Raman Detected Sensing of Volatile Organic Compounds by

Feb 17, 2015 - Raman spectrum confirms the formation of Au−X (X = Cl, Br) bonds by the presence of vAuX peaks at 340 cm. −1. (vAuCl) for 1 and 206...
2 downloads 0 Views 3MB Size
Article pubs.acs.org/cm

Raman Detected Sensing of Volatile Organic Compounds by Vapochromic Cu[AuX2(CN)2]2 (X = Cl, Br) Coordination Polymer Materials Jeffrey S. Ovens and Daniel B. Leznoff* Department of Chemistry, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia V5A 1S6, Canada S Supporting Information *

ABSTRACT: Two vapochromic coordination polymers Cu[AuX2(CN)2]2 (X = Cl, 1; X = Br, 2) were prepared and spectroscopically characterized. Exposure of these solid materials to the volatile organic compounds dimethylformamide (DMF), dimethyl sulfoxide (DMSO), pyridine, 1,4dioxane, and ethylene glycol (glycol) resulted in distinct color, and IR and Raman changes. The thermal stability of the analyte-bound materials was assessed by thermogravimetric analysis. Single-crystal structures of Cu(analyte)4[AuX2(CN)2]2 (analyte = DMF, DMSO; X = Cl, Br) revealed an isostructural set of 1-D coordination polymer chains, where the analyte molecules were equatorially Obound to the Cu(II) centers while axially bound [AuX2(CN)2]− units bridged these Cu(II) centers, while Cu(glycol)4[AuBr2(CN)2]2 is molecular, with monodentate glycol units. The structure of Cu2(OH2)4[AuCl2(CN)2]4·4dioxane is a 2-D coordination polymer network with H2O-bridged Cu(II) centers and dioxane units hydrogen bonded between the 2-D sheets. The intense Raman vCN stretches for 1, 2, and their adducts form distinct, signature patterns. These “antenna” Raman vCN stretches are an effective means for sensing VOCs, and their characteristic patterns can be used to identify the VOC being detected.



INTRODUCTION Over the past decade, coordination polymers have been receiving increased attention due to the ability to tailor them toward various properties such as magnetism,1−6 porosity,7−12 luminescence,13−18 and vapochromism19−25 through the strategic choice of metal centers and bridging and ancillary ligands.26−31 More recently, coordination polymers have been explored as solid-state sensors for the detection of explosive materials32−34 and volatile organic compounds (VOCs) in industrial settings.16,24,25,35 Such devices have important applications in defense, industry, and safety and security.36,37 Coordination polymers such as the Co2+/[Re6Q8(CN)6]4−based (Q = S, Se) Prussian Blue analogues19 can act as vapochromic sensors due to their highly visible color change upon exposure to certain VOCs. The porous metal−organic framework [(WS4Cu4)I2(dptz)3]·3DMF (dptz = 3,6-di(pyridin4-yl)-1,2,4,5-tetrazine; DMF = dimethylformamide) exhibits solvatochromism when immersed in solvents such as MeCN and CHCl3, lending it the potential to sense the presence of these small molecules in solution.38 In another example, cyanoaurate-based coordination polymers have demonstrated high potential as VOC sensors. For example, the M[Au(CN)2]2 (M = Co, Cu) series exhibits stark color changes when exposed to a range of VOCs,24,25 whereas the four polymorphs of Zn[Au(CN)2]2 display a significant change in luminescence in the presence of ammonia vapor.16 Although not substantially © XXXX American Chemical Society

porous, these materials bind and detect certain vapor-phase molecules due to the flexible nature of their coordination polymer frameworks.39 Gold-containing materials in general have also been extensively employed as vapochromic sensors.40−60 For example, the organometallic Au/Ag-based system AuAg(C6F5)2L (L = pyridine, 2,2′-bipyridine, 1,10-phenanthroline, diphenylacetylene, and others)41 changes from its red or orange color to white when exposed to certain coordinating vapors, such as MeOH, EtOH, and acetone. This change occurs when the coordinating solvent molecules bind to the Ag centers, breaking the Ag−Au bonds. These materials have also been used in conjunction with optical fibers, where the change in the material’s refractive index provides the sensory readout.46−54 Several other examples of mixed Au/M (M = Tl, Ag, Cu) compounds56−59 also exist, such as Tl[Au(C6Cl5)2]57 and [Au(im(CH2py)2)2(Cu(MeCN)2)2](PF6)3.59 The former exhibits vapochromism and vapoluminescence when exposed to vapors such as THF, acetone, and MeCN by virtue of changes to Au−M bonds in the structure, while the latter undergoes ligand exchange when exposed to vapors such as MeOH and MeCN. In another example, dimeric gold-dithiocarbamate Received: August 14, 2014 Revised: October 9, 2014

A

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Infrared spectra were measured on a Thermo Nicolet Nexus 670 FT-IR spectrometer equipped with a Pike MIRacle attenuated total reflection (ATR) sampling accessory. Raman spectra were measured using a Renishaw inVia Raman microscope equipped with a 785 nm laser. Solid-state UV−visible reflectance spectra were measured using an Ocean Optics SD2000 spectrophotometer equipped with a tungsten halogen lamp. Microanalyses (C, H, N) were performed by Frank Haftbaradaran or Paul Mulyk at Simon Fraser University on a Carlo Erba EA 1110 CHN elemental analyzer. Thermogravimetric analysis (TGA) data were collected by Dr. Rajendra Sharma at Simon Fraser University using a Shimadzu TGA-50 instrument in an air atmosphere at 1 or 2 °C per minute. 1·py, 2·py, 1, and 2·dioxane lose analyte quickly, precluding accurate elemental analysis of the analytesaturated materials. Synthetic Procedures. Cu[AuCl2(CN)2]2 (1). Cl2 gas was bubbled through a 20 mL green methanolic suspension of Cu(OH2)2[Au(CN)2]2 (380 mg; 0.63 mmol) for 5 min, resulting in a green solution. Nitrogen gas was then bubbled through this solution for 5 min, with no visible color change. Following gravity filtration, a pale blue powder containing primarily Cu[AuCl2(CN)2]2 (1) was collected by rotary evaporation (332 mg; 74% yield). Recrystallization from H2O generates large crystals of Cu(OH2)4[AuCl2(CN)2]2·2H2O (1·H2O), which were separated manually and washed in clean H2O. Grinding of these crystals or treatment under reduced pressure rapidly regenerates 1. Analyses for 1, IR (cm−1): 2245 (s, vCN). Raman (cm−1): 2259 (s, vCN), 578 (m), 416 (m), 369 (w), 336 (vs, vAuCl), 260 (s), 117 (s). Anal. Calcd for C4N4Au2Cl4Cu: C, 6.83; H, 0.00; N, 7.97. Found: C, 6.89; H, 0.09; N, 8.04. Analyses for 1·H2O, IR (cm−1): 2242 (s, vCN), 2186 (w, vCN), 3579 (s), 3521 (vs), 3250 (s, br), 1630 (m, br), 854 (w, br). Raman (cm−1): 2251 (s, vCN), 2197 (s, vCN), 538 (m), 466 (s), 399 (w), 371 (w), 340 (vs, vAuCl), 238 (m), 206 (w), 153 (s), 137 (vs). Anal. Calcd for C4N4Au2Cl4Cu·6H2O: C, 5.92; H, 1.49; N, 6.90. Found: C, 5.85; H, 1.41; N, 6.67. 1·H2O can also be obtained by vapor absorption of H2O in a H2O-saturated environment by 1. Cu[AuBr2(CN)2]2 (2). 1.5 mL of neat Br2 was added to a 20 mL green aqueous suspension of Cu(OH2)2[Au(CN)2]2 (206 mg; 0.34 mmol) and stirred for ca. 15 min, resulting in a mixture of a brown solution and excess unreacted Br2. Nitrogen gas was then bubbled through this mixture for 10 min, resulting in a green solution. Following gravity filtration, a dark green powder of Cu[AuBr2(CN)2]2 (2) was collected by rotary evaporation (199 mg; 66% yield). IR (cm−1): 2238 (s, vCN). Raman (cm−1): 2250 (s, vCN), 572 (s), 414 (s), 329 (s), 253 (m), 206 (vs, vAuBr), 120 (m). Anal. Calcd for C4N4Au2Br4Cu: C, 5.45; H, 0.00; N, 6.36. Found: C, 5.41; H, 0.00; N, 6.29. Cu(DMF)4[AuCl2(CN)2]2 (1·DMF). X-ray quality crystals of 1·DMF were obtained by slow evaporation of 1 from a DMF solution. IR (cm−1): 2193 (m, vCN), 2950 (w), 1644 (vs), 1435 (m), 1376 (s), 1251 (w), 1117 (m), 1060 (w). Raman (cm−1): 2205 (s, vCN), 2182 (s, vCN), 1498 (w), 1433 (s), 1423 (s), 1373 (w), 1120 (m), 1013 (w), 941 (w), 869 (s), 702 (w), 682 (w), 471 (w), 442 (s), 415 (w), 393 (m), 338 (vs, vAuCl), 286 (w), 233 (m), 172 (m), 135 (s). Anal. Calcd for C4N4Au2Cl4Cu·4C3H7NO: C, 19.30; H, 2.83; N, 11.25. Found: C, 18.64; H, 2.63; N, 10.80. The same product can also be obtained by vapor absorption of DMF by 1 or 1·H2O. Cu(DMSO)4[AuCl2(CN)2]2 (1·DMSO). 1·DMSO powder was obtained by vapor absorption of DMSO by 1. IR (cm−1): 2235 (m, vCN), 3004 (w), 1427 (m), 1410 (w), 1321 (w), 1035 (m), 992 (s), 944 (s). Raman (cm−1): 2243 (s, vCN), 2185 (s, vCN), 1416 (m), 996 (m), 958 (w), 721 (s), 684 (s), 535 (m), 450 (s), 338 (vs, vAuCl), 237 (m), 214 (w), 142 (s). Anal. Calcd for C4N4Au2Cl4Cu·4C2H6OS: C, 14.19; H, 2.38; N, 5.52. Found: C, 14.40; H, 2.30; N, 5.44. Cu2(OH2)4[AuCl2(CN)2]4·4dioxane (1·dioxane). X-ray quality crystals of 1·dioxane were obtained by layering dioxane over an H2O solution of 1. IR (cm−1): 2234 (w, vCN), 2204 (m, vCN), 2974 (m), 2923 (m), 2863 (m), 1455 (m), 1295 (w), 1253 (s), 1115 (s), 1076 (s), 1043 (w), 890 (m), 865 (vs). Raman (cm−1): 2253 (m, vCN), 1444 (w), 1310 (w), 1218 (w), 1130 (w), 1013 (w), 848 (w), 834 (s), 543 (w), 452 (w), 340 (vs, vAuCl), 241 (m). 1·dioxane desolvates rapidly upon removal from saturated dioxane atmosphere, precluding

