An Organolanthanide Building Block Approach to Single-Molecule

May 19, 2016 - (D) Arrhenius fit (red) of the χ″ data. Blocking of the magnetization was investigated through hysteresis measurements between −50...
0 downloads 9 Views 5MB Size
Download PDF
Article pubs.acs.org/accounts

An Organolanthanide Building Block Approach to Single-Molecule Magnets Katie L. M. Harriman and Muralee Murugesu* Department of Chemistry and Biomolecular Sciences, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada CONSPECTUS: Single-molecule magnets (SMMs) are highly sought after for their potential application in high-density information storage, spintronics, and quantum computing. SMMs exhibit slow relaxation of the magnetization of purely molecular origin, thus making them excellent candidates towards the aforementioned applications. In recent years, significant focus has been placed on the rare earth elements due to their large intrinsic magnetic anisotropy arising from the near degeneracy of the 4f orbitals. Traditionally, coordination chemistry has been utilized to fabricate lanthanide-based SMMs; however, heteroatomic donor atoms such as oxygen and nitrogen have limited orbital overlap with the shielded 4f orbitals. Thus, control over the anisotropic axis and induction of f−f interactions are limited, meaning that the performance of these systems can only extend so far. To this end, we have placed considerable attention on the development of novel SMMs whose donor atoms are conjugated hydrocarbons, thereby allowing us to perturb the crystal field of lanthanide ions through the use of an electronic π-cloud. This approach allows for fine tuning of the anisotropic axis of the molecule, allowing this method the potential to elicit SMMs capable of reaching much larger values for the two vital performance measurements of an SMM, the energy barrier to spin reversal (Ueff), and the blocking temperature of the magnetization (TB). In this Account, we describe our efforts to exploit the inherent anisotropy of the late 4f elements; namely, DyIII and ErIII, through the use of cyclooctatetraenyl (COT) metallocenes. With respect to the ErIII derivatives, we have seen record breaking success, reaching blocking temperatures as high as 14 K with frozen solution magnetometry. These results represent the first example of such a high TB being observed for a system with only a single spin center, formally known as a single-ion magnet (SIM). Our continued interrelationship between theoretical and experimental chemistry allows us to shed light on the mechanisms and electronic properties that govern the slow relaxation dynamics inherent to this unique set of SMMs, thus providing insight into the role by which both symmetry and crystal field effects contribute to the magnetic properties. As we look to the future success of such materials in practical devices, we must gain an understanding of how the 4f elements communicate magnetically, a subject upon which there is still limited knowledge. As such, we have described our work on coupling mononuclear metallocenes to generate new dinuclear SMMs. Through a building block approach, we have been able to gain access to new double,- triple- and quadruple-decker complexes that possess remarkable properties; exhibiting TB of 12 K and Ueff above 300 K. Our goal is to develop a fundamental platform from which to study 4f coupling, while maintaining and enhancing the strict axiality of the anisotropy of the 4f ions. This Account will present a successful strategy employed in the production of novel and high-performing SMMs, as well as a clear overview of the lessons learned throughout.

1. INTRODUCTION Single-molecule magnets (SMMs) exhibit slow relaxation of the magnetization of purely molecular origin, making them excellent candidates for electronics-based applications, such as high-density information storage, molecular spintronics, and quantum computing.1−3 Implementation of these types of materials into practical devices will rely heavily on increasing both their energy barrier to spin reversal (Ueff) and their magnetic blocking temperature (TB). Throughout the past decade, growing research efforts have been directed toward maximizing Ueff and TB through careful choice of magnetic ions and finite tailoring of their crystalline environments in order to increase single-ion anisotropy.4−11 The 4f elements represent excellent candidates for the targeted design of SMMs due to their large magnetic moments and intrinsic magnetic anisotropy; however, they are often plagued with significant © XXXX American Chemical Society

ground state (GS) quantum tunneling of the magnetization (QTM), which drastically reduces Ueff. This problem becomes even more difficult to overcome in non-Kramers ions (integer spin systems) where GS QTM is not formally forbidden, as it is with Kramers ions (half-integer spin systems). In order to circumvent these challenges, a design perspective has been developed toward the fabrication of 4f based SMMs. This strategy is based on the fact that single-ion anisotropy is greatly influenced by the symmetry of the coordination environment of the magnetic ion.9,12−14 It has been shown through careful molecular design that it is possible to use this strategy in order to manipulate the alignment of anisotropy axes, while Received: February 24, 2016

A

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 1. Synthetic Routes to Complexes Discusseda

a

(A) 1,4-Bis(trimethylsilyl)cyclooctatetraenyl, (B) cyclooctatetraenyl, and (C) cycloheptatrienyl.

tems, particularly with TbIII, have been extensively studied, one of which currently possess one of the largest energy barriers reported to date (Ueff = 938 K).19 Conversely, the first report of an organometallic SMM, in 2010 utilized cyclopentadienyl ligands to support a N-bridged dinuclear DyIII unit.20 This lanthanide−cyclopentadienyl framework has since been exploited to support fascinating pnictogen and chalcogen chemistry leading to novel SMMs.21,22 Inspired by these

simultaneously suppressing QTM effects, eliciting significant enhancement of Ueff and TB.9,15,16 With respect to this strategy, sandwich-type complexes have shown great promise as SMM architectures for harnessing 4f single-ion anisotropy.6,7,17,18 Interestingly, the first report of a lanthanide single-ion magnet (SIM) in 2003 was a sandwichlike structure, comprised of phthalocyanine ligands with nitrogen-based donating atoms.4 These phthalocyanine sysB

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research systems, we became interested in further understanding the role of the inherent anisotropy of lanthanides and how this property could be tuned and harnessed with an appropriate ligand field and, more importantly, how this could be exploited with aromatic organometallic ligands and the role by which the πelectron cloud would affect the overall magnetic properties. In this Account, we discuss our efforts in the field of molecular magnetism, explicitly within the realm of organolanthanide SMMs. We begin with a description of our initial work on the exploitation of 4f single-ion anisotropy within sandwich complexes, followed by our efforts in extending these model systems into higher nuclearity compounds.

