Exploiting Pressure To Induce a “Guest-Blocked ... - ACS Publications

Oct 6, 2016 - X-ray Science Division, Advanced Photon Source (APS), Argonne National Laboratory, Lemont, Illinois 60439, United States. §. Institut t...
0 downloads 10 Views 4MB Size
Article pubs.acs.org/IC

Exploiting Pressure To Induce a “Guest-Blocked” Spin Transition in a Framework Material Natasha F. Sciortino,† Florence Ragon,† Katrina A. Zenere,† Peter D. Southon,† Gregory J. Halder,‡ Karena W. Chapman,‡ Lucía Piñeiro-López,§ José A. Real,§ Cameron J. Kepert,† and Suzanne M. Neville*,† †

School of Chemistry, The University of Sydney, Sydney, New South Wales 2006, Australia X-ray Science Division, Advanced Photon Source (APS), Argonne National Laboratory, Lemont, Illinois 60439, United States § Institut the Ciencia Molecular, Universitat de València, 46980 Paterna, València, Spain ‡

S Supporting Information *

ABSTRACT: A new functionalized 1,2,4-triazole ligand, 4-[(E)-2(5-methyl-2-thienyl)vinyl]-1,2,4-triazole (thiome), was prepared to assess the broad applicability of strategically producing multistep spin transitions in two-dimensional Hofmann-type materials of the type [FeIIPd(CN)4(R-1,2,4-trz)2]·nH2O (R-1,2,4-trz = a 4-functionalized 1,2,4-triazole ligand). A variety of structural and magnetic investigations on the resultant framework material [FeIIPd(CN)4(thiome)2]·2H2O (A·2H2O) reveal that a high-spin (HS) to low-spin (LS) transition is inhibited in A·2H2O due to a combination of guest and ligand steric bulk effects. The water molecules can be reversibly removed with retention of the porous host framework and result in the emergence of an abrupt and hysteretic one-step spin transition due to the removal of guest internal pressure. A spin transition can, furthermore, be induced in A·2H2O (0−0.68 GPa) under hydrostatic pressure, as evidenced by variable-pressure structure and magnetic studies, resulting in a two-step spin transition at ambient temperatures at 0.68 GPa. The presence of a two-step spin crossover (SCO) in A·2H2O under hydrostatic pressure compared to a one-step SCO in A at ambient pressure is discussed in terms of the relative ability of each phase to accommodate mixed HS/LS states according to differing lattice flexibilities.



INTRODUCTION

families, examples of room temperature and cooperative spin transitions exist. Alongside the intrinsic lattice cooperativity of the Hofmann family of SCO materials,3a the pillared layered topology with interlayer pore space allows for rational tuning of the SCO properties through the strategic incorporation of a range of potential hydrogen-bonding and π-stacking interaction sites within the framework scaffold.3b,5 These systems are of fundamental interest because their robust nature allows direct magnetostructural rationalization in terms of a simple guest size/shape argument (“internal pressure”) and guest electronic effects. While the majority of Hofmann materials reported are based on pyridyl donor ligands, we showed previously that the incorporation of 1,2,4-triazole-type ligands [bound through N(1)] into two-dimensional (2D) Hofmann-type materials is of interest, particularly because it provides a unique approach for generating layer distortions and multistability.6 For example, [FeIIPd(CN)4(thtrz)2]·(EtOH,H2O) (thtrz = N-thiophenylidene-4H-1,2,4-triazol-4-amine) shows a two-step SCO character with an unusually wide intermediate-spin-state temperature

Spin crossover (SCO) is an electronic switching phenomenon that can occur in d4−7 transition-metal complexes of appropriate intermediate ligand-field strength under external perturbation (T, P, and hν).1 It is well established that the total ligand field imparted on such metal ions is comprised of both inner- and outer- coordination-sphere characteristics (i.e., hydrogen bonding and p−p stacking).2 Effective propagation of spinstate changes in a solid (i.e., cooperativity in SCO materials) is governed primarily by elastic interactions between the switchable metal ions. Increasing strength and cooperativity of elastic interactions between SCO centers has the tendency to sharpen the spin transition from broad to abrupt and lead commonly to the observation of thermal hysteresis. The incorporation of SCO centers into coordination polymers is a strategy anticipated to increase the strength of interactions between metal ions and has been largely successful when short bridges are employed,3 for example, in one-dimensional chain species of the type [FeII(R-trz)3](A)2·nH2O (where R-trz = a 1,2,4triazole-type ligand and A = anion) and the family of Hofmann materials [FeIIMII(CN)4(L)]·n(guest) [where MII = Pt, Pd, and Ni and L = bis(4-pyridyl) bridging ligand].4 In both of these © 2016 American Chemical Society