chains, such as Au2[S2CN(C5H11)2]2, exhibit both a vapochromic response and a “switching on” of luminescence when exposed to polar aprotic solvents such as CH2Cl2, CHCl3, and MeCN.42 This change is a result of changes in the intermolecular Au···Au interaction distances.42,61 The related compound, [Au2(dppm)(TU)][CF3COO]2 (dppm = bis(diphenylphosphino)methane; TU = 2-thiouracyl), exhibits loss of emission when exposed to CF3COOH vapor.44 Another binuclear material, (C6F5Au)2(μ-1,4-diisocyanobenzene), exhibits mechanochromism,62 where the emission wavelength changes upon grinding of the material. However, once a solvent or vapor is exposed to it, the emission returns to its original color. The trimeric Au-based system Au3(CH3NCOCH3)3 exhibits emission solvatochromism,45 where direct exposure of the material to liquid solvent elicits a change in the emissive properties. In this case, exposure to neat CHCl3 “turns on” yellow emission. Although color changes are ideal for visual alert purposes, the visible reflectance peaks associated with such materials tend to be broad, resulting in a large overlap, thus limiting their absolute sensitivity in electronic monitoring applications. Another potentially useful detection readout handle of cyanometallate-based coordination polymers is their characteristic stretches between 2100 and 2300 cm−1, usually assessed by IR spectroscopy.63 The high sensitivity of these vibrational signals to the cyanide symmetry and chemical environment provides an additional means of monitoring the change in the material as VOCs interact with it.24,25 For example, when exposed to a particular VOC, the aforementioned M[Au(CN)2]2 displays a unique pattern of vCN peaks characteristic to that specific VOC. Raman spectroscopy is complementary to IR, but although Raman-based techniques such as surface-enhanced Raman scattering, resonance Raman, and direct measurement of the analyte Raman spectrum have been widely employed as detection methods in sensor applications,64−66 the use of distinct vibrationally active moieties in a sensory material, for example, the cyano units in cyanometallate coordination polymers, as a detection readout has not been investigated to our knowledge. A strong Raman signal characteristic of the material (e.g., symmetric vCN stretches) and sensitive to the chemical environment could act as a transducer of the Raman signal of the analyte. Thus, herein we report the synthesis and spectroscopic and structural characterization of the two new vapochromic [AuX2(CN)2]−-based (X = Cl, Br) coordination polymers Cu[AuCl2(CN)2]2 and Cu[AuBr2(CN)2]2 in their parent form as well as the materials that result from exposure to several common VOC analytes. We show that Raman detection of the antenna vCN bands in the materials is an effective method to achieve sensing of VOC analytes.



EXPERIMENTAL SECTION

General Procedures and Materials. Caution! Chlorine, bromine, ammonia, and volatile organic compounds (VOCs) should only be used in a well ventilated fumehood. Although cyanoaurate is significantly less toxic than other cyanometallates (because the Au− CN formation constant is very high67) and it also exhibits bioactivity,68 typical precautions regarding handling of hazardous materials should still be followed. All reactions were performed in air. Cu(OH2)2[Au(CN)2]2 and K[AuCl2(CN)2] were synthesized using literature procedures.25,69 All other reagents were obtained from commercial sources and used as received. B

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials Table 1. Crystallographic Data for Compounds 1·H2O, 1·DMF, 2·DMF, and 2·DMSO empirical formula formula weight (g mol−1) crystal dimensions (mm) crystal system space group a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) V (Å 3) Z T (K) ρcalcd (g cm−3) μ (mm) 2θmax (deg) total/unique reflns reflns [I0 ≥ 2σ(I0)] R1, wR2 [I0 ≥ 2σ(I0)]a GOF largest difference peak/hole (e−/Å3) CCDC no. a

1·H2O

1·DMF

2·DMF

2·DMSO

C4H12Au2Cl4CuN4O6 811.45 0.36, 0.44, 0.68 triclinic P1̅ 5.0900(2) 9.2800(3) 9.3946(3) 84.733(1) 84.283(1) 89.235(1) 439.68(3) 1 150(2) 3.065 18.472 72.362 22 353/4226 3941 0.0211, 0.0524 1.177 3.214/−2.060 1019370

C16H28Au2Cl4CuN8O4 995.74 0.14, 0.18, 0.19 triclinic P1̅ 8.4718(1) 8.6120(1) 10.6612(1) 86.056(1) 81.895(1) 86.063(1) 766.869(15) 1 296(2) 2.156 10.612 66.282 17 417/5805 4646 0.0263, 0.0674 1.059 1.107/−0.773 1019373

C16H28Au2Br4CuN8O4 1173.58 0.067, 0.14, 0.22 triclinic P1̅ 8.2765(2) 8.4243(2) 10.8307(3) 90.153(2) 92.457(2) 92.463(1) 762.86(3) 1 150(2) 2.555 15.556 68.674 16 849/6359 5336 0.0266, 0.0656 1.065 2.792/−1.780 1019372

C12H24Au2Br4CuN4O4S4 1193.70 0.040, 0.094, 0.096 triclinic P1̅ 8.2802(7) 8.5080(7) 11.2279(9) 80.1800(19) 79.306(2) 86.8250(19) 765.63(11) 1 296(2) 2.589 15.761 52.742 13 944/3131 2358 0.0422, 0.1138 1.037 3.417/−1.570 1019374

Function minimized: ∑w(F2o − F2c )2. R1 = ∑∥Fo| − |Fc∥/∑|Fo| and wR2 = [∑w(F2o − F2c )2/∑wF4o]1/2.

elemental analysis. The same product can also be obtained by vapor absorption of dioxane and ambient H2O by 1, or by direct immersion of 1 in dioxane solvent. Cu(py)x[AuCl2(CN)2]2 (1·py; x > 3). 1·py powder was obtained by vapor absorption of pyridine (py) by 1. IR (cm−1): 3098 (m), 3065 (m), 3037 (m), 3006 (m), 1606 (s), 1580 (m), 1488 (m), 1454 (s), 1419 (w), 1224 (w), 1205 (m), 1148 (w), 1102 (m), 1075 (s), 1043 (m), 1017 (w), 991 (w), 771 (s). Raman (cm−1): 2185 (m, vCN), 1606 (m), 1571 (w), 1566 (w), 1483 (w), 1218 (w), 1143 (w), 1068 (w), 1046 (m), 1022 (s), 989 (w), 659 (m), 645 (w), 439 (m), 921 (s), 338 (s, vAuCl), 301 (w), 221 (s), 198 (w), 124 (m). Anal. Calcd for C4N4Au2Cl4Cu·3.5C5H5N: C, 26.34; H, 1.80; N, 10.72. Found: C, 26.60; H, 1.82; N, 10.50. Cu(glycol)x[AuCl2(CN)2]2 (1·glycol). 1·glycol powder was obtained by vapor absorption of ethylene glycol (glycol) by 1. IR (cm−1): 2236 (m, vCN), 2206 (m, vCN), 3360 (s, br), 2976 (m), 2957 (m), 2941 (m), 2885 (m), 2780 (w), 1454 (m), 1413 (m), 1386 (m), 1362 (m), 1316 (m), 1260 (w), 194 (w), 1080 (vs), 1033 (vs), 871 (s). Raman (cm−1): 2221 (m, vCN), 1046 (s), 1020 (s), 862 (m), 651 (w), 337 (vs, vAuCl), 235 (m), 163 (m). Because of the low vapor pressure of ethylene glycol, it was not possible to fully convert Cu[AuCl2(CN)2]2 to its glycol adduct, 1·glycol, precluding accurate elemental analysis. Cu(DMF)4[AuBr2(CN)2]2 (2·DMF). X-ray quality crystals of 2·DMF were obtained by slow evaporation of 2 from a DMF solution. IR (cm−1): 2183 (vw, vCN), 2955 (w), 1644 (vs), 1493 (w), 1434 (m), 1370 (s), 1247 (w), 1118 (m), 1060 (w). Raman (cm−1): 2196 (m, vCN), 2184 (m, vCN), 1492 (w), 1434 (s), 1422 (s), 1368 (w), 1120 (m), 869 (s), 704 (m), 451 (s), 413 (w), 396 (m), 306 (w), 233 (w), 206 (vs, vAuBr). Anal. Calcd for C4N4Au2Br4Cu·4C3H7NO: C, 16.38; H, 2.40; N, 9.55. Found: C, 16.09; H, 2.32; N, 9.32. The same product can also be obtained by vapor absorption of DMF by 2. Cu(DMSO)4[AuBr2(CN)2]2 (2·DMSO). X-ray quality crystals of 2·DMSO were obtained by slow evaporation of 2 from a DMSO solution. IR (cm−1): 2189 (vw, shoulder, vCN), 2175 (w, vCN), 3011 (w), 2996 (w), 2921 (w), 1423 (m), 1411 (w), 1395 (w), 1320 (m), 1297 (w), 1024 (m), 989 (s), 945 (s), 719 (w). Raman (cm−1): 2226 (s, vCN), 2190 (m, vCN), 1417 (w), 989 (w), 941 (w), 725 (s), 688 (s), 520 (m), 455 (s), 410 (m), 315 (w), 205 (vs, vAuBr). Anal. Calcd for C4N4Au2Br4Cu·4C2H6OS: C, 12.07; H, 2.03; N, 4.69. Found: C,