2. HARNESSING SINGLE-ION ANISOTROPY When we first began, the magnetism of 4f bis-cyclooctatetraenylcompounds, affectionately referred to as metallocenes, was virtually unknown. We focused our attention toward the synthesis of sandwich complexes of DyIII with 1,4-bis(trimethylsilyl)cyclooctatetraenyl dianion (COT″).23 The substituted COT framework represented an excellent candidate to begin studying the effects of the ligand field on the slow relaxation dynamics since the trimethylsilyl (TMS) substituents provided both increased stability and solubility for the subsequent complexes. We utilized a synthetic strategy first outlined by Edelmann and co-workers,24,25 generating a homoleptic sandwich complex, [Dy(COT″)2Li(THF)(DME)], 1-Dy (Scheme 1A). Single-crystal X-ray diffraction (SCXRD) revealed that the lithium counterion was bound in an η2 fashion with one of the COT″ rings, resulting in a longer Dy−centroid bond, yielding an unsymmetrical compound. To probe the effects of the ligand field on the magnetic properties, investigation of the ac magnetic susceptibility of 1Dy was performed under zero applied dc field. With decreasing temperature, a frequency dependent signal was observed, indicating slow relaxation of the magnetization. Below 3.75 K, the behavior becomes frequency independent as a result of dominant QTM. In an effort to suppress the effects of QTM, the dynamic susceptibility was probed as a function of dc field. By variation of the applied dc field, multiple relaxation processes were revealed. These findings are rather rare, given that lanthanide-based systems often produce overlapping relaxation pathways as a consequence of their dense energy spectra produced from multiple Kramers doublets (KDs). However, in this case very distinct processes were observed, originating from a single spin center. A careful selection of dc fields (0, 100, 200, and 600 Oe) allowed us to individually probe the relaxation processes occurring (Figure 1). Analysis of the out-of-phase (χ″) component collected under a 100 Oe dc field resulted in a reduction of the QTM and the observation of a secondary process below 100 Hz and 4.25 K. Under a 200 Oe static field, an increase in peak intensity for the low frequency process, thereby promoting this process, and a subsequent decrease in intensity for the high frequency process were observed. Application of an optimal field of 600 Oe reveals uniform, frequency dependent peaks. Closer inspection of the peak shape reveals broadening, which is indicative of mixing of both relaxation processes. Nonetheless, a Ueff can be extracted, under the aforementioned optimal dc field, through application of the Arrhenius law (τ = τ0 exp[Ueff/(kBT)]), assuming a single relaxation regime, to yield Ueff = 43 K. The increase in Ueff from 18 to 43 K highlights the significant role of QTM in lanthanide SMMs and their relaxation dynamics.

Figure 1. Frequency dependence of χ″ for 1-Dy in the indicated temperature range under applied dc fields of 0, 100, 200, and 600 Oe.

The investigation of 1-Dy proved to be fruitful in the elucidation of different relaxation pathways originating from a single-spin center; however; the most enticing part of this compound lies within the ability to tune the magnetic properties through peripheral structural modifications, that is, finite tailoring of ligand substituents, and alkali metal ion interactions. The gradual deviations from ideal symmetry, which result as a consequence of these external moieties, diminish the axiality and are responsible for the prevalent QTM observed in 1-Dy.26 In high symmetry crystal fields, the splitting of the GS into ±mJ doublets often yields low lying doublets with large mJ values, thus leading to largely axial anisotropy. This feature is often attributed to large Ueff values observed in lanthanide SIMs.27 With this in mind, we focused C

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

with very strong slow relaxation dynamics. Accordingly, we turned our attention toward ErIII in an effort to promote axial anisotropy over easy-plane type anisotropy. The χ″ data of 2-Er demonstrated its superior anisotropy generating a single relaxation process under zero dc field, producing a Ueff (187 K) nearly eight times larger in magnitude than that of 2-Dy (Figure 3). The discrepancy in the barriers is likely attributed to the multiple overlapping relaxation processes occurring in 2Dy.

our attention on systems that exhibit high rotational symmetry, generating two points for increasing overall symmetry, (1) separation of the alkali metal cation from the sandwich complex and (2) replacement of the TMS substituents on the COT framework. Focusing on the isolation of the alkali metal cation, we successfully produced [Li(DME)3][Ln(COT″)2], Ln = Dy (2-Dy), Er (2-Er) (Scheme 1A).12 Interestingly, the analogous ErIII complex displays remarkable properties (vide inf ra).28 SCXRD of 2-Dy afforded a discrete sandwich complex. The coordination of three molecules of DME to the lithium ion limits possible close contacts with the paramagnetic species, producing a Li−CCOT″ distance nearly double that observed for 1-Dy. A Dy−COTcentroid distance of 1.90 Å is achieved, leading to equivalent electron donation from each COT″ ring. Similarly to 1-Dy, 2-Dy exhibited frequency dependent χ″ signals under zero applied field in the temperature range of 1.8−9 K, demonstrating shifting of peak maxima above 4.5 K (Figure 2). While a full peak was observed below 4.5 K, the