Received: July 19, 2016 Published: October 6, 2016 10490

DOI: 10.1021/acs.inorgchem.6b01686 Inorg. Chem. 2016, 55, 10490−10498

Article

Inorganic Chemistry Table 1. Single-Crystal X-ray Diffraction Data and Refinement Details A·2H2O temperature/K spin state formula FW/g mol−1 cryst syst space group a/Å b/Å c/Å α/deg β/deg γ/deg V/Å3 ρcalcd/Mg m−3 data/restraints/param R(F) [I > 2σ(I), all] Rw(F2) [I > 2σ(I), all] GOF

200 HS C20H20FeN12O2PdS2 682.82 triclinic P1̅ 7.3030(4) 7.6321(4) 12.5116(7) 104.824(3) 98.914(3) 90.167(3) 665.35(6) 1.704 2724/0/176 0.0738 [0.0798] 0.2232 [0.2332] 1.068

A 375 HS C20H16FeN12PdS2 650.82 triclinic P1̅ 7.3047(11) 7.5681(11) 12.7975(19) 102.893(6) 101.806(6) 90.378(7) 674.04(17) 1.603 2755/0/167 0.0859 [0.1055] 0.2400 [0.2753] 1.066

100 LS C20H16FeN12PdS2 650.82 triclinic P1̅ 6.9882(7) 7.3465(8) 12.6944(14) 104.275(4) 101.253(4) 90.262(4) 618.50(11) 1.518 2528/0/167 0.0619 [0.0742] 0.1656 [0.1801] 1.225

Calcd for C8H8N4S: C, 49.98; H, 4.19; N, 29.15. Found: C, 49.94; H, 4.24; N, 29.21. Complex Synthesis. Crystals were grown by vial-in-vial slow diffusion. Powders of thiome (22.7 mg, 0.118 mmol) and K2[Pd(CN)4] (17.0 mg, 0.0589 mmol) were placed at the base of a small vial. Fe(ClO4)2·6H2O (15.0 mg, 0.0589 mmol) was placed at the base of a large vial. The small vial was placed inside the large vial, and both were slowly filled with a 50:50 EtOH/H2O mixture, being careful to fill the vials without disturbing the reactants. The components were allowed to diffuse over a period of 2 weeks to form yellow crystals. IR (cm−1; A·2H2O): 3381 (br), 2171 (s), 1632 (s), 1591 (s), 1515 (s), 1474 (m), 1459 (s), 1328 (m), 1302 (m), 1256 (m), 1221 (m), 1172 (m), 1161 (m), 1052 (s), 964 (m), 854 (m), 815 (m), 619 (s), 546 (m), 513 (m), 497 (m), 410 (s). IR (cm−1, A): 417 (s), 489 (m), 510 (m), 548 (m), 616 (s), 811 (s), 858 (m), 938 (m), 1050 (s), 1166 (m), 1219 (m), 1460 (m), 1509 (s), 1593 (s), 1619 (m), 2174 (s). Microanalysis. Calcd for C20H16N12S2 (A·2H2O): C, 34.97; H, 2.35; N, 24.48. Found: C, 34.84; H, 2.58; N, 24.35. Single-Crystal X-ray Diffraction. Single-crystal X-ray diffraction data were collected on A·2H2O at 200 K. In situ removal of guest H2O molecules was carried out on a single crystal heated to 375 K, which was mounted on a loop without the use of oil to ensure that guest molecules could escape. Single-crystal X-ray diffraction data were collected on A at 375, 200, and 100 K. All single-crystal data were collected using a Bruker APEX2 diffractometer equipped with a rotating anode (λ = 0.7017 Å), and data integration and reduction were performed using the Bruker software suite.13 Structural solution for all materials was completed within SHELXS-97 and refined using SHELXL-9714 within the X-SEED user interface.15 All atoms were refined anisotropically, and hydrogen atoms were fixed using the riding model. Further structural analysis details and thermal ellipsoid diagrams (ORTEP 50% probability) are presented in Figures S3− S5. The single-crystal structural refinement parameters and structural details are summarized in Table 1, and selected bond lengths and angles and other parameters are listed in Table 2. CCDC 1480120− 1480123 contain the supplementary crystallographic data for A at 100, 230, and 375 K and A·2H2O at 200 K, respectively. These data are provided free of charge by The Cambridge Crystallographic Data Centre. Thermogravimetric Analysis. A sample of A·2H2O was heated from room temperature to 600 °C at a rate of 1 °C min−1 under a dry nitrogen gas flow (20 mL min−1) using a TA Instruments Discovery thermogravimetric analyzer. Gravimetric Analysis. A H2O sorption isotherm was obtained on a Hiden-Isochema intelligent gravimetric analyzer. Prior to adsorption, the sample was outgassed under vacuum (10−5 bar) at 100 °C

