Zi-Yi
Du
a,
Ling
Zhang
a,
Bao-Ying
Wang
b,
Sui-Jun
Liu
*c,
Bo
Huang
b,
Cai-Ming
Liu
*d and
Wei-Xiong
Zhang
*b
aCollege of Chemistry and Chemical Engineering, Gannan Normal University, Ganzhou 341000, China
bSchool of Chemistry, Sun Yat-Sen University, Guangzhou 510275, China. E-mail: zhangwx6@mail.sysu.edu.cn
cSchool of Metallurgy and Chemical Engineering, Jiangxi University of Science and Technology, Ganzhou 341000, China. E-mail: liusuijun147@163.com
dBeijing National Laboratory for Molecular Sciences, Center for Molecular Science, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China. E-mail: cmliu@iccas.ac.cn
First published on 17th January 2017
The hydrothermal reaction of Mn(II) or Co(II) ions with 2-carboxyethyl(phenyl)phosphinic acid (H2L) afforded Mn(II) and Co(II) carboxylate–phosphinates containing μ3-OH− as a co-ligand, namely [M3(L)2(OH)2] (M = Mn (1) or Co (2)). Such two compounds feature a new layered structure in which Δ-type chains built from alternating corner- and edge-sharing M3(μ3-OH) triangles are further connected via the Y-shaped “–(CH2)2–C(–O–)–” spacers. Magnetic studies reveal that there are dominant antiferromagnetic interactions between the metal ions. In 1, the complicated magnetic couplings in the Δ-type chains result in spin competition, displaying spin-glass behaviors. In 2, spin fluctuation behavior was observed and the critical field at 2 K is 25 kOe.
On the one hand, in frustrated magnets, localized magnetic moments, or spins, interacting through competing exchange interactions cannot be simultaneously satisfied, which may lead to some interesting phenomena such as spin liquid at low temperature.4 As one of the archetypes of spin frustration, Δ-type chain magnets that contain chains of corner- and/or edge-sharing triangles of M3(μ3-bridge) may provide a theoretical model to study magneto-structural correlations and act as secondary building blocks to construct novel frustrated materials. However, up to now, examples of coordination polymers (CPs) with Δ-type chain structures are still limited in number.5–7 To create magnetic CPs with Δ-type chain structures, the introduction of a μ3-OH− co-ligand has been proved to be an effective strategy.
On the other hand, metamagnetic or spin fluctuation behaviour may occur if the couplings are weak enough to be overcome by an external field and enter into another magnetic state along with the reorientation of spin.8 Frequently, such couplings are not strong and are exhibited in intermolecule, interchain, and interlayer interactions mediated by long linkers, blocking groups, hydrogen bonds, and π–π interactions.8a,9
Recently, we have undertaken research works on the coordination chemistry of a bifunctional ligand, namely 2-carboxyethyl(phenyl)phosphinic acid (H2L, see Scheme 1), which features two formally analogous carboxylate and phosphinate moieties (i.e., both contain two potential O donors) separated by a flexible ethylene spacer. The use of such a conformationally flexible ligand with multiple donor sites has demonstrated the structural diversity for the construction of CPs.10,11 In our current studies on magnetic CPs of metal carboxylate–phosphinates, here we present two unique examples of magnetic materials with Δ-type chain structures, i.e. [Mn3(L)2(OH)2] (1) and [Co3(L)2(OH)2] (2). Herein, we report their syntheses, crystal structures, and magnetic properties.
