Heterobimetallic Lantern Complexes and Their Novel Structural and

Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

Heterobimetallic Lantern Complexes and Their Novel Structural and Magnetic Properties Stephanie A. Beach and Linda H. Doerrer* Department of Chemistry, Boston University, Boston, Massachusetts 02215, United States

CONSPECTUS: As the scale of microelectronic circuit devices approaches the atomic limit, the study of molecular-based wires and magnets has become more prevalent. Compounds with quasi-1D geometries have been investigated for their electronic conductivity and magnetic properties with potential use as nanoscale circuit components and information storage devices. To increase the number of compositionally tailored molecular systems available to study, we have taken a building-block, bottom-up approach to the development of improved electronic structure and magnetic properties of quasi-1D arrays. Over the past decade, a large family of asymmetric complexes that can assemble into extended arrays has resulted. Lantern (or paddle-wheel) complexes with conventional {O, O} donor carboxylates are legion, but by the use of monothiocarboxylate ligands and hard−soft Lewis acid−base principles, dozens of new lantern complexes of the form [PtM(SOCR)4(L)] (M = Mg, Ca, Cr, Mn, Fe, Co, Ni, Zn; R = Ph (tba = thiobenzoate), CH3 (SAc = thioacetate); L = neutral or anionic ligand) have been prepared. Depending on M and L, new intermolecular arrangements have resulted, and the magnetic properties have proven particularly interesting. In the solid state, the [PtM(SOCR)4(L)] building blocks are sometimes isolated, sometimes form dimers, and can be induced to form infinite chains. The versatility of the lantern motif was demonstrated with a range of axial ligands to form both terminal and bridged complexes with various 3d metals and two different substituted thiocarboxylate backbone ligands. Within the dozens of crystallographically characterized compounds that make up this family of lanterns, several different structural motifs of solid-state dimerization were observed and divided into four distinct categories on the basis of their Pt···Pt and Pt···S distances and relative monomer orientations. Among all of these compounds, three novel magnetic phenomena were observed. Initially, long-range antiferromagnetic coupling between two metals more than 8 Å apart was observed in solid-state dimers formed via metallophilic Pt···Pt interactions and could induced by choice of the terminal L group. An infinite chain was prepared in [PtCr(tba)4(NCS)]∞ that displays ferromagnetic coupling between Cr centers with J/kB = 1.7(4) K. Homobimetallic quasi-1D chains of the form [Ni2(SOCR)4(L)]∞ (R = Ph, CH3; L = DABCO, pyz) were also prepared with S = 1 {Ni2} building blocks in which the Ni centers have two different spin states with weak antiferromagnetic coupling along the chain, such that −0.18 > J/kB > −0.24 K. In the [Ni2(tba)4(quin)] derivative, a solid-state dimer forms with a bridging square conformation by interlantern Ni2S2 interactions and displays unusual S = 1 configurations on both Ni centers and weak antiferromagnetic coupling between them.

1. INTRODUCTION

application of their arrays in nanoscale devices and microelectronics. One-dimensional systems require an anisotropic building block with carefully chosen ligands to promote desired electronic and magnetic properties. Examples of 1D geometries shown in Scheme 1 include arrays bridged through metallophilic interactions, coordination polymers,6 and the more

Complexes that are able to assemble in potentially infinite onedimensional (1D) arrays with useful electronic and magnetic properties have created a diverse, multidisciplinary field of study.1 Combinations of solid-state physics, synthetic inorganic chemistry, and organometallic and polymer chemistry have led to a vast array of 1D materials that have been used for vapochromic sensors,2 luminescent materials,3,4 and magnetic resonance (MR) imaging,5 to name a few. Nanowires6 and molecular magnets7 are intensely investigated materials for © XXXX American Chemical Society

Received: November 21, 2017

A

DOI: 10.1021/acs.accounts.7b00585 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Scheme 1. 1D Systems from the Literature: (a) EMAC;9 (b) Krogmann Salt;10 (c) Chugaev’s Red Salt;12 (d) Semiconducting Chain;13 (e) CHAC Ferromagnetic Chain;14 (f) Ferromagnetic Bimetallic Chain15

with both hetero- and homobimetallic cores via a lantern or paddlewheel motif. Similar to Chugaev’s red salt and the Krogmann salts, the lantern complexes described herein can display metallophilic Pt···Pt interactions that lead to solid-state dimers with interesting electronic and magnetic properties. With the careful choice of bridging ligand and 3d metal, various quasi-1D chains with anisotropic magnetic properties have also been synthesized and studied.

recently studied bridged oligomeric compounds known as extended metal atom chains (EMACs).8,9 One of the first classes of 1D compounds to be prepared are Krogmann-type salts, such as K2[Pt(CN)4], which stack in a linear array of anions via Pt···Pt bonding.10 Similarly, Chugaev’s red salt, [(CH3NC)2Pt(C4H9N4)]Br·4H2O, forms both cation pairs and zigzag chains in the solid state, depending on the anion.1,11 Spectroscopic studies of Chugaev’s cation showed cation selfassociation through metallophilic Pt···Pt interactions in solution as well. The close Pt···Pt contacts in this case heavily depend on the chosen anion and the hydrogen-bonding ability of the outer N−H groups.12 Alongside structural studies, interest in promoting electronic conductivity in 1D arrays is continuous, an example of which can be seen in [Cr(isoq)2(NCS)4]−, which through close S···S contacts between the anion and donor (BDH-TPP) forms paramagnetic and semiconducing chains.13 The magnetic properties of 1D systems have also been of interest. For example, ferromagnetically coupled 1D chains of (CH3)3NHCuCl3·2H2O (CHAC) with J/kB ≥ 70 K have been synthesized for the study of their spin dynamics.14 More recently, oxalate-based bimetallic ferromagnetic chains [Mn(H2O)2Cr(ox)3]nn− with [K(18-crown-6)]+ have been synthesized, in which 2D anionic networks are formed via hydrogen bonding of a H2O bound to Mn in {Mn(ox)Cr} chains.15 Building blocks with a heterobimetallic {MM′} core are inherently asymmetric, which can lead to useful electronic and magnetic properties with careful choice of ligands and metal. In this Account, the assembly of quasi-1D chains is accomplished