11.90; H, 1.93; N, 4.59. The same product can also be obtained by vapor absorption of DMSO by 2. Cu(dioxane)x[AuBr2(CN)2]2 (2·dioxane). 2·dioxane powder was obtained by vapor absorption of dioxane by 2. IR (cm−1): 2234 (m, br, vCN), 2957 (m), 2917 (m), 2887 (w), 2856 (m), 1453 (m), 1438 (w), 1364 (w), 1291 (m), 1254 (s), 1118 (s), 1079 (m), 1045 (w), 888 (m), 866 (vs). Raman (cm−1): 2241 (m, br, vCN), 2197 (w), 1088 (w), 1011 (w), 848 (w), 829 (m), 551 (m), 481 (w), 329 (w), 202 (vs, vAuBr). 2·dioxane desolvates rapidly upon removal from saturated dioxane atmosphere, precluding elemental analysis. The same product can also be obtained by direct immersion of 2 in dioxane solvent. Cu(py)x[AuBr2(CN)2]2 (2·py; x > 3). 2·py powder was obtained by vapor absorption of pyridine (py) by 2. IR (cm−1): 2169 (m, vCN), 2154 (w, vCN), 2144 (w), 3068 (m, br), 1606 (vs), 1488 (s), 1447 (vs), 1238 (w), 1216 (m), 1158 (w), 1070 (s), 1043 (m), 1016 (m), 759 (vs). Raman (cm−1): 2219 (w, vCN), 2186 (m, vCN), 1608 (m), 1574 (m), 1488 (w), 1219 (w), 1158 (w), 1068 (w), 1046 (m), 1023 (s), 1019 (s), 650 (m), 642 (m), 455 (m), 299 (w), 241 (m), 206 (vs, vAuBr). Anal. Calcd for C4N4Au2Br4Cu·3C5H5N: C, 20.40; H, 1.35; N, 8.77. Found: C, 20.97; H, 1.41; N, 8.72. Cu(glycol)4[AuBr2(CN)2]2 (2·glycol). X-ray quality crystals of 2·glycol were obtained by slow evaporation of 2 from a 5:95 ethylene glycol/H2O solution. IR (cm−1): 2233 (m, vCN), 2206 (vw, vCN), 2188 (w, vCN), 3350 (s, br), 2976 (m), 2954 (m), 2939 (m), 2881 (m), 1438 (m, br), 1371 (m), 1327 (m, br), 1263 (w), 1183 (w), 1079 (vs), 1033 (s), 866 (s). Because of its sensitivity to the Raman laser, only the region between 2300 and 2100 cm−1 (i.e., the vCN region) was able to be collected. Raman (2300−2100 cm−1): 2234 (m, vCN), 2186 (w, vCN). Anal. Calcd for C4N4Au2Br4Cu·3C2H6O2 (partially desolvated from C4N4Au2Br4Cu·4C2H6O2): C, 11.25; H, 1.70; N, 5.25. Found: C, 11.21; H, 1.74; N, 5.00. The same product can also be obtained by vapor absorption of ethylene glycol by 2. [Cu(py)4]2[Cu(OH2)2(py)2][Au(CN)2]6·4H2O (3). Cu(NO3)2·6H2O (37 mg; 0.10 mmol) was dissolved in 10 mL of 5% v/v pyridine (py) in H2O, forming a dark blue solution. To this was added a colorless aqueous solution of K[AuCl2(CN)2] (71 mg; 0.20 mmol), resulting in no visible change. This solution was partially covered and set aside. After 1 day, blue needle-shaped crystals of [Cu(py)4]2[Cu(OH2)2(py)2][Au(CN)2]6·2H2O (3) formed and were collected by C

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials Table 2. Crystallographic Data for Compounds 2·glycol, 1·dioxane, and 3 empirical formula formula weight (g mol−1) crystal dimensions crystal system space group a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) V (Å3) Z T (K) ρcalcd (g cm−3) μ (mm) 2θmax (deg) total/unique reflns obsd reflns [I0 ≥ 2σ(I0)] R1, wR2 [I0 ≥ 2σ(I0)]a GOF largest difference peak/hole (e−/Å3) CCDC number a

2·glycol

1·dioxane

3

C12H24Au2Br4CuN4O8 1129.46 0.090, 0.29 0.41 monoclinic P21/n 11.62640(15) 8.0910(1) 14.85960(19) 90 104.8570(6) 90 1351.10(3) 2 299(2) 2.776 17.567 80.498 40 766/8489 5424 0.0345, 0.0636 1.007 1.987/−1.702 1019375

C12H20Au2Cl4CuN4O6 915.59 0.15, 0.22, 0.23 monoclinic P21/c 19.8327(4) 13.9821(3) 19.8634(5) 90 118.152(1) 90 4856.51(19) 8 150(2) 2.504 13.395 52.742 34 750/10 062 8782 0.0588, 0.0608 1.067 1.326/−1.164 1019371

C62H62Au6Cu3N22O6 2583.75 0.037, 0.052, 0.43 hexagonal P6222 18.5292(3) 18.5292(3) 19.6529(4) 90 90 120 5843.5(2) 3 150(2) 2.203 12.112 52.722 95 320/3994 3434 0.0300, 0.0713 1.202 1.745/−1.683 1019376

Function minimized: ∑w(F2o − F2c )2. R1 = ∑∥Fo| − |Fc∥/∑|Fo| and wR2 = [∑w(F2o − F2c )2/∑wF4o]1/2.

vacuum filtration. Crystals of 3 rapidly desolvated to the previously reported Cu(py)2[Au(CN)2]2 (confirmed by IR)25 upon exposure to air (15 mg; 21% yield based on Cu(py)2[Au(CN)2]2) precluding IR and elemental analyses. Rapid transfer to a cold stream from the mother liquor facilitated single-crystal X-ray diffraction analysis of 3. The Raman spectrum of a single crystal of 3 immersed in paratone oil was collected. Raman (cm−1): 2215 (w, vCN), 2200 (m, vCN), 2173 (w, vCN), 2158 (s, vCN), 1685 (m), 1608 (w), 1573 (w), 1255 (w), 1216 (w), 1069 (vw), 1045 (s), 1020 (vs), 650 (m), 639 (w), 341 (m), 302 (m), 174 (m), 137 (w), 113 (s). Compound 3 can also be made using K[AuBr2(CN)2] instead of K[AuCl2(CN)2] or by slow evaporation of a Cu[AuCl2(CN)2]2 or Cu[AuBr2(CN)2]2 solution in 5% v/v py in H2O. Analyte Identification Experiments. Uniform powder samples of 1, 2, 1·DMF, 2·DMF, 1·DMSO, 2·DMSO, 1·py, and 2·py were placed on a slide. The Raman spectra were then collected for each between 2300−2100 cm−1 using the lowest magnification objective (5×) with the 785 nm laser set to 1% power. Multiple spectra were taken (4−6) of separate powder samples for averaging purposes. Immediately after each data collection run, a set of spectra of K2[Zn(CN)4] was collected as an external sample using the same run conditions. Each spectrum was integrated ±5 cm−1 of the vCN stretches corresponding to Cu[AuX2(CN)2]2 and their DMF, DMSO, and pyridine adducts and then averaged over all runs. The spectra for K2[Zn(CN)4] were integrated ±5 cm−1 of its vCN stretch (2150 cm−1) and averaged over all runs. The final reported values are the quotient of the values obtained from Cu[AuX2(CN)2]2 and adducts and that from the external standard. X-ray Crystallographic Analysis. All crystallographic data (powder and single crystal) were collected on a Bruker SMART ApexII Duo diffractometer equipped with a Mo Kα (single crystal; λ = 0.7109 Å) TRIUMPH-monochromated source and a Cu Kα (powder; λ = 1.54184 Å) Incoatec microsource using ω and ϕ scans. All samples were mounted on MiTeGen sample holders using paratone oil. Data for 1·dioxane, 2·DMSO, and 3 were collected at 150 K; data for all others were collected at room temperature. Additional crystallographic information can be found in Tables 1−3. All single-crystal diffraction data were processed and initial solutions found with the Bruker ApexII software suite. Subsequent refinements

Table 3. Crystallographic Data for 1·DMSO empirical formula formula weight (g mol−1) crystal system space group a (Å) b (Å) c (Å) α (deg) β (deg) γ (deg) V (Å3) Z T (K) ρcalcd (g cm−3) 2θmin/2θmax (deg) Rexp Rp, wRp GOF

C12H24Au2Br4CuN4O4S4 1015.89 triclinic P1̅ 8.2402(3) 8.3175(3) 11.3369(3) 78.805(2) 77.630(2) 85.840(2) 744.13(4) 1 296(2) 2.213 5/45 0.00823 0.00753, 0.00956 1.162

were performed in ShelXle.70 Hydrogen atoms were placed geometrically and refined using a riding model. All PXRD powder pattern data were processed using the Bruker ApexII software suite. Indexing and Pawley and Rietveld refinements for 1·DMSO were performed using Topas Academic.71 Compound 1·dioxane crystallized as a pseudomerohedral twin. Analysis of twinning was performed using Rotax.72 Compound 3 crystallized as a merohedral twin. Twinning in this case was accounted for using Xprep (a part of the Bruker ApexII software suite). Diagrams were made using ORTEP-373 and POV-ray.74



RESULTS AND DISCUSSION Synthesis of Analyte-Free Cu[AuX2(CN)2]2 (X = Cl, Br) Materials. Cu[AuX2(CN)2]2 (X = Cl, Br) can be synthesized by mixing solutions of Cu[An]2·xH2O ([An]− = NO−3 , ClO−4 ) D

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 1. Powder samples of Cu[AuCl2(CN)2]2 (1 and 2) and Cu[AuBr2(CN)2]2, the results of their exposure to H2O vapor, and a variety of volatile organic compounds (VOCs) in the gas phase.