Compound | 1 | 2 |
---|---|---|
a R 1 = ∑‖Fo| − |Fc‖/∑|Fo|, wR2 = {∑w[(Fo)2 − (Fc)2]2/∑w[(Fo)2]2}1/2. | ||
Empirical formula | C18H20O10P2Mn3 | C18H20O10P2Co3 |
Formula weight | 623.10 | 635.07 |
Space group | P21/n | P21/n |
a (Å) | 5.6315(4) | 5.4625(4) |
b (Å) | 6.7154(5) | 6.5901(6) |
c (Å) | 29.169(3) | 29.293(3) |
β/deg | 91.174(6) | 90.889(6) |
V/Å3 | 1102.88(16) | 1054.38(15) |
Z | 2 | 2 |
D calcd/g cm−3 | 1.876 | 2.000 |
μ/mm−1 | 1.890 | 2.541 |
GOF on F2 | 1.083 | 1.001 |
R 1, wR2 [I > 2σ(I)]a | 0.0887, 0.1933 | 0.0414, 0.0824 |
R 1, wR2 (all data) | 0.1072, 0.2018 | 0.0603, 0.0895 |
1 | |||
---|---|---|---|
Symmetry codes: #1 x − 1, y, z; #2 −x + 1, −y + 1, −z; #3 −x + 1, −y, −z. | |||
Mn(1)–O(2)#1 | 2.138(6) | Mn(1)–O(5)#2 | 2.156(6) |
Mn(1)–O(4) | 2.168(7) | Mn(1)–O(5) | 2.217(7) |
Mn(1)–O(1) | 2.258(6) | Mn(1)–O(1)#3 | 2.268(6) |
Mn(2)–O(5) | 2.097(6) | Mn(2)–O(3)#3 | 2.213(6) |
Mn(2)–O(1)#1 | 2.266(7) | ||
2 | |||
Co(1)–O(2)#1 | 2.054(2) | Co(1)–O(5)#2 | 2.058(2) |
Co(1)–O(4) | 2.096(3) | Co(1)–O(5) | 2.136(3) |
Co(1)–O(3)#3 | 2.171(3) | Co(1)–O(1) | 2.224(3) |
Co(2)–O(5) | 2.009(3) | Co(2)–O(3)#3 | 2.138(2) |
Co(2)–O(1)#1 | 2.172(2) |
The interconnection of the Mn2+ ions by the L2− and μ3-OH− ligands results in the formation of a complicated layered structure. As shown in Fig. 2, one Mn2 and two equivalent Mn1 ions are connected by the μ3-OH− group, forming a Mn3(μ3-OH) triangle unit. In such a triangle unit, the Mn⋯Mn distances (Å) are 3.4407(2) for Mn1⋯Mn1A, 3.2520(2) for Mn1⋯Mn2, and 3.3468(2) for Mn1A⋯Mn2, respectively. The Mn1–O5–Mn1A, Mn1–O5–Mn2 and Mn1A–O5–Mn2 angles (°) are 103.762(2), 97.790(2) and 103.783(2), respectively. The Mn1⋯Mn2 and Mn1A⋯Mn2 edges are also bridged by the carboxylate oxygen (O3) and phosphinate oxygen (O1A) of the L2− ligand, respectively. The corresponding Mn1–O3–Mn2 and Mn1A–O1A–Mn2 angles (°) are 93.053(2) and 95.419(2), respectively. The neighbouring Mn3(μ3-OH) triangle units are both corner-shared at the Mn2 site and edge-shared at the Mn1⋯Mn1A borderline, thus leading to an infinite Δ-type chain running along the a-axis (Fig. 2, top). The neighbouring chains are further bridged by the Y-shaped “–(CH2)2–C(–O–)–” spacers to form a layer in the ab plane. The shortest inter-chain Mn⋯Mn distance within the layer is 4.6801(3) Å.
Fig. 2 View of the layered structure of 1 down the c-axis. Phenyl groups of the L2− ligands and all hydrogen atoms have been omitted for clarity. For other display details, see Fig. 1. |
Furthermore, the layers in 1 are assembled through C–H⋯π interactions between the phenyl groups of the L2− ligands at two adjacent layers, forming a three-dimensional supramolecular structure (Fig. 3). The inter-layer distance is 14.929(1) Å.