Figure 3. (A) Molecular structure of 2-Er. Color code: green (ErIII), teal (Si), and gray (C); H atoms omitted for clarity. (B) Magnetic hysteresis between 1.8 and 10 K, collected at a fixed sweep rate of 22 Oe·s−1. (C) χ″ under a zero-applied dc field at indicated temperatures. (D) Arrhenius fit (red) of the χ″ data. Figure 2. (A) Molecular structure of 2-Dy. Color code: red (DyIII), teal (Si), and gray (C); H atoms omitted for clarity. (B) Temperature dependence of χ″, frequencies in the range of 10−1500 Hz, and (C) frequency dependence of χ″, temperatures in the range 2−9 K, collected under zero applied dc field.

Blocking of the magnetization was investigated through hysteresis measurements between −50 to 50 kOe at a fixed sweep rate of 22 Oe·s−1. A large coercive field of 6250 Oe was observed at 1.8 K and Hdc = 0 Oe for 2-Er. Upon increasing temperature, the coercivity gradually lessens, and above 8 K, a phonon bottleneck-type hysteresis is observed. Above 9 K, small openings observed at higher fields are no longer detectable, rendering a TB of 8 K. The steps observed likely arise from QTM; specifically, the step at ±4500 Oe presumably occurs via a metastable potential well. Whereas the step at zero field is attributed to GS tunneling. As discussed previously, GS tunneling is formally forbidden for Kramers ions, ErIII (S = 3/ 2); however, when the system is subjected to hyperfine and dipolar coupling, it may result in non-negligible QTM.14 To improve upon the magnetic properties of 2-Er via tailoring the molecular symmetry, we synthesized sandwich complexes devoid of peripheral substituents, [K(18-crown6)][Ln(COT)2], Ln = ErIII (3-Er), DyIII (3-Dy) (Scheme 1B).29,30 SCXRD of 3-Er and 3-Dy revealed that the potassium counterion asymmetrically interacts with the COT ring. This causes inequivalent electron donation to the LnIII ion, as observed in the Ln−Ccentroid distances of 3-Er (1.85 vs 1.89 Å) and 3-Dy (1.89 vs 1.91 Å). As anticipated, the substituent free COT permits the preservation of 8-fold symmetry in the SCXRD structures of 3-Er (D8d) and 3-Dy (D8h) collected at 200(2) K.

frequency independent nature of the process suggests a dominant QTM regime. In the thermally activated regime a Ueff = 25 K and τ0 = 6 × 10−6 s can be extracted. A Ueff of this magnitude is not surprising given the significant QTM present below 4.5 K. Field dependent measurements were completed elucidating multiple relaxation processes for 2-Dy, akin to 1Dy. This was intriguing given that the structural distortions in 1-Dy did not appear to account for the observed behavior (i.e., QTM, multiple pathways, low Ueff); thus, the properties could be attributed to the effects of the ligand field acting on the DyIII center. Ab initio calculations were completed in order to determine the orientation of the anisotropic axis of 2-Dy. The presence of the silyl groups on COT″ proved dominant over the pseudo-C8 axis of the COT″ rings, resulting in diagonal anisotropic axes. Investigation of the symmetrized derivative, [Dy(COT)2]−, surprisingly revealed a GS doublet of mJ = 1/2, with a significant transverse component. Additionally, the highest energy doublet (E = 1074.2 cm−1) was the most axial in nature, (gz = 19.9927), suggesting that replacement of DyIII with ErIII within the same ligand field would produce a system D

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Figure 4. (A) Molecular structure and calculated GS electronic density (red surface) of 3-Er and 3-Dy. Dashed lines show the calculated orientation of the magnetic axis on LnIII in the ground (1) and the first excited (2) KD. (B) χ″ collected under zero field and magnetization hysteresis collected at fixed sweep rate of 35 Oe·s−1 at indicated temperatures. (C) The calculated magnetization blocking barriers. The black lines represent the KD as a function of their magnetic moment along the axis connecting the centers of the COT rings. The dashed lines correspond to relaxation processes; QTM (green); Orbach (blue); the most probable path for magnetic relaxation (red).

highly axial GS (gx,y-tensors ≈ 10−6), and well separated lowlying excited states (Figure 4). Conversely, 3-Dy exhibits a more dense low-lying spectrum, where the ground mj state is not the largest in magnitude, leading to noncollinear axes of the GS and the first excited state. Additionally, the GS gx,y-tensors are greater in magnitude (∼10−1), demonstrating the reduced axiality of 3-Dy. Energy spectra of the associated KDs were determined in order to define the magnetization blocking barriers. The barriers are the shortest possible pathways, comprised of energy levels with the largest transverse magnetic moments. The transverse magnetic moments of the ground and first excited state for 3-Er are minimal (10−6μB and 10−4μB), owing to the collinear anisotropic axes. The off-diagonal matrix elements for the ground and first excited state exhibited a very small difference between the components of opposite magnetization, resulting in a suppressed Orbach mechanism. Due to this suppression, the relaxation for 3-Er must occur through a thermally activated process via the second excited state. With 3Dy, non-negligible transverse magnetization (10−1μB) in the GS allows for fast QTM; this same tunneling also occurs in the