2. HETEROBIMETALLIC LANTERN SYNTHESIS The controlled synthesis of heterobimetallic lanterns was achieved using a thiocarboxylate backbone ligand with O and S donors to selectively coordinate hard and soft metals, respectively, to form complexes that precipitate from water and are crystallized from polar organic solvents. The lantern motif contains three variables that lead to a variety of complexes, as shown in the generic compound in Scheme 2. The 3d metal, the axial ligand, L, and the carboxylate substituent, R, can all be Scheme 2. Lantern Motif Structural Variables

B


Article

Accounts of Chemical Research Scheme 3. Heterobimetallic Pt−M Thiocarboxylate Lantern Complexes

Scheme 4. Homobimetallic [Ni2(SOCR)4] Species and Their Spin States

of Cu(0) and apparent S oxidation.16 In the isolated complexes, the soft Pt center is bound in a {PtS4} environment and the hard alkaline-earth or 3d metal in a {MO4X} or {MO4L} environment. The non-Pt metal determines the dia- and paramagnetic character of each complex as well as exhibiting some size-dependent effects. The axial ligands vary in their ability to be bridging or terminal, allowing for the formation of quasi-1D arrays. The thioacetate and thiobenzoate ligands both

independently varied to tune the electronic structure and magnetism of the system. Dozens of new lantern complexes of the form [PtM(SOCR)4(L)] (M = Mg, Ca, Cr, Mn, Fe, Co, Ni, Zn; R = Ph, CH3; L = neutral or anionic ligand) have been prepared, as summarized in Scheme 3. To date, Cu-containing {PtCu} species are missing from this array of complexes. In synthetic attempts, selective Pt coordination followed by addition of a Cu(II) salt consistently indicates the formation C


Article

Accounts of Chemical Research contain the mixed S/O donor system required for selective coordination but show little effect of this substitution on the formation of intermolecular interactions between monomeric units, as described in section 3.17 Homobimetallic complexes of the form [Ni2(tba)4(L)] (tba = thiobenzoate) can also be synthesized in a similar manner as the heterobimetallic species. There are numerous examples of Ni-based homobimetallic {O,O} carboxylate lantern systems in the literature, but only a handful of {S,O} complexes.18−21 Specifically, the lantern [Ni2(SOCR)4(EtOH)]21 is used as a precursor for the synthesis of lanterns with quinuclidine (quin) and pyridine terminal ligands as well as for the formation of quasi-1D arrays with 1,4-diazabicyclo[2.2.2]octane (DABCO) and pyrazine bridging ligands.22 As a nonbridging control for the pyrazine-ligated monomeric lantern unit, pyridine was used as a terminal ligand. This effort did not afford the dinuclear complex [Ni2(SOCR)4(py)] but instead gave the known complex [Ni(SOCR)2(py)2].23,24 From the above variables, over 40 lantern species have been crystallographically characterized and their electronic and magnetic properties explored, as can be seen in the range of compounds prepared to date in Schemes 3 and 4 and collected in Table 1.

Table 1. Compound Formulae and Numbers

3. INTRA- AND INTERMOLECULAR STRUCTURES OF LANTERN COMPLEXES The basic intramolecular coordination geometry shown in Scheme 2 was confirmed by single-crystal X-ray diffraction studies for all of the compounds in Schemes 3 and 4. In each case, the Pt−S (or Ni−S) and M−O bond lengths are unexceptional, and the five-membered chelate ring is twisted as a result of the size difference between these two distances, preventing the {M2SOC} atoms from being coplanar. The slight variation among M(II) ion sizes from 0.97 Å in Mn(II) to 0.83 Å in Ni(II) causes some subtle variations in intralantern metrical parameters. A comparison of the Pt−M distances in Co, Ni, and Zn compounds and their formal shortness ratios (FSRs) are given in Table 2. These data indicate that no proper covalent bonds exist in any of the lanterns, but the distances are influenced by the 3d metal and the terminal ligand. The FSRs in the Co complexes are the largest, and those with Ni and Zn are similar, though the Ni series has a more narrow range. In each group, the largest FSRs are in the salt complexes with NCS ligands, and those with Na bound to {12C4} are larger than with {15C5}. The species with pyNH2 ligands have the next longest ratios, but the remaining three ligands in these series (H2O, py, and pyNO2) do not show any clear pattern. These complexes also form dimers in the solid state, but which dimer forms does not correspond straightforwardly to M or L (vide infra). Lanterns of the form [PtM(SOCR)4(L)], in which L is bound to the alkaline-earth/3d metal, have a square-planar environment at Pt(II) with no axially coordinated ligand on the Pt. Two exceptions have been observed to date. There is strong indirect evidence that the [PtM(SOCR)4] lanterns associate intermolecularly through binding of the carboxylate donors of one lantern to the metal centers of adjacent lantern units.25 These insoluble species can be broken up only with strongly donating solvents that bind to M (e.g., DMSO (27), DMF (28)),26 or pyridine can also be bound axially to M and Pt, as seen in [(py)PtM(SOCR)4(py)] (M = Co (17), Ni (19), Zn (22)). These are highly rare examples of an octahedrally coordinated Pt(II) species in which the additional py