Fig. 3 View of the three-dimensional supramolecular structure of 1 down the a-axis. All hydrogen atoms except for the H4A atom have been omitted for clarity. The C–H⋯π interactions between the inter-layer L2− ligands are drawn as dashed lines. The dihedral angle of two adjacent inter-layer phenyl rings is ca. 75.4°. For other display details, see Fig. 1. |
For 1, the χmT value for a Mn3 triangle unit at 300 K is 12.1 cm3 mol−1 K, which is smaller than the spin-only value for three Mn(II) ions (S = 5/2, 13.12 cm3 mol−1 K for g = 2.0). With decreasing temperature, the χmT value first decreases gradually to a minimum value of ca. 5.85 cm3 mol−1 K at 18 K, then increases sharply to a maximum value of 34.2 cm3 mol−1 K at 4 K, and finally drops down to 26.0 cm3 mol−1 K at 2 K (Fig. 4a). The data at 30–300 K can be fitted to the Curie–Weiss law, giving C = 13.61 cm3 mol−1 K and θ = −34.0 K. The negative θ value indicates that there are dominant antiferromagnetic interactions between the Mn(II) ions. On the other hand, the increase of the χmT value below 18 K implies the uncompensating magnetic moment owing to the spin-frustrated/spin-competing interaction in μ3-OH−-bridged Mn3 units; whereas, the decrease of the χmT value below 4 K is likely due to a saturation effect (vide infra) and/or a weak interchain antiferromagnetic interaction.
Fig. 4 (a) Temperature dependence of χmT () and χm−1 (○) for the microcrystalline sample of 1. The red solid line is Curie–Weiss fitting; (b) magnetization measured at 2 K for 1. |
The field dependence of the magnetization at 2.0 K is shown in Fig. 4b. The magnetization curve shows a rapid increase at very small external fields, and then exhibits a linear increase at the high-field range. The value (4.82 Nβ) at 5 T per Mn3 unit is close to the moment of one Mn(II) ion (5 Nβ), suggesting a non-compensating resultant moment of one Mn(II) ion per Mn3 unit, a typical characteristic of topological ferrimagnets.17 No significant remnant magnetization and coercive field were observed in the hysteresis loop at 2.0 K, suggesting a soft magnetic characteristic for 1 (Fig. S4, ESI†). Moreover, the FC and ZFC curves diverge at below 3.6 K (Fig. S5, ESI†), suggesting the occurrence of irreversibility of magnetization. AC susceptibility measurements revealed that both in-phase (χ′) and out-of-phase (χ′′) signals (Fig. S6, ESI†) are somewhat frequency-dependent, which should be ascribed to spin-glass behaviours that often happen on triangle spin-competing systems.18
The complicated magnetic behaviour at low temperature should be mainly ascribed to the complicated magnetic exchange topology in 1. As shown in Fig. 2, a Δ-type chain is built from alternating corner- and edge-sharing Mn3(μ3-OH) triangles with the carboxylate and phosphinate as co-bridges. As revealed by temperature susceptibilities at high temperature, the intra-triangle magnetic interaction should be of antiferromagnetic type, which is consistent with the Mn–O–Mn angles in the range of 92.9(3)–103.9(4)°. In such a Δ-type chain, spin competition is expected as the intrachain magnetic interactions cannot be satisfied at the same time (Fig. 5). This spin-competing behaviour is also supported by a moderate value of the frustration parameter (f = |θ|/TN = 9.2) in 1.19 As the Mn(II) ion usually has a weak magnetic anisotropy, the magnetic structure in such a Δ-type chain is strongly determined by the coupling between the sites. For this case, a single phase transition from the paramagnetic state to a magnetically ordered state was predicted.20 Taking into consideration the large χ′T value (131 cm3 mol−1 K) observed at 3.8 K and the magnetization of 4.82 Nβ at 5 T, one of the possible magnetic structures of a single chain at low temperature can be proposed for 1. As shown in Fig. 6, for each single chain, the moments on the edge-sharing (Mn1) and vertex-sharing (Mn2) sites form two individual sub-lattices, which are anti-parallel to each other, and thus, a non-compensating resultant moment of one Mn(II) ion per Mn3 unit could be formed, leading to a topologically ferrimagnetically ordered state for 1.