plateau (120 K). This unusual high-spin (HS)/low-spin (LS) state thermal stability is provided by a range of antagonistic hydrogen-bonding interactions7 and the ability of the Hofmann topology to stabilize the resultant lattice distortion. Here, we explore the broad applicability of this approach by investigating a new 1,2,4-triazole-type ligand. To date, the vast majority of multistep spin transitions are generated serendipitously but may be enabled by this overall approach because of the strategically imbedded range of ferro- and antiferroelastic interactions.7a,8 Here, we find, through a range of structure and magnetic studies conducted under temperature and hydrostatic pressure variation, that a multistep SCO is evident, but the spin transition is inhibited when guest molecules are present due to internal pressure effects. Indeed, the ca. 10 % volume change necessary for the HS-to-LS transition to proceed is only possible under hydrostatic pressure application or with the removal of guest molecules. Despite many publications dealing with the application of hydrostatic pressure to SCO polymeric materials,9 few focus on guest-loaded species.10 We show here that such studies are beneficial for uncovering “hidden” SCO properties11 but may also be advantageous for revealing new SCO properties because both host−host and host−guest interactions may be perturbed and/or enhanced in synergy.10,12



230 HS C20H16FeN12PdS2 650.82 triclinic P1̅ 7.2754(9) 7.6058(9) 12.6147(16) 102.524(5) 101.255(5) 90.394(5) 667.44(14) 1.619 2752/0/167 0.0486 [0.0614] 0.1252 [0.1392] 1.085

EXPERIMENTAL SECTION

General Procedures. All reagents were commercially available and were used as received [iron(II) perchlorate was handled carefully and in small amounts to avoid any potential explosions]. Ligand Synthesis. 4-Amino-1,2,4-triazole (1.5 g, 17.5 mmol) and 5-methyl-2-thiophenecarboxaldehyde (2.76 g, 21.9 mmol) were dissolved in 50 mL of ethanol (EtOH) along with 4 drops of concentrated H2SO4. The solution was refluxed and stirred for 5 h. With cooling to room temperature, a white precipitate formed and was washed with cold water (H2O) followed by EtOH. The crude product was recrystallized from EtOH. Mw: 192.25. Yield: 1.7 g (51%). 1H NMR (200 MHz, DMSO-d6, δ/ppm): 9.15 (s, 1H), 9.09 (s, 2H), 7.51 (d, 3JHH = 3.6 Hz, 1H), 6.98 (d, 3JHH = 3.6 Hz, 1H), 2.53 (s, 3H). 13C NMR (300 MHz, DMSO-d6, δ/ppm): 152.8, 146.9, 138.8, 136.0, 133.9, 127.3, 15.5. ESI-MS: m/z 192.87 (M + H peak), 214.87 (M + Na peak). IR (solid, ν/cm−1): 3108 (s), 2960 (m), 1601 (s), 1508 (s), 1490 (s), 1475 (s), 1452 (m), 1379 (m), 1295 (m), 1247 (m), 1220 (m), 1205 (m), 1164 (s), 1058 (s), 1048 (s), 975 (m), 838 (m), 806 (s), 621 (m), 573 (s), 545 (m), 518 (m), 485 (m). Microanalysis. 10491

DOI: 10.1021/acs.inorgchem.6b01686 Inorg. Chem. 2016, 55, 10490−10498

Article

Inorganic Chemistry

nonpenetrating hydrostatic pressure medium Fluorinert with CaF2 as an internal pressure standard. For all data collected, the raw images were processed using Fit-2D.16 Le Bail analyses of the diffraction data were performed within TOPAS.17 Unit cell volume evolution data are presented in the manuscript text, and the thermal dependence of the individual lattice parameters is given in Figures S7−S10. The pressure compression data (Figure S11) was fitted using the software EOS-FIT 5.2.18