compound

number

ref

[PtMg(SAc)4(OH2)] [PtMg(tba)4(OH2)] [PtCa(tba)4(OH2)] [PtFe(tba)4(OH2)] [PtCo(SAc)4(OH2)] [PtCo(tba)4(OH2)] [PtNi(SAc)4(OH2)] [PtNi(tba)4(OH2)] [PtZn(SAc)4(OH2)] [PtZn(tba)4(OH2)] [PtCo(SAc)4] [PtNi(SAc)4] [PtZn(SAc)4] [PtCo(SAc)4(pyNO2)] [PtNi(SAc)4(pyNO2)] [PtZn(SAc)4(pyNO2)] [PtFe(SAc)4(py)] [PtCo(SAc)4(py)] [PtNi(SAc)4(py)] [PtZn(SAc)4(py)] [(py)PtCo(SAc)4(py)] [(py)PtNi(SAc)4(py)] [(py)PtZn(SAc)4(py)] [PtMn(SAc)4(pyNH2)] [PtFe(SAc)4(pyNH2)] [PtCo(SAc)4(pyNH2)] [PtNi(SAc)4(pyNH2)] [PtZn(SAc)4(pyNH2)] [PtCo(SAc)4(DMSO)]DMSO [PtNi(SAc)4(DMF)]DMF [PtCr(tba)4(NCS)]∞ {Na(12C4)}2[PtMn(SAc)4(NCS)] {Na(15C5)}[PtCo(SAc)4(NCS)] {Na(12C4)}2[PtCo(SAc)4(NCS)] {Na(15C5)}[PtNi(SAc)4(NCS)] {Na(12C4)}2[PtNi(SAc)4(NCS)] {Na(15C5)}[PtZn(SAc)4(NCS)] {Na(12C4)}2[PtZn(SAc)4(NCS)] [Ni2(tba)4(quin)] [Ni2(SAc)4(DABCO)]∞ [Ni2(tba)4(DABCO)]∞ [Ni2(SAc)4(pyz)]∞ [Ni2(tba)4(pyz)]∞ [(py)Ni(tba)2(py)]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 42 18 20 23 17 19 22 43 44 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 21

30 30 30 16 17 16 17 16 17 30 17 17 17 17 17 17 31 26 26 26 26 26 26 31 31 26 26 26 26 26 27 27 27 27 27 27 27 27 22 22 22 22 22 22

coordination causes a shortening of the Pt···M contact by ∼0.04 Å. Thermographic analysis demonstrated that the Ptbound pyridine ligand is easily removed with heat, leaving the more common [PtM(SOCR)4(py)] structure.26 Intermolecular solid-state dimers were first discovered in the complexes [PtM(tba)4(OH2)] (M = Co (6), Ni (8)),16 which have close intermolecular Pt···Pt metallophilic interactions. Many, but not all, of the neutral [PtM(SOCR)4(L)] complexes exhibit dimeric interactions in the solid state. These interactions have been organized into four categories, as depicted in Scheme 5: staggered, totally eclipsed, partially eclipsed, and square. Dimers are placed into these categories on the basis of (i) the intermolecular Pt···Pt distances, (ii) the intermolecular Pt···S distances, and (iii) the M−Pt−Pt interlantern angle. The staggered configuration was first observed in these complexes, D


Article

Accounts of Chemical Research Table 2. Metal−Metal Distances within Lantern Complexes

Scheme 5. Classification of Dimeric Lanterns

contacts allowing backbone eclipse, with a M−Pt−Pt angle close to 180°, whereas the partially eclipsed complexes have shorter Pt···S contacts and angles around 160°. Similar to the partially eclipsed complexes, square dimers have close Pt···S contacts between lanterns, but they differ in their 140° M−Pt−

with short Pt···Pt intermolecular contacts forcing a staggered conformation of the thiocarboxylate backbone when viewed along the M−Pt−Pt−M vector.16,17 The totally and partially eclipsed cases differ in the number and orientation of Pt···Pt versus Pt···S contacts and the magnitude of the M−Pt−Pt angle. The totally eclipsed complexes have shorter Pt···Pt E


Article


LMCT is primarily from S to Pt.28 All of the paramagnetic complexes except Mn(II)-containing ones also have 3d−3d transitions, as expected, as well as an intermetallic charge transfer band in the near-infrared (NIR) region indicative of Pt−M bonding.26 No features corresponding to the solid-state dimerization are observed in the solid-state UV−vis−NIR spectra. A series of [PtCo(SAc)4(L)] spectra are shown in Figure 1.

