Open AccessCCS ChemistryRESEARCH ARTICLE7 Dec 2022Air-Stable Dy(III)-Macrocycle Enantiomers: From Chiral to Polar Space Group Zhenhua Zhu†, Chen Zhao†, Quan Zhou, Shuting Liu, Xiao-Lei Li, Akseli Mansikkamäki and Jinkui Tang Zhenhua Zhu† State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Chinese Academy of Sciences, Beijing 100049 †Z. Zhu and C. Zhao contributed equally to this work.Google Scholar More articles by this author , Chen Zhao† State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 †Z. Zhu and C. Zhao contributed equally to this work.Google Scholar More articles by this author , Quan Zhou State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Google Scholar More articles by this author , Shuting Liu State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Google Scholar More articles by this author , Xiao-Lei Li State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 Google Scholar More articles by this author , Akseli Mansikkamäki *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] NMR Research Unit, University of Oulu, Oulu FI-90014 Google Scholar More articles by this author and Jinkui Tang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Rare Earth Resource Utilization, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.022.202101604 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Magnetoelectric (ME) multiferroic materials have unique advantages in low-power and high-density information storage, because they can simultaneously display ferroelectricity and ferromagnetism. However, research on how to construct air-stable high-performance ME single-molecule magnets (SMMs) is nonexistent. Herein, by introducing homochirality while reducing molecular symmetry, two double-decker Dy(III) enantiomers adopting the polar space group P21 and exhibiting excellent thermal stability were obtained. They displayed zero field SMM behavior with an anisotropy barrier (Ueff) of ca. 100 cm−1. This work establishes a rational chemical design strategy for crystallizing SMMs in polar space groups and elucidates the direction for future research, that is, engineering small-size high-performance SMMs. Download figure Download PowerPoint Introduction Single-molecule magnets (SMMs) exhibit fascinating potential in information storage and molecular spintronics applications.1–3 In recent years, Dy(III) metallocene SMMs have made numerous impressive achievements,4–8 among which is the discovery of [(Cpi-Pr5)Dy(Cp*)][BPh4] (Cpi-Pr5 = penta-isopropyl cyclopentadienyl, Cp* = pentamethylcyclopentadienyl) with magnetic bistability at temperatures as high as 80 K.9 In addition, multifunctional molecular magnetic materials have also attracted attention in this field because of their structural flexibility, tunability, and important applications in sensors and optical devices,10 including luminescent,11,12 conducting,13–15 and chiral SMMs.16–19 In addition, the development of SMMs that respond to physical stimuli and substrate-SMM hybrids strongly drives SMM research toward the realization of functioning proof-of-principle devices.3,20–23 In 2020, Long et al.24 reported a Yb(III)-based chiral complex, R,R-[Zn(OAc)(L)Yb(NO3)2], (H2L = [6,6′-((1E,1′E)-(((1R,2R)-1,2-diphenylethane-1,2-diyl)bis(azaneylylidene))bis(methaneylylidene)), which exhibited strong room temperature magnetoelectric (ME) coupling resulting from the combination of ferroelectric behavior and magnetostrictive effect. However, in terms of SMM properties, only a typical field-induced slow relaxation of the magnetization at very low temperature (below 3 K) was observed. This pioneering work motivates us to explore ME coupling in air-stable high-performance Dy(III) SMMs for applications in low-power and high-density information storage, which requires fast electrical writing and magnetic reading at the same time.25,26 To achieve this goal, the compound needs to meet the following requirements: (1) a local axial crystal field (CF), (2) a polar space group, and (3) switchable