These compounds are characterized by very different magnetic properties, such that 3-Er exhibits strong magnetic blocking. Below 12 K, a χ″ signal cannot be observed because the relaxation time is longer than 10 s. Surprisingly, 3-Dy showed comparatively weak blocking. This contrast is exemplified by the zero field behavior producing a Ueff of 286 and 11 K for 3-Er and 3-Dy, respectively. Magnetic hysteresis measurements at a sweep rate of 35 Oe·s−1 depicted an absence of coercivity and remnant magnetization for 3-Dy, even at 1.8 K. The coercive field for 3-Er at 1.8 K (7000 Oe) is even larger than that obtained for 2-Er. This was the first time that such a large coercive field had been observed for a mononuclear system.4−7,31,32 Thus, replacement of the TMS substituents to afford greater molecular symmetry provided a potentially viable approach toward fine-tuning single-ion anisotropy. Ab initio calculations of the CASSCF/RASSI/SINGLE_ANISO type were completed via MOLCAS to gain insight on the electronic and magnetic structure of 3-Er and 3-Dy.33 Calculated anisotropic axes for the ground and first excited state reveal that 3-Er has almost collinear axes arising from the E

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research excited state (10−1μB). For 3-Er, the axial parameters (B02, B04, B06) for the multiplet-specific crystal field energies were the largest, and the GS is primarily composed of a single Jprojection, ±15/2. These results are significantly different in 3Dy, where the GS has contributions from different Jprojections, with the largest contribution from ±9/2. The resulting spectra appear to be inversely related, a consequence of the axial parameters, B02 and B04, which are opposite in sign. These studies provide a unique perspective on harnessing single-ion anisotropy by exploiting the symmetry of carefully chosen organometallic ligands and metal ions.

3. THE BUILDING BLOCK APPROACH If larger barriers are required for technological applications, then single-ion anisotropy combined with the limited spin of a single metal center will not be sufficient to commercialize SMMs. Thus, inducing significant interactions between highly anisotropic lanthanide centers is vital. To this end, we have directed efforts toward utilizing the Ln-COT monomers in the design of new dinuclear SMMs, because dimers represent the most fundamental unit for studying magnetic exchange interactions. The COT scaffold represents a unique bridging motif with which to probe magnetic communication, thereby taking advantage of the π-electron cloud, which may facilitate non-negligible interactions between shielded 4f orbitals. To this end, we prepared triple-decker compounds [Ln2(COT″)3], Ln = DyIII (4-Dy), ErIII (4-Er) of monomers 2-Dy and 2-Er (Scheme 1A).12,34−36 SCXRD depicted unequal electronic donation from the COT″ rings of 4-Dy and 4-Er, because the central COT″ provides electronic donation to both LnIII sites, vide inf ra. Additionally, the three COT″ rings form a near parallel arrangement, giving rise to a tilt angle of 1.86° for 4-Dy, demonstrating the preservation of rotational symmetry in this molecule. The electronic structure and nature of bonding of 4-Dy was investigated using DFT calculations utilizing the geometry obtained from SCXRD. Optimized wave functions for the singlet state found the terminal COT″ rings to be sufficiently polarized with spin density values of −0.39 au. The bridging and terminal COT″ rings were also examined for charge density distributions, yielding values of −0.12 and −0.30 au, respectively, indicating that the terminal COT″ rings donate 1.70e− to their respective DyIII ion and the central ring effectively donates 0.94e− to each DyIII ion. Interestingly, a Dy−Dy covalent interaction was determined with a Mayer bond order of 0.04. While this value is relatively small, it is surprising to observe such interaction between 4f elements. In order to investigate this interaction magnetically, dc magnetic susceptibility of the isotropic derivative, GdIII−GdIII, was fit through application of the Van Vleck equation to Kambe’s vector coupling method using the spin Hamiltonian, H = −JSa· Sb, Sa = Sb = 7/2. Best fit parameters of J = −0.448 cm−1 and g = 2.00 were obtained; the small coupling constant is not surprising given the weak nature of lanthanide−lanthanide interactions. Magnetic hysteresis measurements were completed at a fixed sweep rate of 22 Oe·s−1. No hysteretic behavior was observed for 4-Dy. Comparatively, compound 4-Er displays hysteresis with s-shaped loops and coercivity at Hdc = 0 Oe (Figure 5). A TB of 12 K was achieved, displaying a 4 K increase over 2-Er. Frozen solution measurements of 4-Er, resulted in an increase in TB to 14 K, likely resulting from the dilution of intermolecular interactions. Furthermore, the frequency

Figure 5. (A) Molecular structure of 4-Er. Color code: green (ErIII), teal (Si), and gray (C); H atoms omitted for clarity. (B) χ″ under a zero-applied dc field at indicated temperatures. (C) Magnetic hysteresis completed at an average sweep rate of 22 Oe·s−1 between 1.8 and 13 K. (D) Solid state and solution (4 mM cyclopentane) magnetic hysteresis at 1.8 K.