and [Cat][AuX2(CN)2] ([Cat]+ = K+, nBu4N+). However, because of the high solubility of the Cu[AuX2(CN)2]2 product in many solvents, it was very challenging to isolate this material free of the salt byproduct. To circumvent this purification issue, halogen (X2 = Cl2 or Br2) was added directly to a suspension of Cu(OH2)2[Au(CN)2]2,25 resulting in the in situ oxidative addition of X2 to the Au(I) center to yield the Au(III) building block [AuX2(CN)2]−.69,75,76 This solubilized the suspension, and pure light blue Cu[AuCl2(CN)2]2 (1) and olive green Cu[AuBr2(CN)2]2 (2), free of any other salts, could be isolated upon removal of the solvent. Product formation was evidenced by a shift of three vCN peaks in the IR spectrum from 2217, 2195, and 2171 cm−1 for Cu(OH2)2[Au(CN)2]225 to a single, weaker signal at 2245 cm−1 for 1 and 2238 cm−1 for 2. The absence of any signal at ca. 3500 or 1600 cm−1 indicated that the products are anhydrous, in contrast to the Au(I)-containing parent compound, which contains two bound H2O units. The Raman spectrum confirms the formation of Au−X (X = Cl, Br) bonds by the presence of vAuX peaks at 340 cm−1 (vAuCl) for 1 and 206 cm−1 (vAuBr) for 2.77 The vCN stretches of each product in the IR and Raman spectra differ by ca. 10 cm−1, with the Raman vCN signal (the symmetric stretch) at higher energy versus the IR vCN signal (the asymmetric stretch). Although single crystals of 1 and 2 could not obtained, because there is only a single vCN stretch in the IR and Raman spectra, all [AuX2(CN)2]− units are in a similar chemical environment. In addition, the blue-shifted nature of the peak with respect to free [AuX 2 (CN) 2 ] − (2185 cm −1 for [AuCl2(CN)2]−69 and 2167 cm−1 for [AuBr2(CN)2]−75) suggests that all [AuX2(CN)2]− units are likely bound in a bridging fashion, creating a multidimensional framework. The similarity in the powder X-ray diffractograms obtained for 1 and 2 suggests they adopt similar structures (see Supporting Information Figure S1). Vapochromic Behavior of Cu[AuCl2(CN)2]2 (1) and Cu[AuBr2(CN)2]2 (2). The previously reported Au(I)-based coordination polymer Cu(OH2)2[Au(CN)2]2 shows visible vapochromism when exposed to ammonia and organic vapors with donor atoms. Thus, 1 and 2 (which are Au(III)-based analogues of Cu(OH2)2[Au(CN)2]2)) were exposed to a variety of vapors to test whether they also exhibit vapochromism and to compare their behavior to the Au(I) analogue. Compounds 1 and 2, as prepared above, contain no

solvent and no Cu-bound analyte, and remain so when left exposed to ambient air. However, when exposed to an atmosphere saturated with water vapor, 1 picks up H2O, forming Cu(OH2)4[AuCl2(CN)2]2·2H2O (1·H2O), which exhibits a visible lightening in its color. This H2O adduct retains the H2O for several days once removed from the saturated H2O atmosphere, but eventually returns to the dehydrated form, Cu[AuCl2(CN)2]2. This behavior is different from Cu(OH2)2[Au(CN)2]2, which retains the bound units H2O indefinitely. Furthermore, the anhydrous Au(I)-based material, Cu[Au(CN)2]2, quickly picks up H2O from ambient air. Thus, the Au(III)-containing 1 and 2 have a distinct advantage in the application of vapor sensing, because ambient moisture is far less likely to interfere with their operation. Exposure of light blue 1 in the solid state to dimethylformamide (DMF) vapor results in a visible color change to the lighter, green-blue Cu(DMF)4[AuCl2(CN)2]2 (1·DMF), while olive green 2 changes to the light yellow-green Cu(DMF)4[AuBr2(CN)2]2 (2·DMF) upon the same treatment, as can be seen in Figure 1. The conversion of 1 or 2 to their DMF adducts is spectroscopically observable with DMF concentrations as low as 500 ppm. Exposure of 1·H2O to DMF vapor results in the same transformation to 1·DMF as evidenced by IR spectroscopy. However, this change cannot be reversed by exposing the DMF adduct to increased humidity; in fact, both 1·DMF and 2·DMF are stable for over a month in ambient air before any loss of DMF can be detected. Similarly, when 1 or 2 is exposed to a range of other VOCs, including dimethyl sulfoxide (DMSO), pyridine (py), and ethylene glycol (glycol), they exhibit other striking color changes. Most of these solvent adducts have a characteristic color (Figure 1), which can be quantified by a λmax in the visible reflectance spectrum (Tables 4 and 5). The response time for the uptake of vapor is within a few minutes for pyridine, DMF, dioxane, and H2O, but significantly longer (several hours) for DMSO and ethylene glycol, presumably due to their low vapor pressures. Exposure to some VOCs triggers a decomposition of the parent compound, usually into a combination of AuCN and other unidentified substances, as observable by the absence of a vAuX signal in the Raman spectrum and a strong vCN signal at 2235 cm−1 for AuCN. These include acetic acid (1 only; exposure to acetic acid has no effect on 2), ammonia, diethylamine, triethylamine, PhSH, and tetrahydrothiophene. E

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

strongly the VOCs are retained by the materials after exposure and uptake, thermogravimetric analysis (TGA) data were examined. The TGA plots for 1 and 2 are shown in Figure 2. In

Table 4. IR, Raman, and UV−Vis Reflectance Data for 1 and 1·analyte vCN absorption(s) (cm−1) analyte none H2O

IR

DMF

2245 (s) 2241 (s), 2186 (w) 2193 (w)

DMSO

2235 (m)

pyridine dioxane ethylene glycol

2234 (w), 2204 (m) 2234 (m), 2206 (m)

Raman

maximum visible reflectance (nm) 492 ± 3 485 ± 9

2259 (s) 2251 (s), 2197 (s) 2205 (s), 2182 (s) 2243 (s), 2185 (s) 2185 (m) 2253 (m)

468 ± 3 483 ± 7

2221 (m)

508 ± 6

530 ± 13 (broad) 504 ± 6

Table 5. IR, Raman, and UV−Vis Reflectance Data for 2 and 2·analyte vCN absorption(s) (cm−1) analyte

IR

none DMF

2238 (s) 2183 (vw)

DMSO

2189 (vw, shoulder), 2175 (w)

pyridine dioxane ethylene glycol

2234 (m) 2233 (m), 2206 (vw), 2188 (w)

Raman 2250 (s) 2197 (m), 2184 (m) 2227 (s), 2190 (m) 2219 (w), 2186 (m) 2245 (m, br) 2234 (s), 2186 (m)

maximum visible reflectance (nm)

Figure 2. Thermogravimetric analysis (TGA) plots of 1 and 2.

555 ± 5 543 ± 10 (broad)

both cases, decomposition begins with the loss of halogen at 190−210 °C for 1 and at 140−160 °C for 2, resulting in Cu[Au(CN)2]2. This product then loses cyanogen in two stages between 250 and 375 °C in both cases, consistent with data previously reported for Cu[Au(CN)2]2 and similar to other cyanometallate decompositions.25,81 TGA data for the H2O, DMF, DMSO, and ehtylene glycol adducts (Supporting Information Figures S2−S5) illustrate that with gentle heating (ca. 100 °C), thermal desorption of the analytes from 1 and 2 occurs over several hours, regenerating the parent materials. However, increased heating risks loss of halogen, which occurs immediately after loss of analyte in these cases; 1·dioxane and 2·dioxane both lose dioxane readily in ambient air and are thus reversible back to the parent compounds with little or no heating. In contrast, loss of pyridine from 1·py or 2·py occurs in several stages, completely overlapping the loss of halogen, precluding thermal desorption of pyridine from these complexes. In addition to being reversible by gentle heating, in some cases the adsorbed analyte can be replaced by a different VOC in the solid state. For example, exposure of either solid 1·DMSO or 2·DMSO to DMF vapor results in 1·DMF or 2·DMF, respectively. 1·DMF and 2·DMF can, in turn, be converted to 1·py and 2·py via exposure to pyridine vapor. The reverse transformations (e.g., conversion of 1·py to 1·DMF), however, do not occur. Structural Studies of the Cu(analyte)n[AuX2(CN)2]2 Adducts. To gain further insight into how the various analytes are structurally incorporated, several of their crystal structures were determined using either single crystal or powder X-ray diffraction techniques, as described below. In general, the single-crystal structure is the same as that adopted by the powder product generated by vapor adsorption of the analyte by 1 or 2, except for the pyridine adducts, as confirmed by powder X-ray diffraction and IR and Raman spectra. Analyte = H2O. The crystal structure of 1·H2O reveals a 1-D coordination polymer of octahedral Cu(II) centers with two equatorial and two axially bound H2O units, and equatorially bridging [AuCl2(CN)2]− building blocks as shown in Figure 3. Unbound [AuCl2(CN)2]− units situated between the chains

581 ± 5 539 ± 5 529 ± 3 598 ± 3

On the other hand, benzene, toluene, CH2Cl2, ether, dimethoxyethane, tetrahydrofuran, CH3NO2, MeOH, or EtOH exhibited no effect on either 1 or 2 as determined by IR and Raman spectroscopy. Of particular note, MeCN also has no effect on 1 or 2, which is surprising given that MeCN molecules readily bind to M(II) centers in [AuCl2(CN)2]− and [AuBr2(CN)2]−-containing systems,76,78 and exposure of the Au(I)-based Cu(OH2)2[Au(CN)2]2 to MeCN does cause a color change.25 In general, only VOCs containing donors conducive to binding to the Cu(II) center are taken up: this induces a change in the ligand field splitting and a consequent color change;25 analytes without appropriate donor groups had no effect on 1 and 2. In addition to their distinct colors, the change in the environment of the Cu(II) centers upon VOC binding also elicits a change in the π-backbonding from the Cu(II) center to the bound cyano group and potentially changes the symmetry of the Cu(II) center, thereby directly affecting the energy and number of stretches observed in the IR and Raman spectra. As a result, the VOC adducts of 1 and 2 display signature patterns in the vCN region in both the IR and the Raman spectra (Tables 4 and 5). A characteristic of cyanoaurate(III) units is the low intensity of vCN peaks in the IR,79,80 which greatly decreases the sensitivity of changes in the IR spectrum to VOC binding. On the other hand, the vCN stretches in the Raman spectra show significantly stronger intensities, and are more well resolved than their IR counterparts; therefore, these can better serve as a readout for sensory purposes. Thermal Stability and Reversibility. To determine the stability of the parent Cu[AuX2(CN)2]2 compounds and how F