Pt angle, preventing the previously observed partial eclipsing of the thiocarboxylate backbone ligands. In the typical lantern, the alkaline-earth/3d metal M is square-pyramidal, although 3 forms a tail-to-tail carboxylate dimer in which the Ca atom not only binds the axial water ligand but also forms a contact with a carboxylate oxygen of an adjacent lantern. This six-coordinate Ca has a trigonal-prismatic geometry instead of the usual square-pyramidal configuration. This displacement may be a result of the size difference of Ca(II), whose ionic radius of 1.14 Å is roughly 68% larger than those of the previously used divalent 3d metals. Mg(II) has an ionic radius of 0.97 Å, which is comparable to that of Mn(II), and consequently, this distortion is not observed in the Mg lanterns. The geometric adjustment required for the oxygen atoms to support a six-coordinate Ca environment may also be a contributor to the observed distortion.28 Scheme 5 shows that not all complexes with the same axial ligand dimerize in the same fashion. There are only two families of compounds that all have the same motif: pyridine and pyNH2 ligands, which form square dimers. In the [PtM(SOCR)4(OH2)] lantern species, all but the thiobenzoate lanterns of Ca, Mg, and Fe (2−4, respectively) are staggered. The Co (14) and Zn (16) NO2py analogues are partially eclipsed, and the Ni complex (15) is staggered. The thiocyanate-ligated lanterns fall into two categories: totally eclipsed and partially eclipsed. The {15C5}-encapsulated Na cations of 31, 33, and 35 bind to the thiocyanate that bridges Na and the {MO4} center, and the dimers have the short interlantern Pt···S contacts of the partially eclipsed category. This interaction is absent with encapsulated {Na(12C4)2} cations, and totally eclipsed geometries with shorter Pt···Pt interactions are observed in 32, 34, and 36. The geometries adopted by the dimers are clearly influenced by the M-bound ligand L, which influences the electron density at Pt and therefore affects the van der Waals dispersion forces. Intermolecular interactions also influence the relative energies of different configurations, and further computational work is underway. In the solid state,22 homobimetallic [Ni2(tba)4(quin)] (37) forms a dimer through intermolecular Ni2S2 linkages with a Ni−Ni−Ni angle of ∼134.6° and intermolecular Ni···Ni < Ni··· S distances that form a square geometry similar to the heterobimetallic species, as shown in Scheme 4.

Figure 1. Visible−NIR spectra of [PtCo(SAc)4(L)] compounds. Adapted from ref 26. Copyright 2013 American Chemical Society.

An initial molecular orbital (MO) description16 was formed for the general complex [PtCo(SAc)4(L)] via density functional theory (DFT) calculations, showing that greater intramolecular interaction between M and Pt orbitals leads to greater electron density in the Pt-based antibonding orbitals. The intermolecular overlap of these Pt-based orbitals then promotes greater Pt···Pt coupling, resulting in the formation of the solid-state dimers.17 Later structures bore out this hypothesis in the formation of the four motifs from Scheme 5. The pyridine-based axial ligands can donate by different degrees directly to the 3dz2 orbital of M and indirectly to the Pt 5dz2 orbital, as shown in Scheme 6. An increase in donation with pyNH2 (24−26, 43, and 44) stabilizes the Pt−M σ orbital, leading to the observed decrease in Pt−M distance with decreasing Pt−Laxial distance. An increase in Pt−M distance is consistent with the weaker donation and short intermolecular Pt···S contacts seen in [PtCo(SAc)4(pyNO2)] (14) and [PtZn(SAc)4(pyNO2)] (16). As the calculated pKa of the pyridine-based conjugate acid increases, there is a corresponding M−N bond length decrease (Figure 2) due to the increased electron density being donated from the pyridine. The pyNO2 derivatives deviate from the other pyridine-based ligands, with much shorter M−N distances than expected, possibly because the diminished σ-donating ability of Npy favors the formation of short Pt···Pt contacts, the effect of which can be seen in the different dimer geometries for the pyNO 2 and pyNH 2 complexes. The {15C5} thiocyanate lantern derivatives 31, 33, and 35 also display an inversely proportional relationship between the Pt−M and Pt···S distances due to axial ligand donation. Interactions between Na and SNCS pull electron density from Pt and promote shorter Pt···S contacts and longer Pt−M distances. When this interaction is removed in the {12C4} analogues 32, 34, and 36, the thiocyanate ligand donates more electron density and stabilizes the Pt−M σ orbital, causing

4. LANTERN MONOMER AND DIMER ELECTRONIC STRUCTURES The electronic structures of the lantern complexes have been revealed through a combination of spectroscopic and computational data. The effect of the axial ligand L, carboxylate substituent R, and 3d metal M have been evaluated. The chosen paramagnetic M centers are all high-spin, consistent with the relatively weak field of the carboxylate O donors. The {PtS4} center is always low-spin and diamagnetic, and it combines with the high-spin, paramagnetic 3d metal center {MO4} to form an overall high-spin system.16 Electron density calculations show delocalization of unpaired electron density from the 3dz2 orbital to Pt, allowing for magnetic communication across the monomeric lantern unit, as discussed in detail in section 5.16 Solution-phase UV−vis spectroscopy showed ligand-to-metal charge transfer (LMCT) from the backbone ligand in every case. Comparison of these absorptions to those of both {K2[Pt(SAc)4]} and the 3d metal acetates indicates that the F