dependent χ″ data of 4-Er revealed a Ueff of 323 K under zero field in the temperature range of 26−14 K; this barrier is significantly larger than that obtained for 2-Er (Ueff = 187 K). Not unlike 2-Dy, 4-Dy was plagued with significant QTM, drastically reducing Ueff, demonstrating the sensitivity of 4f ions to certain ligand fields. Under zero applied field, frequency dependent studies displayed a peak below 5 K (Ueff = 9 K); combined with the absence of signal in the χ″ vs T, suggests that QTM is significant in this system. Thus, the application of a small dc field (600 Oe) allows for the removal of degeneracy in the mj states, preventing QTM and allowing for the extraction of Ueff. Shifting of the peak maxima was observed under these conditions yielding Ueff = 24 K and τo = 3.6 × 10−6 s, which is significantly smaller than that obtained for 4-Er (Figure 6). The reduced performance of 4-Dy can be explained via computational studies, which revealed that the anisotropic axes of the DyIII ions deviated from the axis imposed by the COT″ rings, suggesting the strong influence of the peripheral silyl groups; as previously observed in 2-Dy. The crystal field parameters (B0n) obtained are positive, demonstrating that the equatorial component of the ligand field is stronger than the axial one for 4-Dy. Having understood the critical role of symmetry on the overall SMM properties, we synthesized [K2Er2(COT)4(THF)4] (5-Er), a quadruple-decker compound containing the symmetrized monomer, 3-Er (Scheme 1B). The two individual [Er(COT)2]− monomers are bridged by a potassium ion (K2), with a second potassium ion (K1) bound to one of the monomer units.36 SCXRD reveals that Er2 experiences uneven electronic donation from COTa and COTb, with Er2−CCOT average distances of 2.58 and 2.66 Å, respectively (Figure 7). Comparatively, Er1 experiences more equivalent electronic donation with average Er1−CCOT bond lengths of 2.61 Å, as a consequence of the η8-bound K ions. F

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

process. Given the distance between the ErIII ions, one may expect that the two ions would relax independently, producing two peaks in the χ″ data. However, due to the frequency limitations of the instrument, independent processes were only observed in the temperature dependent χ″ data between 32 and 1.8 K. Energy barriers for these processes were determined to be 170 K (7−12 K) and 293 K (19−32 K). The computationally derived ground and excited state anisotropic axes for 5Er are axial in nature, which is in agreement with 3-Er. This arises from the negative sign of the crystal field parameters, B02 and B04, causing stabilization of the ±15/2 state. Overall, the higher Ueff observed for 5-Er can be attributed to the increased symmetry of the Er1 monomer.

4. PROBING f−f ORBITAL INTERACTIONS While considerable efforts have been employed in order to understand the fundamental interactions between 4f ions, there remains limited knowledge of how the 4f ions communicate magnetically. The high blocking temperatures that are recorded for lanthanide-based systems are largely a result of single-ion properties. Recently, we have begun to explore the effects of magnetic communication between 4f ions as a consequence of charge and electronic density within the bridging moiety. We found success in inducing covalent interactions within the [Ln2(COT″)3] framework (4-Dy and 4-Er); as such we used the π-electron cloud of aromatics as a means to perturb the shielded 4f orbitals. To this end, we utilized six and seven membered rings to vary the charge and electronic density within the bridging moiety. Both COT and arene provide bridging moieties with 10π electrons, adopting formal charges of 2− and 4−, respectively. Our work with these bridges shows that a weak, yet non-negligible interaction is observed with coupling constants of −0.644 and −0.488 cm−1, respectively.12,37 However, if the bridging ligand is a seven-membered ring, it must adopt a formal charge of 3− in order to maintain Hückel aromaticity (10π electrons), giving rise to the cycloheptatrienyl trianion. Complexes of η-cycloheptatrienyl are rare with 4f ions; there exists only a single example.38 Inspired by this example, we utilized bulky ancillary ligands, bis(trimethylsilyl)amido, to promote the facile formation of the trianion, yielding [KEr2(η7-C7H7)(N(SiMe3)2)4], 6-Er (Scheme 1C). SCXRD revealed the coordination of the potassium counterion to the amido nitrogens of Er2 (Figure 8), making this molecule unsymmetrical. An intramolecular Er−Er distance of 3.96 Å is slightly shorter than what was obtained in 4-Er (4.07 Å), which may provide an efficient pathway for magnetic exchange. The χ″ data of 6-Er revealed the presence of significant QTM. Application of an 800 Oe optimal dc field minimized these effects, producing frequency dependent behavior in the temperature range 3−7 K (Figure 8). Upon fitting the data, a Ueff of 58 K and τo of 2.9 × 10−8 s were obtained. The observed relaxation dynamics of 6-Er are likely attributed to single-ion behavior, because below 3 K a second relaxation process begins to emerge. Investigation of the isotropic derivative revealed a magnetic coupling constant of J = −0.134 cm−1, using the spin Hamiltonian H = −JSa·Sb, Sa = Sb = 7/2. The presence of the potassium cation reduces equivalent electronic donation to each ErIII ion. Collectively, the weak magnetic coupling constant, noncentrosymmetric structure, and presence of two independent relaxation processes strongly suggest that the two ErIII ions are independently relaxing. Nonetheless, this compound represents the first example of a cycloheptatrienyl

Figure 6. (A) Molecular structure of 4-Dy. Color code: red (DyIII), teal (Si), and gray (C); H atoms omitted for clarity. Frequency dependence of χ″ between 2.5 and 9 K under dc fields of (B) 0 Oe and (C) 600 Oe.

Figure 7. (A) Molecular structure of 5-Er. Color code: green (ErIII), purple (K), red (O), and gray (C); H atoms omitted for clarity. (B) Magnetic hysteresis at temperatures 1.8−13 K, measurements completed at an average sweep rate of 18 Oe·s−1. (C) χ″ under a zero-applied dc field at indicated temperatures.