Table 2. Variable-Temperature Structural Parameters A·2H2O temperature/K ⟨dFe−N⟩/Åa ∑Fe/degb Fe−NC/deg host···host/Å S(1)···S(1) host···guest/Å N(4)···O(1) O(1)···N(6)

A

200 2.179 32 171.5, 178.2

375 2.178 28.8 170.9, 177.6

230 2.160 29.5 171.8, 177.4

100 1.947 17.6 173.1, 177.7

3.630

4.089

4.004

3.931



RESULTS AND DISCUSSION Synthesis and Characterization. 4-[(E)-2-(5-Methyl-2thienyl)vinyl]-1,2,4-triazole (thiome) was prepared by the same methodology as that of the previously reported analogues.19 Single crystals of A·2H2O were readily prepared via slow diffusion techniques and A by heating to 375 K for >1.5 h. For each, the IR spectra show the strong stretching vibration mode of the cyano groups around 2170 cm−1, consistent with the formation of [FeIIPd(CN)4] grids. Thermogravimetric analysis reveals that the two guest H2O molecules in A·2H2O are removed over the range 100−140 °C to produce the guest vacant framework, A (Figure S1). Gravimetric H2O sorption analyses reveal a very slow uptake of H2O molecules, with the fully hydrated phase being stable to low pressures at ambient temperature (Figure S2). Structure and Magnetic Properties of A·2H2O. Singlecrystal X-ray diffraction on the yellow square-plate crystals of A· 2H2O at 200 K revealed a 2D Hofmann material with H2Ofilled pores located in the interlayer spacing (Table 1). The asymmetric unit consists of a single iron(II) site, one thiome ligand (Figure 1a) bound through the N(1) position of the 1,2,4-triazole ring, a [Pd(CN)4] unit bound through the cyanido nitrogen groups, and a single H2O molecule (Figure 1b). The average Fe−N bond lengths are ca. 2.2 Å, indicative of iron(II) in the HS state (Table 2). Typical Hofmann-type square grid layers of the type [FeIIPd(CN)4] result from the bridging of iron(II) sites by four-connecting [Pd(CN)4] anions (ab plane; Figure 1c). The octahedral coordination of iron(II) sites is completed by axially bound thiome ligands, which protrude above and below each [FeIIPd(CN)4] layer to form an overall 2D layer topology (Figure 1d). The near-planarity of these Hofmann layers is highlighted by the idealized Fe−NC angles of near 180° (Table 2). The layers pack efficiently along the c direction with an almost complete head-to-tail overlap between thiome ligands of adjacent layers; the latter form weak π-stacking interactions along the a axis (Figure 1e). The unbound nitrogen atoms of the 1,2,4-triazole rings are located in a trans orientation across each iron(II) site, and the thiophene sulfur groups of each ligand are also trans with respect to this unbound nitrogen atom. Thus, within the interlayer spacing, there are small voids lined with hydrogen-bond donor groups from both the unbound triazole nitrogen and thiophene sulfur atoms. Indeed, the sulfur atoms form contacts with adjacent thiome ligands along the b direction and the N(2) triazole atoms form hydrogen-bonding interactions with guest H2O molecules, as illustrated in Figure 1f. In a comparison of this overall structure with the previously reported analogue [FeIIPd(CN)4(thtrz)2]·(EtOH,H2O),19 the most notable difference is that the layers are planar in A·2H2O rather than undulating. In [FeIIPd(CN)4(thtrz)2]·(EtOH,H2O), the layer undulation results in pairs of interacting ligands within each layer spaced by large solvent-filled pores. Here, all of the ligands are equally spaced and do not show any interactions