Article

Accounts of Chemical Research Scheme 6. Orbital Path for Electron Donation and Withdrawal along the Pt−M−L Vector

similar Pt···Zn and Pt···Pt distances, suggesting that the luminescence is a result of excitation from a Pt···Zn-based orbital to the π* orbital of the substituted pyridine axial ligand or thioacetate backbone ligand, indicating a Pt···Zn metallophilic interaction in these compounds. The Pt−Zn calculated bond orders are demonstrably larger than those of the Mg and Ca species, suggesting increasing covalency with increasing atomic number. From these data, the PtMg complexes show exclusively ionic metal−ligand interactions while the PtZn species are more dative and covalent in character, with the Ca species lying between the two. The study of these diamagnetic complexes definitively shows the large effect of the Lewis acidic 3d metal and bridging ligand on the Pt environment, which in turn can be used to fine-tune the desired electronic and magnetic properties.28 In the Ni2 homobimetallic lanterns, the choice of axial ligand has a large effect on the spin of the complex. The terminally ligated lantern [Ni2(SOCR)4(EtOH)] has been previously studied and shown to exhibit a high-spin S = 1 {NiO4} unit and low-spin S = 0 {NiS4} center.21 This EtOH-based starting material was utilized to synthesize new Ni2 lanterns with both terminal and bridging axial ligands. Superconducting quantum interference device (SQUID) measurements of [Ni2(SOCR)4(L)] with the bridging ligands DABCO and pyrazine show an S = 1 lantern core.22 DFT calculations indicate that the spin-localized structure is more energetically stable by ∼6 kcal/mol than the spin-delocalized case in which each Ni is S = 1/2.22 When the terminal ligand quinuclidine is used in [Ni2(tba)4(quin)] (37), however, SQUID data for the dimer show evidence of a highly unusual high-spin S = 1 on both the {NiO4} and {NiS4} centers, as indicated in Scheme 4. Compared with the EtOH-containing precursor, the effect of the heteroatom (O vs N) coordination and the longer Ni···Ni distance observed in 37 could explain this spin-state difference. In 37, the interlantern Ni2S2 interactions lead to a weaker {NiS4} ligand field and therefore much longer Ni···Ni distances, similar to the previously described heterobimetallic square complexes. The Ni···Ni distance in dimer 37 (2.5747 Å) falls between the Ni···Ni distances observed in the Ni2 1D chains 38−41 (2.5316−3.595 Å).22 When compound 37 was studied as a [Ni2(tba)4(quin)] monomer with a triplet ground state and no fifth Lewis basic ligand on the {NiS4} Ni atom, an electronic structure similar to that observed in the lanterns composing chains 38−41 was observed.22 These data suggest a large effect of the capping axial ligand on the S-ligated Ni when two such {Ni2} units dimerize. The additional S donor and additional Lewis acid (Ni) on one carboxylate may create a more

Figure 2. M−N distances in [PtM(SAc)4(pyX)] as functions of conjugate acid axial ligand pKa: (red) {PtCo}, (blue) {PtNi}, (black) {PtZn}, (green) {PtFe}, (purple) {PtMn}; (triangles) pyNO2, (circles) py, (squares) pyNH2.

shorter Pt−M distances and shorter Pt···Pt interactions.27 Furthermore, DFT calculations performed on the water-ligated species 4, 6, and 8 confirm the donor (Pt 5dz2)−acceptor (M) character expected with the electron-withdrawing thiobenzoate oxygen atoms, resulting in additional donation to the 3d metal center and the observed shortening of the Pt−M bond.16 More evidence for the influence of the electron density at the Pt center on both the electronic and solid-state structures of the lantern complexes was obtained from 195Pt NMR studies of complexes with M = Mg, Ca, and Zn. The chosen array of complexes again tested the influences of the axial ligand L, the carboxylate substituent R, and M on the electron density of the Pt center. Both the thiocarboxylate substitution and terminal axial ligand choice (1−3, 9, 10, 16, 22, 23, and 26) show a modest influence on the Pt center, as seen via the 195Pt NMR shift. With thiocarboxylate variation, there is a downfield shift of ∼6 ppm in 2 (R = Ph) versus 1 (R = CH3) and a downfield shift of ∼10 ppm in 10 (R = Ph) versus 9 (R = CH3). Comparison of the terminal ligands in the various Zn species (9, 10, 16, 22, 23, and 26) shows a range of ∼11 ppm in their shifts, with the most upfield being L = OH2 (9) and the most downfield L = pyNH2 (26). There is a more significant change due to the choice of the Lewis acidic 3d metal, with the largest shift observed in the Zn family of lanterns.28 Despite the fact that the Zn and Mg complexes are isostructural, the chemical shift data suggest greater interaction of Pt with Zn than with Mg. This result is further confirmed in the comparison of nonluminescent [PtZn(SAc) 4 (py)] (23) and [PtZn(SAc)4(pyNO2)] (16) to luminescent [PtZn(SAc)4(pyNH2)] (26) and [PtZn(SAc)4(OH2)] (9). The four complexes have G


Article


Figure 3. Antiferromagnetic coupling constants of solid-state dimers containing Co (circles) and Ni (triangles). All of the structures are shown as monomers for clarity. Dimer structures are shown in Scheme 5.

octahedral ligand-field splitting on the {NiS5} center. Work is currently underway to test this hypothesis.

have led to the exploration of these building blocks for 1D chain complexes.