The magnetic interaction was quantified with the isotropic derivative, using the spin Hamiltonian H = -JSa·Sb, Sa = Sb = 7/ 2, yielding J = −0.007 cm−1 and g = 1.99. Stronger coupling was obtained for the isotropic analogue of 4-Dy and 4-Er; however, this likely arises from the separation of the magnetic ions (8.82 Å vs. 4.11 Å). Hysteresis measurements at a fixed sweep rate of 18 Oe·s−1 elicited a butterfly like hysteresis with an absence of coercivity at Hdc = 0 Oe; above 12 K, openings are no longer observed (Figure 7). The relaxation mechanisms responsible for this behavior were investigated with ac susceptibility measurements, revealing frequency dependent behavior between 32 and 13 K, producing a Ueff = 306 K and τo = 5.0 × 10−9 s for a single G

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

energy barriers. Hence, the ability to promote strong magnetic coupling is essential, because the current properties achieved from single metal centers, namely, single-ion anisotropy and spin, are not sufficient to commercialize SMM technology. Ideally, stacking monomers in a linear chain-like array should yield a material with axial orientation of anisotropy, thus resulting in a large net vector. This type of arrangement would theoretically give rise to even larger uniaxial anisotropies, thereby leading to increased Ueff values and potentially larger TB values (Figure 9). While theoretically the chain-like

Figure 8. (A) Molecular structure of 6-Er. Color code: green (ErIII), purple (K), teal (Si), blue (N), and gray (C); H atoms omitted for clarity. (B) Frequency dependence of χ″ for 6-Er between 3.75 and 7 K, under 800 Oe applied dc field.

Figure 9. Hypothetical chain-like arrangement of Ln(COT) 2 monomers. The axial anisotropy of each monomer is depicted by blue vectors, and the vector addition of the monomeric axial anisotropies yields the purple vector.

based SMM. It is vital that we continue to look for new systems that provide unique exchange pathways for magnetic interactions.

arrangements of these sandwich compounds are supported by the planarity of the bridging COT ligands, the synthetic feasibility of such materials remains an exciting and challenging goal for the future of lanthanide−COT based SMMs. Through taking advantage of their ability to be uniformly attached to a variety of surfaces via π−π stacking, these aromatic organometallic SMMs offer a truly unparalleled direction for the incorporation of molecular based magnetic materials into devices.

5. CONCLUSIONS AND OUTLOOK In this Account, we have highlighted the unique role of aromatic COT and related ligands in the elucidation of late 4f ion single-molecule magnetism. Although all of the systems presented are sensitive to aerial oxidation and moisture; this rare set of SMMs has yielded remarkable and in some cases record-breaking properties. Through utilizing a rational design approach and considering key symmetry elements, the ligand field generated by COT and its derivatives elegantly promotes uniaxial anisotropy and coupling of 4f ions. 2-Er represents just one example discussed in this Account displaying very strict axiality in the GS, eliciting a TB of 8 K, and extension to the triple-decker (4-Er) produced an impressive TB of 12 K. Such magnetic performance is extremely rare for nonradical based SMMs. The impact of employing such ligand systems to promote Ln−Ln interactions was noted by a unique Dy−Dy bond observed in 4-Dy. Only recently have we begun to better understand the potential and tunability in employing π-electron clouds toward garnering desirable Ln−Ln interactions. With this knowledge, we have generated the first example of a cycloheptatrienyl based SMM (6-Er). Thus, we propose that these design strategies be continuously explored and improved in order to further probe the potentially groundbreaking magnetic properties possible for f−f coupled systems. Perhaps most critically, the future development of improved materials for high-density electronics will require even larger

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. Biographies Katie L. M. Harriman obtained her B.Sc. (Hons) in Biopharmaceutical Sciences from the University of Ottawa in 2015. During her undergraduate studies, she undertook two research assistantships, first under the supervision of Prof. Jaclyn Brusso and second under the supervision of Prof. Muralee Murugesu. She is currently pursuing her M.Sc. in the group of Prof. Muralee Murugesu, working on 4f organometallic single-molecule magnets. Muralee Murugesu received his Ph.D. from University of Karlsruhe in 2002 under the guidance of Prof. A. K. Powell. He undertook H