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

[AuCl2(CN)2]− units (Supporting Information Figure S6). Surrounding the Cu(II) centers are four equatorially bound DMF units. Between the chains in 1·DMF are unbound [AuCl2(CN)2]− units (Figure 4a) with interchain Au···N

Figure 3. Crystal structure of the extended network of 1·H2O (hydrogen atoms removed for clarity) showing the hydrogen bonding (red dashes) and Cl···Cl interactions (green dashes) between the chains and unbound [AuCl2(CN)2]− units. Selected bond lengths (Å): Cu(1)−O(1) 1.942(2); Cu(1)−O(2) 2.400(2); Cu(1)−N(1) 1.987(2); Au(1)−C(1) 1.988(3); Au(1)−Cl(1) 2.2839(7); Au(2)− C(2) 1.997(3); Au(2)−Cl(2) 2.2827(9); Cl(1)−Cl(2) 3.396(1). Au, gold; C, gray; Cl, light green; Cu, dark green; N, blue; O, red. Figure 4. Extended 2-D network of (a) 1·DMF and (b) 2·DMF (hydrogen atoms and carbon and nitrogen atoms belonging to the DMF moieties removed for clarity). Au···N interactions shown as dashed blue lines, Au···Br interactions shown as dashed red lines. Au, gold; Br, scarlet; C, gray; Cl, light green; Cu, dark green; N, blue; O, red.

hold them together via an extensive network of hydrogen bonds and a rare Cl···Cl interaction of 3.396(1) Å (cf., sum of van der Waals radii of 3.51 Å for two Cl atoms)82,83 to yield a 2-D sheet superstructure. Few structures with Cl···Cl interactions have been reported, although mixed halogen interactions such as Cl···Br and Cl···I are more common.84,85 The previously reported Co(OH2)4[AuCl2(CN)2]2·2H2O76 is isostructural to 1·H2O. In both cases, an elongation in two trans-axial M−O bonds is observed, but is much more pronounced in this d9 Cu(II) system. Analyte = DMF. Crystals of 1·DMF and 2·DMF can be obtained by slow evaporation of a DMF solution of 1 and 2, respectively. The structures of 1·DMF and 2·DMF are very similar to each other and also to that of 1·H2O; selected bond lengths and angles are listed in Table 6. Thus, 1·DMF also consists of 1-D chains built from Cu(II) centers bridged by

distances of 3.283(4) Å (cf., sum of van der Waals radii of 3.21 Å).82,83 The 1-D chains in 2·DMF are also interspaced by [AuBr2(CN)2]− units (Figure 4b), which weakly link chains together via Au···Br interactions of 3.5218(4) Å (cf., sum of van der Waals radii of 3.56 Å).82,83 In contrast to the Cu−O bonds in 1·H2O, it is the Cu−N (cyano) bonds in 1·DMF and 2·DMF that are Jahn−Teller elongated (Table 6). Also, in 1·H2O the 1-D chains are essentially linear, whereas the chains in 1·DMF and 2·DMF are wavy, likely to accommodate the extra bulk added by the DMF molecules. Analyte = DMSO. Slow decomposition of [AuCl2(CN)2]− to AuCN in DMSO solution hindered access to crystals of 1·DMSO. However, crystals of 2·DMSO could be obtained from the slow evaporation of a DMSO solution of 2. 2·DMSO (shown in Figure 5) exhibits a crystal structure very similar to that of 2·DMF: a 1-D chain of Cu[AuBr2(CN)2]+ units with four equatorially bound DMSO units, and unbound

Table 6. Selected Bond Lengths in 1·DMF, 2·DMF, 1· DMSO, and 2·DMSOa distance (Å) bond

1·DMF

2·DMF

1· DMSOb

Cu(1)−O(1) Cu(1)−O(2) Cu(1)−N(1) Au(1)···N(2) Au(1)···X(2)c Au(2)···X(1)*c

1.995(2) 1.964(2) 2.367(3) 3.283(4)

1.980(2) 1.966(2) 2.414(3)

1.97(2) 2.07(2) 2.85(3)

1.977(7) 1.961(6) 2.650(8)

3.5218(4) 3.7124(3)

3.38(1) 3.75(2)

3.3906(13) 3.6803(15)

3.5965(8)

2·DMSO

Symmetry element. *: x, y + 1, z for 1·DMF and 2·DMF, x − 1, y, z for 2·DMSO. bBond lengths based on Rietveld refinements of powder pattern data. Because of refinement conditions, bond lengths related to C and N atoms are approximate. cX = Cl for 1·DMF and 1·DMSO, X = Br for 2·DMF and 2·DMSO. a

Figure 5. 1-D chain of 2·DMSO (hydrogen atoms removed for clarity). Au, gold; Br, scarlet; C, gray; Cu, dark green; N, blue; O, red; S, yellow. G

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials Table 7. Selected Bond Lengths in 2·glycola

[AuBr2(CN)2]− units between the chains. The Cu−N(cyano) distance of 2.659(8) Å is substantially longer in comparison with the Cu−N bond distances of 2.367(3)−2.414(3) Å seen in the DMF adducts of 1 and 2, respectively (selected bond lengths and angles are in Table 6). The powder X-ray diffractogram of 1·DMSO is very similar to the pattern generated from the crystal structure for 2·DMSO, and a similar unit cell is obtained. Using the atomic parameters of 2·DMSO as a starting point (and replacing the Br atoms with Cl atoms), Rietveld refinement yields a structure for 1·DMSO with the same motif as 1·DMSO (see Supporting Information Figure S7 and Table 3 for PXRD data). Analyte = Ethylene Glycol. Upon crystallization of 2 from a 5% v/v ethylene glycol/H2O solution, a molecular strucuture different from the 1-D chain structures obtained with H2O, DMF, and DMSO results, although the geometry about the Cu(II) (Figure 6) center in 2·glycol is analogous to the

bond

distance (Å)

Cu(1)−O(1) Cu(1)−O(3) Cu(1)−N(1) O(1)−H(1)···O(2)* O(2)−H(2)···N(2)† O(3)−H(3)···O(4)‡ O(4)−H(4)···O(2)§

2.290(2) 2.024(2) 1.969(2) 2.809(4) 2.768(4) 2.622(4) 2.800(4)

Symmetry elements. *: −x + 3/2, y − 1/2, −z + 1/2. †: x + 1, y, z. ‡: −x + 1, −y + 1, −z. §: −x + 3/2, y + 1/2, −z + 1/2. a

There are only a few reported structures of transition-metal ethylene glycol complexes, which demonstrate a variety of coordination modes, the most common being bidentate chelating86,87 or bridging between88,89 two metal centers; there are only a handful of complexes with monodentate ethylene glycol units90,91 as observed in 2·glycol. Crystals of 1·glycol could not be obtained, and the PXRD pattern thereof did not resemble that simulated for 2·glycol, indicating that a different (unknown) structure is generated in this case. Analyte = Dioxane. In contrast to the structures of the oxygen donor adducts above, when 1 is exposed to dioxane vapor, a very different structural motif is generated. The 2-D structure of 1·dioxane (crystals were obtained by layering 1,4dioxane over an H2O solution of 1) contains nodes of H2Obridged [Cu2(OH2)4]4+ moieties as shown in Figure 8. These

Figure 6. Crystal structure of 2·glycol (hydrogen atoms removed for clarity). Au, gold; Br, scarlet; C, gray; Cu, dark green; N, blue; O, red.

aforementioned adducts. Rather than chelate, the glycol remains monodentate, and structurally in place of a repeating coordination polymer geometry is a mononuclear complex connected by a very extensive hydrogen-bonding network facilitated by both the bound and the unbound ends of the glycol units (Figure 7 and Table 7). This structure clearly demonstrates that the difference in energy between an extended network supported by a coordination polymer framework and an extensive network of weak interactions such as hydrogen bonds is very low. Figure 8. A section of the crystal structure of 1 dioxane showing coordination about the Cu(II) centers (hydrogen atoms and dioxane molecules removed for clarity). Au, gold; C, gray; Cl, light green; Cu, dark green; N, blue; O, red.

moieties are then linked in two dimensions by [AuCl2(CN)2]− units, resulting in the 2-D framework shown in Figure 9a. The close proximity of the H2O-bridged Cu(II) centers results in a similar set of “bridged” Au(III) centers through Au···Cl interactions of 3.222(2) Å. Similar Au···Cl interactions of 3.167(2), 3.191(2), and 3.452(2) Å and Au···N interactions of 2.958(7) Å support unbound [AuCl2(CN)2]− units within the square pockets of this 2-D framework (Figure 9(a)).69,75,76,92,93 Uncoordinated dioxane moieties are present between these sheets, and are hydrogen bonded to H2O units, creating the overall 3-D network shown in Figure 9b. In each of the cases presented thus far, the vapochromic response of each compound was attributable to the

Figure 7. Extended hydrogen-bonding network of 2·glycol (hydrogen atoms and [AuBr2(CN)2]− units removed for clarity). Au, gold; C, gray; Cu, dark green; O, red. H

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 9. Extended network of 1 dioxane (hydrogen atoms removed for clarity) showing (a) 2-D sheets with interactions to unbound [AuCl2(CN)2]− units (dioxane molecules removed for clarity) and (b) intersheet hydrogen bonding through dioxane moieties (unbound [AuCl2(CN)2]− units removed for clarity). Au···N interactions shown as dashed blue lines, Au···Cl interactions shown as dashed green lines, hydrogen bonds shown as dashed red lines. Au, gold; C, gray; Cl, light green; Cu, dark green; N, blue; O, red.