2.860 3.005

a

Average Fe−N bond distance. bOctahedral distortion parameter calculated by the sum of |90 − θ| for the 12 N−Fe−N angles in the octahedron.2 overnight. The sample was cooled to the analysis temperature (25 °C) and the dry mass recorded. The sample chamber was pressurized to a set pressure and allowed to equilibrate for 240 min before moving to the next pressure point. Magnetic Susceptibility Measurements. Data for magnetic susceptibility measurements were collected on a Quantum Design Versalab Measurement System with a vibrating sample magnetometer attachment. Measurements were taken continuously under an applied field of 0.3 T over the temperature range 300−50−300 K, at a ramp rate of 2 K min−1. In situ desolvation was carried out to produce A with heating to 375 K for 1.5 h, followed by measurements over the range of 375−50−300 K. A variety of scan rates were applied to the measurements, but only the 2 K min−1 data are presented here because there was no rate dependence. Magnetic measurements under pressure were performed on A· 2H2O and A using a hydrostatic pressure cell, specially designed for the SQUID setup in the València laboratory (SQUID specifications: Quantum Design MPMS2 SQUID magnetometer operating at 1 T and 1.8−300 K), made of hardened beryllium bronze with silicon oil as the pressure-transmitting medium and operating over the pressure range of 1 bar to 12 kbar. The compound (13 mg) was packed in a cylindrically shaped sample holder (1 mm diameter and 5−7 mm length) made up of very thin aluminum foil. The pressure was calibrated using the transition temperature of superconducting lead of high purity (99.999%). Differential Scanning Calorimetry (DSC). Calorimetric measurements were performed on A by using a differential scanning calorimeter (Mettler Toledo DSC 823e). Measurements were carried out by using ca. 3 mg of the sample sealed in a 40 μL aluminum pan with a mechanical crimp (an empty pan was used as a reference). The sample holder was kept in a drybox under a flow of dry nitrogen gas (50 mL min−1) to avoid H2O condensation. The differential scanning calorimeter was liquid-nitrogen-cooled. Temperature and heat-flow calibrations were made with standard samples of sapphire. Sample desolvation was achieved within the DSC machine by heating a sample of A·2H2O to 140 °C. Holes were placed in the lid of the aluminum pan to allow H2O vapor to escape. Powder X-ray Diffraction Measurements. The X-rays (17.03 keV and 0.72808 Å) available at the 17-BM beamline at the APS at Argonne National Laboratory were used in combination with a PerkinElmer area detector with a carbon window to record diffraction patterns. For variable-temperature measurements, a polycrystalline sample of A·2H2O was ground as a slurry and loaded into a quartz capillary (0.7 mm diameter) that was open at one end to enable in situ desolvation. The sample temperature was controlled using an Oxford Cryosystems open-flow cryostream, and the data were collected in 20 s exposures during continuous ramping over the range of 200−375 K for A·2H2O and 375−100−200 K for A at 120 K h−1. This corresponds to the collection of diffraction images at 2 K intervals. LaB6 was used as a standard. Variable-pressure measurements were carried out on a polycrystalline sample of A·2H2O loaded into a diamond anvil cell using the 10492

DOI: 10.1021/acs.inorgchem.6b01686 Inorg. Chem. 2016, 55, 10490−10498

Article

Inorganic Chemistry

Figure 2. χMT versus temperature for A·2H2O (blue) and A (red) measured at 2 K min−1.

observed with a hysteresis loop of 20 K (T1/2↓ = 184 K; T1/2↑ = 204 K). A DSC measurement was also carried out on A in the 120−300 K temperature range, and the critical temperatures obtained from the exothermic (T1/2↓ = 183 K) and endothermic (T1/2↑ = 206 K) curves are in good agreement with the magnetic data (Figure S6). Structural characterization of the dehydrated phase was performed in situ by heating a single crystal to 375 K and also by powder X-ray diffraction. There was no loss in the diffraction intensity or crystallinity with guest removal. The achievement of such robustness is relatively uncommon in 2D materials,20 being more prevalent in three-dimensional (3D) framework materials, and exemplifies the pseudo 3D nature imparted by the network of intermolecular interactions. The crystal symmetry and lattice parameters of the dehydrated phase remain unchanged compared to those of A·2H2O, with slight thermal expansion at 375 K (Table 2). Single-crystal analysis at 375 K confirms retention of the Hofmann 2D topology and host−host interactions compared to A·2H2O (Figure 3). As an initial overview of the magnetostructural behavior of A over the spin-transition temperature range, variable-temperature synchrotron powder X-ray diffraction studies were carried out on a bulk crystalline sample. Plotting the peak position evolution versus temperature clearly indicates a single-phase behavior over this abrupt SCO profile, where discontinuous (first-order) shifts in the Bragg peaks occur at the temperatures of a HS ↔ LS transition (Figure 4a). While there is a cooperative shift between the two spin states within the sample, there is a small temperature interval (