5. MAGNETIC INTERACTIONS When no interlantern magnetic coupling is observed, the complexes display Curie−Weiss behavior. The four different dimer geometries of Scheme 5 give rise to a range of different magnetic interactions between the two 3d metal centers. The staggered dimeric lanterns with the shortest Pt···Pt distances (5−8, 14, and 15) display antiferromagnetic coupling between 3d metal centers more than 8 Å apart as the temperature is decreased. In these cases, the Ni-containing species always has larger antiferromagnetic coupling than the Co analogue, as shown in Figure 3. This increase in coupling is evidence for better intramolecular Pt−Ni overlap. When the axial ligand favors the square dimer configuration, as in 4, 18, 20, and 24−28, no coupling is observed. In these cases, the close Pt···S intermolecular distances dominate, causing a much smaller degree of Pt···Pt overlap, such that zero-field splitting is the dominant magnetic effect over any antiferromagnetic coupling.26 The shift of the axial ligand from water to pyNO2 leads to a reduction in the antiferromagnetic coupling in both the Co and Ni species 14 and 15; however, the degree of this reduction is not consistent, with a 2-fold decrease in the Co case (14) and a 4-fold decrease in the Ni case (15) compared with 5 (Co) and 7 (Ni) respectively. The coupling is not significantly changed in the dehydrated lanterns 11−13, however, signifying either that the water ligand has a minimal role in the coupling or that the interaction with an adjacent thiocarboxylate ligand leads to similar coupling. There are minute differences in coupling between the thioacetate and thiobenzoate bridging ligands, indicating that the axial ligand has the largest and most consistent effect on the observed coupling.17 The observed coupling over 8 Å apart shows the potential for long-range coupling in these systems, perhaps through close Pt···Pt metallophilic interactions. The choice of terminal L group is able to control this coupling. When the axial ligand favors a square configuration, as with L = py or pyNH2, the intermolecular Pt···S distances are shorter than the Pt···Pt distances, and the antiferromagnetic coupling has less effect than the zero-field splitting as seen in the χmT data fitting.16,17,26 These combined structural and magnetic studies

6. QUASI-1D ARRAYS The isolation of three [(py)PtM(SOCR)4(py)] compounds, 17, 19, and 22, indicated that axial coordination of small molecules to the Pt centers is possible and therefore that chains might be made with the correct choice of ligand. To explore this idea, thiocyanate (NCS) was chosen as a bridging ligand to favor a {M−NCS−Pt} bridging motif. When divalent Mn−Ni and Zn are used, an overall negatively charged species is formed, requiring a countercation. The first cation studied was {Na(15C5)}+, which did not result in 1D assembly of the lantern units because of Na+ interactions with SNCS. The {Na(12C4)2}+ complexes completely isolate Na and prevent the S···Na interaction, but chains were not formed in these complexes with any {Pt−M(II)} pairing.27 The lack of extended structure formation even with complete isolation of Na suggested that the electrostatic repulsion between lantern units prevented formation of the desired 1D array. To decrease this repulsion, neutral lanterns were synthesized with Cr(III) in place of M(II). Single-crystal X-ray diffraction studies confirmed the formation of a zigzag-based chain of the form [PtCr(tba)4(NCS)]∞ (29) in which Cr is bound to NNCS and Pt is bound to SNCS with unexceptional CN and C−S bonds.27 Magnetic studies of this new quasi-1D chain indicated ferromagnetic interactions between S = 3/2 Cr(III) centers.27 Although the shortest intrachain distance between Cr centers is 7.772(1) Å, there is still electronic communication between lantern units with J/kB = 1.7(4) K and g = 2.16(2). A Curie− Weiss fit of χm−1 vs T data at lower temperatures suggested that the net exchange coupling can be attributed almost entirely to the 1D interaction as opposed to 3D magnetic ordering of the chains. There is no linear trend in χm−1 versus T at higher temperatures, suggesting some competing intra- and interchain interactions. However, the fact that the shortest interchain Cr− Cr distance is 9.86(1) Å, which is longer than the intrachain Cr−Cr distance by more than 2 Å, further indicates that the 1D Cr−Cr interaction most likely supersedes the 3D effect. The field and temperature dependences of the magnetization of 29 show discontinuities at low temperatures, suggestive of the onset of the aforementioned weak 3D ferromagnetic ordering, H


Article

Accounts of Chemical Research which is expected given the zigzag nature of the chains.27 The predominance of ferromagnetic interactions can be rationalized with an MO description of the Pt(II)−Cr(III) interaction. The Cr(III) ion is in a tetragonally distorted octahedral geometry. The π orbitals through which the bridging NCS ligand can transmit spin density are orthogonal to the Pt-based 5dz2 orbital on an adjacent lantern, resulting in intrachain ferromagnetic interactions between Cr(III) centers. Quasi-1D arrays were also formed with the homobimetallic Ni2 lanterns. The double N-donor linkers in the form of L = DABCO or pyrazine were combined with Ni2 complexes to create chains 38−41 (Scheme 4), which were compared with the terminal quinuclidine derivative 37. All three of these chains displayed the same temperature- and field-dependent magnetization properties, consistent with S = 1 ground states. Their similar χmT products showed a small linear increase at temperatures above 10 K and a lack of curvature at these higher temperatures, consistent with temperature-independent paramagnetism, meaning each S = 1 state is relatively well isolated from adjacent lantern units.22 In each case, a sharper downturn in χmT at lower temperatures indicates very weak antiferromagnetic interactions between lantern units. Compared with the {PtCrNCS} 1D chain (29), the coupling within these Ni2 chains is ∼10 times smaller, indicating that the bridging DABCO and pyrazine do not allow for much magnetic communication. This weak coupling was also confirmed computationally via an estimate of the exchange coupling in [L{Ni(SOCR)4}L{Ni2(SOCR)4}L] with L = DABCO or pyrazine.22

DiSalvo. Her Ph.D. thesis work was done at MIT with Stephen Lippard, after which came postdoctoral studies with Malcolm Green at the University of Oxford. She served on the faculty at Barnard College from 1999 to 2006. Since that time she has been at Boston University, where she is currently an Associate Professor in Chemistry and in the Division of Materials Science and Engineering. She has been recognized with various awards, including an NSF CAREER Award, a Henry Dreyfus Teacher-Scholar Award, and a Fulbright Scholar Award.

