. (b) Bugella-Altamiranoa, E.; GonzalÄez-Per´eza, J. M.; Sicilia-Zafraa, A. G.; Niclos´-Gutier´reza, J.; Castin˜eirasCampos, A. Polyhedron 1999, 18, 3333. (c) Iskander, M. F.; Khalil, T. E.; Werner, R.; Haase, W.; Svoboda, I.; Fuess, H. Polyhedron 2000, 19, 1181. Admas, H.; Bailey, N. A.; de Barbarin, C. O. R.; Fanton, D. E. J. Chem. Soc., Dalton Trans. 1995, 2323. Yokoi, H.; Addison, A. W. J. Chem. Soc., Dalton Trans. 1977, 16, 1341. Sakaguchi, U.; Addison, A. W. J. Chem. Soc., Dalton Trans. 1979, 600. N(H)‚‚‚O range: 2.69 to 2.89 Å. (a) Couchman, S. M.; Jeffery, J. C.; Ward, M. D. Polyhedron 1999, 18, 2633. (b) Kuduva, S. S.; Bla¨ser, D.; Boese, R.; Desiraju, G. R. J. Org. Chem. 2001, 66, 1621. (c) Alam, M. A.; Nethaji, M.; Ray, M. Inorg. Chem. 2005, 44, 1302. N(H)‚‚‚S range: 3.25 to 3.45 Å. (d) Ueyama, N.; Nishikawa, N.; Yamada, Y.; Okamura, T.; Oka, S.; Sakurai, H.; Nakamura, A. Inorg. Chem. 1998, 37, 2515. (a) Ludwig, R. Angew. Chem., Int. Ed. 2003, 42, 3458. (b) Pal, S.; Sankaran, N. B.; Samanta, A. Angew. Chem., Int. Ed. 2003, 42, 1741. (c) Mukherjee, A.; Saha, M. K.; Nethaji, M.; Chakravarty, A. R. Chem. Commun. 2004, 716. (d) Lioyd, G. O.; Atwood, J. L.; Barbour, L. J. Chem. Commun. 2005, 1845.

CG070250S