coordination of the analyte to the Cu(II) center. Here, however, it is the H2O units that are coordinated to the Cu(II) centers rather than the dioxane molecules themselves, resulting in a color for 1·dioxane that closely resembles that of 1·H2O. From PXRD data, 2·dioxane adopts a structure different from that of 1·dioxane. As an accurate elemental analysis could not be obtained (the compound desolvates quickly), it is not possible to determine the number of dioxane units present in 2·dioxane. Analyte = py: Crystal Structure of [Cu(py)4 ] 2 [Cu(OH2)2(py)2][Au(CN)2]6·4H2O (3). Exposure of solid 1 or 2 to pyridine vapor generates Cu(py)x[AuCl2(CN)2]2 (1·py; x > 3) and Cu(py)x[AuBr2(CN)2]2 (2·py; x > 3), as identified by a combination of elemental analysis, TGA, and Raman and IR spectroscopies. In particular, the vAuX stretches in the Raman spectra are present, confirming the presence of the [AuX2(CN)2]− moieties. Attempts to crystallize these pyridine adducts of 1 and 2 were made by dissolving the parent compounds in 5% v/v pyridine in H2O, resulting in a deep blue solution in both cases. After a few days, identical blue needle-shaped crystals appeared in both solutions. Single-crystal X-ray diffraction analysis of both sets of crystals revealed that the same product was formed regardless of starting material: Au(I)-containing [Cu(py)4]2[Cu(OH2)2(py)2][Au(CN)2]6·2H2O (3). In other words, the presence of pyridine in these solutions triggered the reductive elimination of halogen from the Au(III) centers, resulting in the formation of Au(I)-based materials. Thermal treatment of the Au(III) building blocks in solution also induces this unusual reaction.78 Compound 3 consists of two different Cu(II) environments arranged in a linear 1-D chain (Figure 10a). Two thirds of the Cu(II) centers (A) are bound by four equatorial pyridine units, with bridging [Au(CN)2]− units in the axial positions. The remaining Cu(II) centers (B) have two units and two H2O pyridine units bound equatorially. The overall 1-D chains adhere to the AABAAB-type pattern shown in Figure 10a. In addition, the A-site Cu octahedra exhibit a Jahn−Teller distortion parallel to the 1-D chains, whereas that of the Bsite Cu octahedra is oriented perpendicular to the chains. This

Figure 10. Crystal structure of 3 (hydrogen atoms removed for clarity) showing (a) 1-D chains of its two different Cu(II) environments in an AAB pattern and (b) its extended network built upon Au···Au interactions (pyridine units removed for clarity). Hydrogen bonds shown as dashed red lines. Au, gold; C, gray; Cu, dark green; N, blue; O, red.

alternating axial−equatorial orientation of Jahn−Teller axes is a common phenomenon among Cu(II)-containing networks.94 Shown in Figure 10b is the packing arrangement of the 1-D chains. Free [Au(CN)2]− units are bound to the bridging units via Au···Au interactions of 3.2755(1) and 3.2824(1) Å along the c-axis (Table 8). As the Au···Au interactions propagate along the c-axis, the 1-D chains form a left-handed spiral, resulting in an overall chiral structure (space group P6222). The spiral is built of alternating layers of the aforementioned 1-D chains as well as chains of unbound [Au(CN)2]− units I

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials Table 8. Selected Bond Lengths in 3a bond

distance (Å)

Cu(1)−N(1) Cu(1)−N(3) Cu(1)−N(1) Cu(1)−N(12) Cu(2)−N(2) Cu(2)−N(21) Cu(2)−O(1) Au(1)···Au(3) Au(2)···Au(4) O(2)−H(2B)···N(4)* O(2)−H(2A)···N(6)†

2.313(9) 2.393(9) 2.049(6) 2.039(8) 1.988(8) 2.042(9) 2.342(8) 3.2824(1) 3.2755(1) 2.957(9) 2.768(4)

Symmetry elements. *: −x + y + 1, −x + 1, z + 1/3. †: x − y + 1, x, z + 1/3. a

Figure 11. UV−visible reflectance spectra of 1 and its analytecontaining analogues.

connected together via hydrogen-bonding interactions with the unbound H2O units of 2.768(4) and 2.957(9) Å. Complex 3 desolvates rapidly upon removal of the mother liquor; the IR spectrum of the desolvate matches that of the previously reported Au(I)-based material Cu(py) 2 [Au(CN)2]2,25 indicating loss of all H2O molecules and four equivalents of pyridine molecules. Structural Impact of Analyte Molecules. In the analytecontaining structures described above, the analyte molecules are almost always incorporated into the structure through coordination to the Cu(II) center. In the case of H2O, DMF, DMSO, and ethylene glycol, a similar coordination environment is seen around the Cu(II) center: there are four meridional O-bound analyte units and two bridging (pendant in the case of 2·glycol) N-bound [AuX2(CN)2]− units occupying the remaining trans sites. Even in the case of dioxane adsorption, although the dioxane molecules themselves do not coordinate, additional ambient H2O molecules do. The ability of a coordination polymer material such as 1 and 2 to pick up these analyte molecules despite the lack of significant porosity has been attributed in part to the flexibility of the framework,95,96 especially around the Cu(II) coordination sphere (assisted by the Jahn−Teller distortion, which weakens the axial bonds); this was also observed in the 2-D sheet structure of the analogous Cu(OH2)2[Au(CN)2]2.25 In that case, when exposed to VOC molecules, the Cu[Au(CN)2]2 2-D sheets corrugate or flatten to facilitate the addition of the VOC molecules to the Cu(II) coordination sphere; the overall dimensionality, however, remained the same in most cases. In the Cu[AuX2(CN)2]2 analogues 1 and 2, the network becomes 1-D in almost all cases, and even 0-D (molecular) in the case of 2·glycol. This is most likely due to the fact that [AuX2(CN)2]− building blocks are much less Lewis-basic75,76,92,93 than Au(I)based [Au(CN)2]−; the much more labile Cu−N(cyano) bond can easily be severed to open more coordination sites for VOC molecules. This feature of Au(III)-based coordination polymers may have advantages in analyte uptake; however, they also tend to have a lower chemical stability, as exemplified by the formation of 3. Sensor Properties. It is clear that both 1 and 2 exhibit a significant change in physical appearance, Raman, and IR spectra when exposed to a variety of VOCs. One obvious method for detecting the change in the material is visible spectroscopy (or visual observation) due to the variety of colors exhibited. Figure 11 shows the overlaid solid-state reflectance spectra of 1 and its VOC adducts. Although the signals are very

strong (theoretically leading to very good sensitivity), they are also broad, resulting in a significant overlap of all of the spectra (including the parent compound); this reduces overall sensitivity, although complete conversion of the sample (i.e., in a sufficiently high concentration of analyte, or as a dosimeter) is readily observable. On the other hand, the vibrational (Raman and IR) spectra of these compounds (Tables 4 and 5) show distinct vCN stretches with much less overlap for different analytes. Unlike cyanoaurate(I) compounds, cyanoaurate(III) compounds such as those reported here generally have weak vCN stretches in their IR spectra,75,77−79 which can lead to a very low sensitivity when used as a sensor readout. This is, however, not the case in the Raman spectra.63,77 A representative comparison of the vCN peaks of 2·DMSO is shown in Figure 12: the signal-to-noise

Figure 12. A comparison of the vCN stretches of 2 DMSO in the infrared (top) and Raman (bottom) spectra.

ratio of the vCN stretch is much higher in the Raman spectrum than the IR spectrum, and thus these coordination polymers can effectively act as “antennae” for the VOC analytes, transducing the chemical change resulting from forming an adduct to a detectable Raman signal. The distinct, well-resolved patterns of vCN peaks belonging to each individual adduct J

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 13. Chart showing the ability of (a) 1 and (b) 2 to differentiate DMF from DMSO, pyridine, and the parent material by monitoring the Raman vCN stretches at 2205 or 2197 cm−1, respectively.

Figure 14. A representation of the detected signal patterns unique to each of the analytes when multiple Raman νCN signals are monitored simultaneously. Green squares represent detection events, or a “yes” response; red squares represent the absence of a detection event, or a “no” response.

DMF Detection. The standardized Raman spectra for 1, 2, 1·DMF, 2·DMF, 1·DMSO, 2·DMSO, 1·py, and 2·py were collected as described above. The Raman intensities in these spectra were then integrated over a 10 cm−1 range (the approximate width of a given vCN peak at half height) centered on the vCN peak corresponding to signal A for 1·DMF or 2·DMF (2205 and 2197 cm−1, respectively). These integrated values represent the signal resulting from exposure to DMF, DMSO, or py while monitoring 1 at 2205 ± 5 cm−1 (signal A for 1·DMF) or 2 at 2197 ± 5 cm−1 (signal A for 2·DMF), the results of which are summarized in Figure 13. Obviously, exposure of 1 or 2 to DMF vapor results in an intense signal at 2205 or 2197 cm−1, respectively. In contrast, the DMSO or py adducts and the parent compounds have low intensity in this region, as indicated by the bars in Figure 13. Thus, when specifically monitoring for the presence of 1·DMF or 2·DMF in 1 and 2, a detection event is only triggered upon exposure to DMF (i.e., none of the other VOCs give a false positive for the presence of DMF). Therefore, identification of DMF as the analyte bound to 1 or 2 is trivial. As mentioned previously, by monitoring the Raman spectrum of 1 at 2205 cm−1, concentrations as low as 500 ppm of DMF can be detected. DMSO and py Detection. The same procedure was repeated, monitoring the appropriate vCN signals of 1·DMSO,

should in addition allow for accurate identification of the detected VOC, as shown below. Analyte Identification. The ability to uniquely identify a detected analyte when using 1 and 2 as Raman-detected sensors was probed by focusing on the DMF, DMSO, and pyridine analytes. The intensity of the vCN peaks in the Raman spectra was compared to that of an external standard, K2[Zn(CN)4], under the same experimental conditions. Using the ratio of the intensity of the sample vCN peak to that of the external standard allows for a direct comparison of signal strengths between the materials studied. The DMF, DMSO, and py adducts of 1 and 2 exhibit two Raman vCN peaks (see Tables 4 and 5): a higher energy peak above 2200 cm−1 (signal A; this stretch is absent for 1·py), and a lower energy peak at approximately 2185 cm−1 (signal B). Because the spread of signal B is very tight among these adducts (i.e., they are all within a few wavenumbers of each other), this peak is not useful in identifying which analyte caused the sensory response. The spread in signal A, however, is much higher, with vCN peaks approximately 10 cm−1 or more apart within a given family (i.e., 1 or 2 and their respective adducts); thus signal A of each adduct was used to test the ability of 1 and 2 to distinguish between bound DMF, DMSO, and pyridine. K

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

for the detection and identification of VOCs. The concept of harnessing Raman bands of these materials in this fashion could readily be generalized to a wide range of other coordination polymers and MOF materials. Further studies in using this methodology in conjunction with SERS to further amplify the detection signal are also of interest.