■

(1) Marks, T. J.; Kalina, D. W. Highly Conductive Halogenated LowDimensional Materials. In Extended Linear Chain Compounds; Miller, J. S., Ed.; Plenum Press: New York, 1982; Vol. 1, Chapter 6, pp 197− 331. (2) Drew, S. M.; Janzen, D. E.; Buss, C. E.; MacEwan, D. I.; Dublin, K. M.; Mann, K. R. An electronic nose transducer array of vapoluminescent platinum(II) double salts. J. Am. Chem. Soc. 2001, 123, 8414−8415. (3) White-Morris, R. L.; Olmstead, M. M.; Balch, A. L. Aurophilic interactions in cationic gold complexes with two isocyanide ligands. Polymorphic yellow and colorless forms of [(cyclohexyl isocyanide)2AuI](PF6) with distinct luminescence. J. Am. Chem. Soc. 2003, 125, 1033−1040. (4) Balch, A. L. Polymorphism and luminescent behavior of linear, two-coordinate gold(I) complexes. Gold Bull. 2004, 37, 45−50. (5) Contrast Agents I: Magnetic Resonance Imaging; Krause, W., Ed.; Topics in Current Chemistry, Vol. 221; Springer: Berlin, 2002. (6) Givaja, G.; Amo-Ochoa, P.; Gomez-Garcia, C. J.; Zamora, F. Electrical conductive coordination polymers. Chem. Soc. Rev. 2012, 41, 115−147. (7) Benelli, C.; Gatteschi, D. Introduction to Molecular Magnetism: From Transition Metals to Lanthanides; Wiley-VCH: Weinheim, Germany, 2015. (8) Cotton, F.; Murillo, C.; Walton, R. Multiple Bonds between Metal Atoms; Springer: New York, 2005. (9) Berry, J. F.; Cotton, F. A.; Lei, P.; Lu, T.; Murillo, C. A. Additional steps toward molecular scale wires: further study of Ni510/11+ chains embraced by polypyridylamide ligands. Inorg. Chem. 2003, 42, 3534−3539. (10) Bera, J. K.; Dunbar, K. R. Chain Compounds Based on Transition Metal Backbones: New Life for an Old Topic. Angew. Chem., Int. Ed. 2002, 41, 4453−4457. (11) Butler, W. M.; Enemark, J. H.; Parks, J.; Balch, A. L. Chelative addition of hydrazines to coordinated isocyanides. Structure of Chugaev’s red salt. Inorg. Chem. 1973, 12, 451−457. (12) Stork, J. R.; Olmstead, M. M.; Balch, A. L. Effects of Pt···Pt Bonding, Anions, Solvate Molecules, and Hydrogen Bonding on the Self-Association of Chugaev’s Cation, a Platinum Complex with a Chelating Carbene Ligand. Inorg. Chem. 2004, 43, 7508−7515. (13) Ouahab, L.; Enoki, T. Multiproperty Molecular Materials: TTFBased Conducting and Magnetic Molecular Materials. Eur. J. Inorg. Chem. 2004, 2004, 933−941. (14) Bray, J. W.; Interrante, L. V.; Jacobs, I.; Bonner, J. C. The Spin− Peierls Transition. In Extended Linear Chain Compounds; Miller, J. S., Ed.; Plenum Press: New York, 1983; Vol. 3, Chapter 7, pp 353−415. (15) Coronado, E.; Galán-Mascarós, J. R.; Gómez-García, C. J.; Martí-Gastaldo, C. Synthesis, Structure, and Magnetic Properties of the Oxalate-Based Bimetallic Ferromagnetic Chain {[K(18-crown6)][Mn(H2O)2Cr(ox)3]}∞ (18-crown-6 = C12H24O6, ox = C2O42−). Inorg. Chem. 2005, 44, 6197−6202. (16) Dahl, E. W.; Baddour, F. G.; Fiedler, S. R.; Hoffert, W. A.; Shores, M. P.; Yee, G. T.; Djukic, J.-P.; Bacon, J. W.; Rheingold, A. L.; Doerrer, L. H. Antiferromagnetic coupling across a tetrametallic unit through noncovalent interactions. Chem. Sci. 2012, 3, 602−609. (17) Baddour, F. G.; Fiedler, S. R.; Shores, M. P.; Golen, J. A.; Rheingold, A. L.; Doerrer, L. H. Heterobimetallic Lantern Complexes

7. SUMMARY AND FUTURE DIRECTIONS The homo- and heterobimetallic lantern complexes described in this Account can be expanded in many directions. Already Pt- as well as Pd-based heterobimetallic metalloligands, [M(SOCR)4]2−, have been used for the synthesis of highsymmetry trimetallic lanthanide complexes with tunable geometry via a chelating effect of the lantern units.29 The {PtM} complexes described may be bridged into chains with similar or new linkers. The new magnetic and electronic properties will add to the continuously growing field of 1D systems.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Linda H. Doerrer: 0000-0002-2437-6374 Notes