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

(13) Petit, S.; Pilet, G.; Luneau, D.; Chibotaru, L. F.; Ungur, L. A dinuclear cobalt(II) complex of calix 8 arenes exibiting strong magnetic anisotropy. Dalton Trans. 2007, 4582−4588. (14) Habib, F.; Lin, P.-H.; Long, J.; Korobkov, I.; Wernsdorfer, W.; Murugesu, M. The use of magnetic dilution to elucidate the slow magnetic relaxation effects of a Dy-2 single-molecule magnet. J. Am. Chem. Soc. 2011, 133, 8830−8833. (15) Chilton, N. F. Design criteria for high-temperature singlemolecule magnets. Inorg. Chem. 2015, 54, 2097−2099. (16) Gregson, M.; Chilton, N. F.; Ariciu, A.-M.; Tuna, F.; Crowe, I. F.; Lewis, W.; Blake, A. J.; Collison, D.; McInnes, E. J. L.; Winpenny, R. E. P.; Liddle, S. T. A monometallic lanthanide bis(methanediide) single molecule magnet with a large energy barrier and complex spin relaxation behaviour. Chem. Sci. 2016, 7, 155−165. (17) Ishikawa, N. Phthalocyanine-based magnets. Struct. Bonding 2010, 135, 211−228. (18) Ishikawa, N. Single molecule magnet with single lanthanide ion. Polyhedron 2007, 26, 2147−2153. (19) Ganivet, C. R.; Ballesteros, B.; de la Torre, G.; Clemente-Juan, J. M.; Coronado, E.; Torres, T. Influence of peripheral substitution on the magnetic behavior of single-ion magnets based on homo- and heteroleptic TbIII bis(phthalocyaninate). Chem. - Eur. J. 2013, 19, 1457−1465. (20) Layfield, R. A.; McDouall, J. J. W.; Sulway, S. A.; Tuna, F.; Collison, D.; Winpenny, R. E. P. Influence of the N-bridging ligand on magnetic relaxation in an organometallic dysprosium single-molecule magnet. Chem. - Eur. J. 2010, 16, 4442−4446. (21) Pugh, T.; Tuna, F.; Ungur, L.; Collison, D.; McInnes, E. J. L.; Chibotaru, L. F.; Layfield, R. A. Influencing the properties of dysprosium single-molecule magnets with phosphorus donor ligands. Nat. Commun. 2015, 6, 7492. (22) Pugh, T.; Vieru, V.; Chibotaru, L. F.; Layfield, R. A. Magnetostructural correlations in arsenic- and selenium-ligated dysprosium single-molecule magnets. Chem. Sci. 2016, 7, 2128−2137. (23) Jeletic, M.; Lin, P.-H.; Le Roy, J. J.; Korobkov, I.; Gorelsky, S. I.; Murugesu, M. An organometallic sandwich lanthanide single-ion magnet with an unusual multiple relaxation mechanism. J. Am. Chem. Soc. 2011, 133, 19286−19289. (24) Poremba, P.; Reissmann, U.; Noltemeyer, M.; Schmidt, H. G.; Bruser, W.; Edelmann, F. T. Cyclooctatetraenyl complexes of the early transition metals and lanthanides VIII. Anionic sandwich complexes of the lanthanides containing silylated cyclooctatetraenyl ligands. J. Organomet. Chem. 1997, 544, 1−6. (25) Lorenz, V.; Edelmann, A.; Blaurock, S.; Freise, F.; Edelmann, F. T. A surprising solvent effect on the crystal structure of an anionic lanthanide sandwich complex. Organometallics 2007, 26, 6681−6683. (26) Feltham, H. L. C.; Lan, Y.; Kloewer, F.; Ungur, L.; Chibotaru, L. F.; Powell, A. K.; Brooker, S. A Non-sandwiched macrocyclic monolanthanide single-molecule magnet: The key role of axiality. Chem. - Eur. J. 2011, 17, 4362−4365. (27) Blagg, R. J.; Ungur, L.; Tuna, F.; Speak, J.; Comar, P.; Collison, D.; Wernsdorfer, W.; McInnes, E. J. L.; Chibotaru, L. F.; Winpenny, R. E. P. Magnetic relaxation pathways in lanthanide single-molecule magnets. Nat. Chem. 2013, 5, 673−678. (28) Le Roy, J. J.; Korobkov, I.; Murugesu, M. A sandwich complex with axial symmetry for harnessing the anisotropy in a prolate erbium(III) ion. Chem. Commun. 2014, 50, 1602−1604. (29) Ungur, L.; Le Roy, J. J.; Korobkov, I.; Murugesu, M.; Chibotaru, L. F. Fine-tuning the local symmetry to attain record blocking temperature and magnetic remanence in a single-ion magnet. Angew. Chem., Int. Ed. 2014, 53, 4413−4417. (30) Hodgson, K. O.; Mares, F.; Starks, D. F.; Streitwieser, A. Lanthanide(III) complexes wtih cyclooctatetraene dianion- synthetic chemistry, characterization, and physical-properties. J. Am. Chem. Soc. 1973, 95, 8650−8658. (31) Ishikawa, N.; Sugita, M.; Ishikawa, T.; Koshihara, S.; Kaizu, Y. Mononuclear lanthanide complexes with a long magnetization relaxation time at high temperatures: A new category of magnets at the single-molecular level. J. Phys. Chem. B 2004, 108, 11265−11271.

postdoctoral stays at the University of Florida (2003−2005) with Prof. G. Christou and jointly at the University of California, Berkeley, and the University of California, San Francisco, under the supervision of Prof. J. R. Long and the Nobel Laureate Prof. S. Prusiner (2005− 2006). In 2006, he joined the University of Ottawa as an assistant professor, then he became an associate professor in 2011, and in 2015 he became a full professor. He is currently University Research Chair in nanotechnology, and his research focuses on the design and development of new synthetic methods towards novel nanoscale materials.

■

ACKNOWLEDGMENTS We acknowledge the University of Ottawa, Canadian Foundation for Innovation, and NSERC for their financial support. The authors gratefully acknowledge Dr. Jennifer J. Le Roy for her conceptual and practical contributions to the work presented, as well as Rebecca J. Holmberg who is thanked for her many comments in the preparation of this manuscript. The authors also thank all collaborators involved in this work.