2·DMSO, 1·py, and 2·py. The resulting charts are shown in Supporting Information Figure S8. Although, in the case of 1, signals from DMF and py adducts do not interfere with DMSO detection signals, the proximity of its intense vCN stretch (2243 cm−1) to that of the equally intense parent compound (2259 cm−1) significantly reduces its sensitivity. In the case of 2, exposure to py vapor also results in an intense signal at the same place when monitoring for DMSO. Because 1·py only has a signal B at 2185 cm−1, which is also present in the Raman spectrum of all other analytes, py vapor cannot be identified apart from the other VOCs in this manner using 1; using 2 gives a result similar to that of 2·DMSO, above. Thus, 1 and 2 cannot uniquely identify DMSO or py analytes by monitoring a single Raman peak position. However, this limitation can be overcome by monitoring several vCN positions in either 1 or 2 simultaneously. Because each VOC adduct displays a characteristic vCN pattern, unambiguous differentiation between these VOCs becomes straightforward. To achieve this, the aforementioned intensity data can be simplified to yield a simple “yes” or “no” response. For example, exposure of 1 to DMF yields a “yes” response when monitoring at 2205 cm−1 (the peak for 1·DMF; see Figure 13), but yields a “no” response when monitoring, for example, 2243 cm−1 (the vCN peak for 1·DMSO). Using this analysis for all Raman frequencies results in the response charts shown in Figure 14, with green squares representing “yes” and red squares representing “no” responses. Using this straightforward data analysis, a clear, distinct yes/ no pattern is observed when 1 is exposed to DMF, DMSO, or py; each is also distinct from the pattern given by 1 itself. Thus, it is possible to uniquely identify the VOC that is triggering a detection event using 1. Once identified, only a single channel (preferably one that is well separated from the vCN peak corresponding to the parent compound, e.g., 2185−1) is required for quantification, as described above. In the case of 2, because exposure to DMSO or py results in the same yes/no pattern, only DMF is still uniquely identifiable via this methodology. Thus, by this measure, Cu[AuCl2(CN)2]2 can uniquely distinguish between a wider range of analytes than 2.



ASSOCIATED CONTENT

S Supporting Information *

TGA data for 1·DMF, 2·DMF, 1·DMSO, 2·DMSO, 1·py, 2·py, and 1·H2O, results of analyte identification experiments for DMSO and pyridine, powder X-ray diffractograms for 1, 2, and and X-ray crystallographic data (CIFs) for all structures. This material is available free of charge via the Internet at http:// pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: dleznoff@sfu.ca. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS D.B.L. is grateful to NSERC of Canada for support of this research; J.S.O. thanks NSERC for a PGS-D graduate scholarship.



REFERENCES

(1) Clemente-León, M.; Coronado, E.; Martí-Gastaldo, C.; Romero, F. M. Chem. Soc. Rev. 2011, 40, 473−497. (2) Batten, S. R.; Murray, K. S. Coord. Chem. Rev. 2003, 246, 103− 130. (3) Lefebvre, J.; Callaghan, F.; Katz, M. J.; Sonier, J. E.; Leznoff, D. B. Chem.Eur. J. 2006, 12, 6748−61. (4) Dul, M.-C.; Pardo, E.; Lescouëzec, R.; Journaux, Y.; FerrandoSoria, J.; Ruiz-Garca, R.; Cano, J.; Julve, M.; Lloret, F.; Cangussu, D.; Pereira, C. L.; Stumpf, H. O.; Pasán, J.; Ruiz-Pérez, C. Coord. Chem. Rev. 2010, 254, 2281−2296. (5) Wriedt, M.; Yakovenko, A. a.; Halder, G. J.; Prosvirin, A. V.; Dunbar, K. R.; Zhou, H.-C. J. Am. Chem. Soc. 2013, 135, 4040−50. (6) Maxim, C.; Branzea, D. G.; Tiseanu, C.; Rouzières, M.; Clérac, R.; Andruh, M.; Avarvari, N. Inorg. Chem. 2014, 53, 2708−17. (7) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. Science 2013, 341, 1230444. (8) Song, L.; et al. Energy Environ. Sci. 2012, 5, 7508. (9) Kitagawa, S.; Uemura, K. Chem. Soc. Rev. 2005, 34, 109−19. (10) Zhang, J.-P.; Liao, P.-Q.; Zhou, H.-L.; Lin, R.-B.; Chen, X.-M. Chem. Soc. Rev. 2014. (11) Herm, Z. R.; Wiers, B. M.; Mason, J. A.; van Baten, J. M.; Hudson, M. R.; Zajdel, P.; Brown, C. M.; Masciocchi, N.; Krishna, R.; Long, J. R. Science 2013, 340, 960−4. (12) Taylor, J. M.; Dawson, K. W.; Shimizu, G. K. H. J. Am. Chem. Soc. 2013, 135, 1193−6. (13) Cui, Y.; Yue, Y.; Qian, G.; Chen, B. Chem. Rev. 2012, 112, 1126−62. (14) Heine, J.; Müller-Buschbaum, K. Chem. Soc. Rev. 2013, 42, 9232−42. (15) Rocha, J. a.; Carlos, L. D.; Paz, F. a. A.; Ananias, D. Chem. Soc. Rev. 2011, 40, 926−40. (16) Katz, M.; Ramnial, T.; Yu, H.; Leznoff, D. J. Am. Chem. Soc. 2008, 130, 10662. (17) Roberts, R. J.; Li, X.; Lacey, T. F.; Pan, Z.; Patterson, H. H.; Leznoff, D. B. Dalton Trans. 2012, 41, 6992−7. (18) Fernández, E. J.; Laguna, A.; López-de Luzuriaga, J. M. Dalton Trans. 2007, 1969−81.



CONCLUSIONS The two new coordination polymers Cu[AuCl2(CN)2]2 (1) and Cu[AuBr2(CN)2]2 (2) have demonstrated their utility as sensory materials toward several VOCs including DMF, DMSO, pyridine, dioxane, and ethylene glycol. The flexible nature of the Cu(II) coordination sphere and the weak Lewisbasicity of the [AuX2(CN)2]− (X = Cl, Br) building block allows for a dissociation of some of the Cu−N(cyano) bonds and the concomitant coordination of the VOC molecules, resulting in a decrease from a 2-D or 3-D network to a 1-D chain in most cases, and even a 0-D molecule with ethylene glycol. The change in the coordination environment around the Cu(II) center results in a striking visible color change; thus 1 and 2 have potential as visual alert indicators. In addition, the cyano moieties present in these materials give an alternative means of detecting the chemical change: by monitoring a variety of vCN peaks via Raman spectroscopy, the identity of the detected VOC could be uniquely identified by the pattern of vCN shifts corresponding to the adduct. Indeed, using the shift of the strong Raman vCN peaks to amplify a signal generated by the presence of a VOC clearly shows promise as a methodology L