The authors declare no competing financial interest. Biographies Stephanie A. Beach is from Philadelphia, PA. She completed her undergraduate studies in chemistry at Rochester Institute of Technology in the laboratory of Hans Schmitthenner, working on the synthesis of multimodal-targeted MRI agents. She is currently a Ph.D. candidate at Boston University working in the laboratory of Linda Doerrer. She is a 2017/2018 Chateaubriand Fellow and is coordinator of the undergraduate WISE@Warren Program at BU. Linda H. Doerrer was born in Ithaca, NY, and grew up in Pensacola, FL. She completed her undergraduate studies in chemistry at Cornell University, working in the laboratories of Klaus Theopold and Frank I


Article

Accounts of Chemical Research That Couple Antiferromagnetically through Noncovalent Pt···Pt Interactions. Inorg. Chem. 2013, 52, 4926−4933. (18) Bellitto, C.; Dessy, G.; Fares, V. Synthesis, x-ray crystal structure, and chemical and physical properties of the new linear-chain mixed-valence complex (μ-iodo)tetrakis(dithioacetato)dinickel, Ni2(CH3CS2)4I, and X-ray crystal structure of the precursor tetrakis(dithioacetato)dinickel(II), Ni2(CH3CS2)4. Inorg. Chem. 1985, 24, 2815−2820. (19) Melson, G. A.; Greene, P. T.; Bryan, R. F. The crystal and molecular structure of ethanolatotetra-μ-thiobenzoato-dinickel(II), [Ni(C6H5COS)2]C2H5OH. Inorg. Chem. 1970, 9, 1116−1122. (20) Valan Amburose, C.; Deivaraj, T. C.; Lai, G. X.; Sampanthar, J. T.; Vittal, J. J. Syntheses and structures of Group 12 metal thioacetate anions, [M(SC{O}Me)nCl4−n]2− (n = 3 and 4) and [Cd2Cl2(SC{O}Me)4]2−. Inorg. Chim. Acta 2002, 332, 160−166. (21) Joshi, D. K.; Bhattacharya, S. Synthesis and structural characterization of a few thiocarboxylatonickel(II) complexes. Inorg. Chim. Acta 2014, 411, 119−127. (22) Nimthong-Roldan, A.; Guillet, J. L.; McNeely, J.; Ozumerzifon, T. J.; Shores, M. P.; Golen, J. A.; Rheingold, A. L.; Doerrer, L. H. Quasi-1D chains of dinickel lantern complexes and their magnetic properties. Dalton Trans. 2017, 46, 5546−5557. (23) Musaev, F. N.; Movsumov, É. M.; Mamedov, K. S.; Amiraslanov, I. R.; Magerramov, A. I.; Shnulin, A. N. Crystal and molecular structure of di(monothiobenzoate) dipyridinenickel (II). J. Struct. Chem. 1977, 18, 865−869. (24) Savant, V. V.; Iyengar, R. R.; Patel, C. C. Crystallographic data on bis(pyridine)nickel(II) monothiobenzoate. Indian J. Chem. 1970, 8, 195. (25) Cotton, F. A.; Dikarev, E. V.; Stiriba, S.-E. Studies of Dirhodium Tetra(trifluoroacetate). 3. Solid State Isomers of the Compound Rh2(O2CCF3)4(THF) Prepared by Sublimation. Inorg. Chem. 1999, 38, 4877−4881. (26) Baddour, F. G.; Fiedler, S. R.; Shores, M. P.; Bacon, J. W.; Golen, J. A.; Rheingold, A. L.; Doerrer, L. H. Pt···Pt vs Pt···S Contacts Between Pt-Containing Heterobimetallic Lantern Complexes. Inorg. Chem. 2013, 52, 13562−13575. (27) Guillet, J. L.; Bhowmick, I.; Shores, M. P.; Daley, C. J. A.; Gembicky, M.; Golen, J. A.; Rheingold, A. L.; Doerrer, L. H. Thiocyanate-Ligated Heterobimetallic {PtM} Lantern Complexes Including a Ferromagnetically Coupled 1D Coordination Polymer. Inorg. Chem. 2016, 55, 8099−8109. (28) Baddour, F. G.; Hyre, A. S.; Guillet, J. L.; Pascual, D.; Lopez-deLuzuriaga, J. M.; Alam, T. M.; Bacon, J. W.; Doerrer, L. H. Pt−Mg, Pt−Ca, and Pt−Zn Lantern Complexes and Metal-Only Donor− Acceptor Interactions. Inorg. Chem. 2017, 56, 452−469. (29) Soerensen, M. A.; Weihe, H.; Vinum, M. G.; Mortensen, J. S.; Doerrer, L. H.; Bendix, J. Imposing high-symmetry and tuneable geometry on lanthanide centres with chelating Pt and Pd metalloligands. Chem. Sci. 2017, 8, 3566−3575. (30) Baddour, F. G.; Hyre, A. S.; Guillet, J. L.; Pascual, D.; Lopez-deLuzuriaga, J. M.; Alam, T. M.; Bacon, J. W.; Doerrer, L. H. Pt−Mg, Pt−Ca, and Pt−Zn Lantern Complexes and Metal-Only Donor− Acceptor Interactions. Inorg. Chem. 2017, 56, 452−469. (31) Zuckerman, L. A.; Beach, S. A.; Hyre, A. S.; Rheingold, A. L.; Doerrer, L. H. Effects of substituted pyridine axial ligands on the electronic structure of heterobimetallic lantern complexes. 2017, in preparation.

J


Heterobimetallic Lantern Complexes and Their Novel Structural and

Recommend Documents