For 2, the χmT value at room temperature is 9.63 cm3 mol−1 K, which is much larger than the expected spin-only value (5.625 cm3 mol−1 K) for three magnetically isolated Co(II) ions (S = 3/2, g = 2) and this phenomenon can be attributed to the strong spin–orbit coupling of Co(II) ions. As the temperature decreases, the value of χmT slowly decreases at high temperature and quickly decreases down to a minimum value of 0.48 cm3 mol−1 K at 2.0 K. Curie–Weiss fitting of the magnetic data from 2 K to 300 K for 2 results in C = 10.23 cm3 mol−1 K and θ = −17.3 K. Notably, the negative θ value does not indicate dominant antiferromagnetic coupling between the metal centers because of the strong spin–orbit coupling of Co(II) ions, which itself can lead to a negative θ value and a decrease of χmT at high temperature.21 The strength of the antiferromagnetic exchange interaction caused by the spin–orbit coupling of Co(II) at high temperature was estimated based on eqn (1).22
χT = Aexp(−E1/kT) + Bexp(−E2/kT) | (1) |
In eqn (1), A + B equals the Curie constant, and E1 and E2 are the spin–orbit coupling constant and activation energy of antiferromagnetic interactions, respectively. The best fit of the experimental data gives A + B = 10.53 cm3 mol−1 K, −E1/k = −90.24 K, and −E2/k = −4.98 K for 2 (Fig. 6a). The negative value of −E2/k indicates the dominant antiferromagnetic interactions between adjacent Co(II) ions.22
The field-dependent magnetization of 2 at 2 K has a value of 5.92 Nβ (far from the saturation value) at 7 T and exhibits a pronounced sigmoid shape at low field. The latter implies spin fluctuation behaviour for 2 that a magnetic transition occurs from the AF interaction at low field to a FM state at high field (Fig. 6b), and the critical field defined as dM/dH at 2 K is 25 kOe. To investigate the nature of the spin fluctuation of 2, the low temperature magnetic susceptibilities at different fields were measured. The χmvs. T plot (Fig. 6c) shows a cusp around 5.0 K when the applied fields are lower than 25 kOe. Below 5.0 K, χm drops sharply toward zero, suggesting that the ground state of the system is S = 0. The cusp disappears at higher fields and the antiferromagnetic couplings mediated by μ3-OH could be overcome at external fields larger than 25 kOe, from which 2 turns into a ferromagnetic state. In addition, the antiferromagnetic behavior is further confirmed by the temperature-dependent in-phase (χ′) magnetic susceptibilities (Fig. S7, ESI†), but no out-of-phase (χ′′) signals could be observed.
As is known, the usual coordination modes of carboxylates to transfer magnetism could be divided into four types: two of which are syn–syn-μ2-η1:η1 and anti–anti-μ2-η1:η1 modes, which likely produce antiferromagnetism between CoII ions; one is a syn–anti-μ2-η1:η1 mode, which probably produces weak ferromagnetism; the last one is a μ2-η2 mode, which provides more opportunities for generating ferromagnetism.21b A common rule that they transfer ferromagnetic interactions when the Co–O–Co angle is smaller than 110° was widely recognized. For 2, there exists an anti–syn–anti-μ3-η1:η2 mode of carboxylate groups, phosphinate and μ3-OH−. The Co–O–Co angles of 93.54° and 94.51° (O atoms are from phosphinate and carboxylate) and anti–syn–anti-μ3-η1:η2 carboxylates are favorable for the transference of ferromagnetism, however, the existence of a μ3-OH− magnetic exchange pathway may play a key role in the antiferromagnetic interaction between CoII ions.
Footnote |
† Electronic supplementary information (ESI) available: PXRD patterns, TGA curves and magnetic characterization of 1 and 2. CCDC 1054109 and 1418238. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ce02537d |
This journal is © The Royal Society of Chemistry 2017 |