■

REFERENCES

(1) Ardavan, A.; Rival, O.; Morton, J. J. L.; Blundell, S. J.; Tyryshkin, A. M.; Timco, G. A.; Winpenny, R. E. P. Will spin-relaxation times in molecular magnets permit quantum information processing? Phys. Rev. Lett. 2007, 98, 057201. (2) Aromi, G.; Aguila, D.; Gamez, P.; Luis, F.; Roubeau, O. Design of magnetic coordination complexes for quantum computing. Chem. Soc. Rev. 2012, 41, 537−546. (3) Bogani, L.; Wernsdorfer, W. Molecular spintronics using singlemolecule magnets. Nat. Mater. 2008, 7, 179−186. (4) Ishikawa, N.; Sugita, M.; Ishikawa, T.; Koshihara, S.; Kaizu, Y. Lanthanide double-decker complexes functioning as magnets at the single-molecular level. J. Am. Chem. Soc. 2003, 125, 8694−8695. (5) Takamatsu, S.; Ishikawa, T.; Koshihara, S.-y.; Ishikawa, N. Significant increase of the barrier energy for magnetization reversal of a single-4f-ionic single-molecule magnet by a longitudinal contraction of the coordination space. Inorg. Chem. 2007, 46, 7250−7252. (6) Jiang, S.-D.; Liu, S.-S.; Zhou, L.-N.; Wang, B.-W.; Wang, Z.-M.; Gao, S. Series of Lanthanide Organometallic Single-Ion Magnets. Inorg. Chem. 2012, 51, 3079−3087. (7) Jiang, S.-D.; Wang, B.-W.; Sun, H.-L.; Wang, Z.-M.; Gao, S. An organometallic single-ion magnet. J. Am. Chem. Soc. 2011, 133, 4730− 4733. (8) Zhang, P.; Zhang, L.; Wang, C.; Xue, S. F.; Lin, S. Y.; Tang, J. K. Equatorially coordinated lanthanide single ion magnets. J. Am. Chem. Soc. 2014, 136, 4484−4487. (9) Cucinotta, G.; Perfetti, M.; Luzon, J.; Etienne, M.; Car, P.-E.; Caneschi, A.; Calvez, G.; Bernot, K.; Sessoli, R. Magnetic anisotropy in a dysprosium/DOTA single-molecule magnet: beyond simple magneto-structural correlations. Angew. Chem., Int. Ed. 2012, 51, 1606−1610. (10) Zadrozny, J. M.; Xiao, D. J.; Atanasov, M.; Long, G. J.; Grandjean, F.; Neese, F.; Long, J. R. Magnetic blocking in a linear iron(I) complex. Nat. Chem. 2013, 5, 577−581. (11) Poulten, R. C.; Page, M. J.; Algarra, A. G.; Le Roy, J. J.; Lopez, I.; Carter, E.; Llobet, A.; Macgregor, S. A.; Mahon, M. F.; Murphy, D. M.; Murugesu, M.; Whittlesey, M. K. Synthesis, electronic structure, and magnetism of [Ni(6-Mes)2]+: A two-coordinate nickel(I) complex stabilized by bulky N-heterocyclic carbenes. J. Am. Chem. Soc. 2013, 135, 13640−13643. (12) Le Roy, J. J.; Jeletic, M.; Gorelsky, S. I.; Korobkov, I.; Ungur, L.; Chibotaru, L. F.; Murugesu, M. An organometallic building block approach to produce a multidecker 4f single-molecule magnet. J. Am. Chem. Soc. 2013, 135, 3502−3510. I

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research (32) AlDamen, M. A.; Cardona-Serra, S.; Clemente-Juan, J. M.; Coronado, E.; Gaita-Arino, A.; Marti-Gastaldo, C.; Luis, F.; Montero, O. Mononuclear lanthanide single molecule magnets based on the polyoxometalates [Ln(W5O18)2]9‑ and [Ln(β2-SiW11O39)2]13‑ (LnIII = Tb, Dy, Ho, Er, Tm, and Yb). Inorg. Chem. 2009, 48, 3467−3479. (33) Aquilante, F.; De Vico, L.; Ferre, N.; Ghigo, G.; Malmqvist, P.A.; Neogrady, P.; Pedersen, T. B.; Pitonak, M.; Reiher, M.; Roos, B. O.; Serrano-Andres, L.; Urban, M.; Veryazov, V.; Lindh, R. MOLCAS 7: The next generation. J. Comput. Chem. 2010, 31, 224−247. (34) Poremba, P.; Edelmann, F. T. Cyclooctatetraenyl complexes of the early transition metals and lanthanides X. The first organometallic triple decker sandwich complexes of the lanthanides. J. Organomet. Chem. 1998, 553, 393−395. (35) Lorenz, V.; Blaurock, S.; Hrib, C. G.; Edelmann, F. T. The first linear, homoleptic triple-decker sandwich complex of an f-element: A molecular model for organolanthanide nanowires. Organometallics 2010, 29, 4787−4789. (36) Le Roy, J. J.; Ungur, L.; Korobkov, I.; Chibotaru, L. F.; Murugesu, M. Coupling strategies to enhance single-molecule magnet properties of erbium-cyclooctatetraenyl complexes. J. Am. Chem. Soc. 2014, 136, 8003−8010. (37) Huang, W.; Le Roy, J. J.; Khan, S. I.; Ungur, L.; Murugesu, M.; Diaconescu, P. L. Tetraanionic biphenyl lanthanide complexes as single-molecule magnets. Inorg. Chem. 2015, 54, 2374−2382. (38) Arliguie, T.; Lance, M.; Nierlich, M.; Ephritikhine, M. Inverse cycloheptatrienyl sandwich complexes of uranium and neodymium. J. Chem. Soc., Dalton Trans. 1997, 2501−2504.

J

DOI: 10.1021/acs.accounts.6b00100 Acc. Chem. Res. XXXX, XXX, XXX−XXX