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials (19) Beauvais, L.; Shores, M.; Long, J. J. Am. Chem. Soc. 2000, 8, 2763−2772. (20) Kreno, L. E.; Leong, K.; Farha, O. K.; Allendorf, M.; Van Duyne, R. P.; Hupp, J. T. Chem. Rev. 2012, 112, 1105−25. (21) Lim, S. H.; Olmstead, M. M.; Balch, A. L. J. Am. Chem. Soc. 2011, 133, 10229−38. (22) Lim, M. J.; Murray, C. A.; Tronic, T. A.; DeKrafft, K. E.; Ley, A. N.; DeButts, J. C.; Pike, R. D.; Lu, H.; Patterson, H. H. Inorg. Chem. 2008, 47, 6931−47. (23) Ley, A. N.; Dunaway, L. E.; Brewster, T. P.; Dembo, M. D.; Harris, T. D.; Baril-Robert, F.; Li, X.; Patterson, H. H.; Pike, R. D. Chem. Commun. 2010, 46, 4565−7. (24) Lefebvre, J.; Korčok, J. L.; Katz, M. J.; Leznoff, D. B. Sensors 2012, 12, 3669−92. (25) Lefebvre, J.; Batchelor, R. J.; Leznoff, D. B. J. Am. Chem. Soc. 2004, 126, 16117−25. (26) Guillerm, V.; Kim, D.; Eubank, J. F.; Luebke, R.; Liu, X.; Adil, K.; Lah, M. S.; Eddaoudi, M. Chem. Soc. Rev. 2014. (27) Stock, N.; Biswas, S. Chem. Rev. 2012, 112, 933−69. (28) Burchell, T. J.; Puddephatt, R. J. Inorg. Chem. 2005, 44, 3718− 30. (29) Janiak, C. Dalton Trans. 2003, 2781. (30) Hao, Z.; Yang, G.; Song, X.; Zhu, M.; Meng, X.; Zhao, S.; Song, S.; Zhang, H. J. Mater. Chem. A 2014, 2, 237. (31) Li, Y.; Zhang, S.; Song, D. Angew. Chem., Int. Ed. 2013, 52, 710− 3. (32) Pramanik, S.; Hu, Z.; Zhang, X.; Zheng, C.; Kelly, S.; Li, J. Chem.Eur. J. 2013, 19, 15964−71. (33) Tian, D.; Li, Y.; Chen, R.-Y.; Chang, Z.; Wang, G.-Y.; Bu, X.-H. J. Mater. Chem. A 2014, 2, 1465. (34) Lan, A.; Li, K.; Wu, H.; Olson, D. H.; Emge, T. J.; Ki, W.; Hong, M.; Li, J. Angew. Chem., Int. Ed. 2009, 48, 2334−8. (35) Fleischer, M., Lehmann, M., Eds. Solid State Gas Sensors Industrial Application; Springer: New York, 2012. (36) Gründler, P. Chemical Sensors: An Introduction for Scientists and Engineers; Springer: Berlin, Germany, 2007. (37) Yamazoe, N. Sens. Actuators, B: Chem. 2005, 108, 2−14. (38) Lu, Z.-Z.; Zhang, R.; Li, Y.-Z.; Guo, Z.-J.; Zheng, H.-G. J. Am. Chem. Soc. 2011, 133, 4172−4. (39) Bureekaew, S.; Sato, H.; Matsuda, R.; Kubota, Y.; Hirose, R.; Kim, J.; Kato, K.; Takata, M.; Kitagawa, S. Angew. Chem. 2010, 122, 7826−7830. (40) Wenger, O. S. Chem. Rev. 2013, 113, 3686−733. (41) Bariáin, C.; Matas, I. R.; Romeo, I.; Garrido, J.; Laguna, M. Appl. Phys. Lett. 2000, 77, 2274. (42) Mansour, M. A.; Connick, W. B.; Lachicotte, R. J.; Gysling, H. J.; Eisenberg, R. J. Am. Chem. Soc. 1998, 120, 1329−1330. (43) Fung, E. Y.; Olmstead, M. M.; Vickery, J. C.; Balch, A. L. Coord. Chem. Rev. 1998, 171, 151−159. (44) Lee, Y.-A.; Eisenberg, R. J. Am. Chem. Soc. 2003, 125, 7778−9. (45) Vickery, J. C.; Olmstead, M. M.; Fung, E. Y.; Balch, A. L. Angew. Chem., Int. Ed. Engl. 1997, 36, 1179−1181. (46) Elosúa, C.; Bariáin, C.; Matas, I. R.; Arregui, F. J.; Luquin, A.; Laguna, M. Sens. Actuators, B 2006, 115, 444−449. (47) Elosúa, C.; Bariain, C.; Matas, I. R.; Arregui, F. J.; Vergara, E.; Laguna, M. Sens. Actuators, B 2009, 137, 139−146. (48) Elosúa, C.; Bariáin, C.; Luquin, A.; Laguna, M.; Matas, I. R. Sens. Actuators, B 2011, 157, 388−394. (49) Elosúa, C.; Arregui, F.; Zamarreño, C.; Bariáin, C.; Luquin, A.; Laguna, M.; Matas, I. R. Sens. Actuators, B 2012, 173, 523−529. (50) Elosúa, C.; Vidondo, I.; Arregui, F.; Bariain, C.; Luquin, A.; Laguna, M.; Matas, I. Sens. Actuators, B 2013, 187, 65−71. (51) Elosúa, C.; Bariáin, C.; Luquin, A.; Laguna, M.; Matas, I. R. IEEE Sens. J. 2012, 12, 3156−3162. (52) Terrones, S. C.; Aguado, C. E.; Bariáin, C.; Carretero, A. S.; Maestro, I. R. M.; Gutiérrez, A. F.; Luquin, A.; Garrido, J.; Laguna, M. Opt. Eng. 2006, 45, 044401. (53) Bariáin, C.; Matas, I. R.; Romeo, I.; Garrido, J.; Laguna, M. Sens. Actuators, B 2001, 76, 25−31.

(54) Bezunartea, M.; Estella, J.; Echeverra, J.; Elosúa, C.; Bariáin, C.; Laguna, M.; Luquin, A.; Garrido, J. Sens. Actuators, B 2008, 134, 966− 973. (55) Elosúa, C.; Bariáin, C.; Matas, I. R.; Arregui, F. J.; Luquin, A.; Laguna, M. 2009 IEEE Sensors 2009, 154−157. (56) Fernández, E. J.; Laguna, A.; López-de Luzuriaga, J. M.; Montiel, M.; Olmos, M. E.; Pérez, J. Organometallics 2006, 25, 1689−1695. (57) Fernández, E. J.; López-De-Luzuriaga, J. M.; Monge, M.; Olmos, M. E.; Pérez, J.; Laguna, A.; Mohamed, A. A.; Fackler, J. P. J. Am. Chem. Soc. 2003, 125, 2022−3. (58) Laguna, A.; Lasanta, T.; López-de Luzuriaga, J. M.; Monge, M.; Naumov, P.; Olmos, M. E. J. Am. Chem. Soc. 2010, 132, 456−7. (59) Strasser, C. E.; Catalano, V. J. J. Am. Chem. Soc. 2010, 132, 10009−11. (60) Osawa, M.; Kawata, I.; Igawa, S.; Hoshino, M.; Fukunaga, T.; Hashizume, D. Chem.Eur. J. 2010, 16, 12114−26. (61) King, C.; Wang, J. C.; Khan, M. N. I.; Fackler, J. P. Inorg. Chem. 1989, 28, 2145−2149. (62) Ito, H.; Saito, T.; Oshima, N.; Kitamura, N.; Ishizaka, S.; Hinatsu, Y.; Wakeshima, M.; Kato, M.; Tsuge, K.; Sawamura, M. J. Am. Chem. Soc. 2008, 130, 10044−5. (63) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordination Compounds; John Wiley & Sons, Inc.: Toronto, 1997. (64) Qazi, H. H.; bin Mohammad, A. B.; Akram, M. Sensors 2012, 12, 16522−56. (65) Passaro, V. M. N.; de Tullio, C.; Troia, B.; La Notte, M.; Giannoccaro, G.; De Leonardis, F. Sensors 2012, 12, 15558−98. (66) Capitán-Vallvey, L. F.; Palma, A. J. Anal. Chim. Acta 2011, 696, 27−46. (67) Lide, D. R., Ed. CRC Handbook of Chemistry and Physics, 71st ed.; CRC Press, Inc.: Boca Raton, FL, 1990. (68) Shaw, C. F. Chem. Rev. 1999, 99, 2589−2600. (69) Ovens, J. S.; Truong, K. N.; Leznoff, D. B. Dalton Trans. 2012, 41, 1345−51. (70) Hübschle, C. B.; Sheldrick, G. M.; Dittrich, B. J. Appl. Crystallogr. 2011, 44, 1281−84. (71) TOPAS-Academic V5. Coelho Software, 2012. (72) Cooper, R. I.; Gould, R. O.; Parsons, S.; Watkin, D. J. J. Appl. Crystallogr. 2002, 35, 168−74. (73) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565. (74) Fen, T. D.; Ringe, D.; Petsko, G. A. J. Appl. Crystallogr. 2003, 36, 944. (75) Ovens, J. S.; Geisheimer, A. R.; Bokov, A. A.; Ye, Z.-G.; Leznoff, D. B. Inorg. Chem. 2010, 49, 9609−16. (76) Ovens, J. S.; Truong, K. N.; Leznoff, D. B. Inorg. Chim. Acta 2013, 403, 127−135. (77) Jones, L. H.; Smith, M. J. J. Phys. Chem. 1964, 41, 2507. (78) Ovens, J. S.; Leznoff, D. B. Dalton Trans. 2011, 40, 4140. (79) Vitoria, P.; Muga, I. n.; Gutiérrez-Zorrilla, J. M.; Luque, A.; Román, P.; Lezama, L.; Zúniga, F. J.; Beitia, J. I. Inorg. Chem. 2003, 42, 960−9. (80) Katz, M. J.; Shorrock, C. J.; Batchelor, R. J.; Leznoff, D. B. Inorg. Chem. 2006, 45, 1757−65. (81) Chomič, J.; Č ernák, J. Thermochim. Acta 1985, 93, 93−96. (82) Bondi, A. J. Phys. Chem. 1964, 68, 441−451. (83) Alvarez, S. Dalton Trans. 2013, 42, 8617−36. (84) Metrangolo, P.; Resnati, G.; Pilati, T.; Biella, S. Struct. Bonding (Berlin) 2008, 126, 105. (85) Navon, O.; Bernstein, J.; Khodorkovsky, V. Angew. Chem., Int. Ed. Engl. 1997, 36, 601. (86) Wang, Z.; Liu, D.-S.; Zhang, H.-H.; Huang, C.-C.; Cao, Y.-N.; Yu, X.-H. J. Coord. Chem. 2008, 61, 419−425. (87) Pico, A. R.; Houk, C. S.; Weakley, T. J.; Page, C. J. Inorg. Chim. Acta 1997, 258, 155−160. (88) Labádi, I.; Burger, K.; Liptay, G.; Czugler, M.; Kálmán, A. J. Therm. Anal. 1986, 31, 1171−1181. (89) Hunger, M.; Limberg, C.; Zsolnai, L. Polyhedron 1999, 17, 3935−3945. M

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials (90) Klüfers, P.; Vogler, C. Z. Anorg. Allg. Chem. 2007, 633, 908− 912. (91) Zhong, K.-L.; Chen, L. Acta Crystallogr., Sect. C 2011, 67, m62− 4. (92) Geisheimer, A. R.; Wren, J. E. C.; Michaelis, V. K.; Kobayashi, M.; Sakai, K.; Kroeker, S.; Leznoff, D. B. Inorg. Chem. 2011, 50, 1265− 74. (93) Katz, M.; Kaluarachchi, H.; Batchelor, R. J.; Schatte, G.; Leznoff, D. B. Cryst. Growth Des. 2007, 7, 1946−1948. (94) Halcrow, M. A. Chem. Soc. Rev. 2013, 42, 1784−95. (95) Atwood, J. L.; Barbour, L. J.; Jerga, A.; Schottel, B. L. Science 2002, 298, 1000−2. (96) Kitagawa, S.; Kitaura, R.; Noro, S.-I. Angew. Chem., Int. Ed. 2004, 43, 2334−75.

N

DOI: 10.1021/cm502998w Chem. Mater. XXXX, XXX, XXX−XXX