ferroelectric polarization.27 To the best of our knowledge, however, only a few high-performance SMMs crystallize in the polar space group,6,28 and these SMMS were not obtained by a rational design strategy. To date, no chemical design strategy for crystallizing SMMs in a polar space group has been established. In terms of crystallization, among 11 chiral point groups, five of them are also polar (C1, C2, C4, C3, and C6), providing a higher possibility for them to adopt polar point groups with respect to achiral point groups (5/11 vs 10/32).29 Consequently, to induce the complexes to crystallize in the polar space group, we first constructed homochiral SMMs. It should be noted that if the chiral compound crystallizes in low-symmetry triclinic or monoclinic crystal systems, all chiral point groups are also polar point groups (P1, P2, P21, and C2) but none of them belong to polar point groups in the orthorhombic system ( Supporting Information Table S1). Herein, two types of chiral organic ligands, R/S-1-(2-hydroxynaphthalen-1-yl)naphthalen-2-ol (HL1, axial chirality) and RRRR/SSSS-(2E,5E,8E,11E)-3,5,9,11-tetraaza-1,7(2,6)-dipyridina-4,10(1,2)-dicyclohexa-nacyclododecaphane-2,5,8,11-tetraene (LN6R/S, carbon-center chirality), were used as an axial and equatorial ligand, respectively, resulting in two pairs of enantiomers, [DyIIILN5(L1)2](BPh4)·2H2O (LN5 = (2E,13E)-2,14-dimethyl-3,6,10,13-tetraaza-1(2,6)-pyridinacyclotetra-decaphane-2,13-diene) ( 1R/ 1S) and [DyIIILN6R/S(L2)2](BPh4)·xCH2Cl2·yH2O (HL2 = triphenylsilanol; x = 1, y = 0, 2R; x =1, y = 2, 2S) (Scheme 1). However, they all crystallize in the chiral but not polar space group P212121 (D2, 222), which belongs to the orthorhombic system. When a less sterically hindered ligand, 4-(methylthio)phenol (HL3), was used to replace bulky Ph3SiOH in 2R/ 2S, two L3-blocking double-decker Dy(III) enantiomers ( 3R/ 3S) were obtained. Compared with the compound [Dy(LE)(4-MeO-PhO)2](BPh4)·3THF (LE = equatorial Schiff base ligand derived from 2,6-pyridinedicarboxaldehyde and (1R,2R)-(+)-1,2-diphenylethylenediamine),30 it is the equatorial plane without enough steric hindrance that is responsible for the dimeric structures. As anticipated, they crystallize in the chiral polar space group P21 (C2, 2) due to the combination of homochirality and molecular symmetry breaking. Scheme 1 | Synthetic routes of [DyIIILN5(L1)2](BPh4) (1R/1S), [DyIIILN6R/S(L2)2](BPh4) (2R/2S) and [DyIII2(LN6R/S)2(L3)2(μ-OH)2](BPh4)2 (3R/3S). Download figure Download PowerPoint Experimental Methods Chemicals and synthesis All chemicals and solvents were commercially available and used as received without any further purification, and all experiments were performed under aerobic conditions. The starting materials R/S-1-(2-hydroxynaphthalen-1-yl)naphthalen-2-ol (R/S-BINOL), 2,6-diacetylpyridine, N,N′-bis(2-aminoethyl)-1,3-propanediamine, NaBPh4, Et3N, 2,6-diformylpyridine, 1R,2R-diaminocyclohexane, 1S,2S-diaminocyclohexane were purchased from J&K Scientific (Beijing, China). The following general procedure was used to synthesize air-stable Dy(III)-macrocycle enantiomers. 2,6-Diacetylpyridine (0.2 mmol, 32.6 mg), N,N′-bis(2-aminoethyl)-1,3-propanediamine (0.2 mmol, 34.8 mg), and DyCl3·6H2O (0.1 mmol, 37.7 mg) were combined in 15 mL of methanol. The mixture was refluxed for 5 h. After cooling to room temperature, the solvent was removed under vacuum to yield a black solid, which was dissolved in 15 mL of dichloromethane (DCM) and 15 mL of deionized water. Then R-BINOL (0.3 mmol, 85.9 mg) and NaBPh4 (0.1 mmol, 34.2 mg) were added to the mixture. The mixture was then refluxed for 1 h. After cooling to room temperature, the DCM phase was separated. Orange crystals suitable for X-ray measurement were obtained by layering with n-pentane at room temperature for several days. X-ray crystallography and magnetic measurements Single-crystal X-ray data for all complexes were recorded on a Bruker SMART APEX diffractometer (Bruker AXS, Ettlingen, Germany) equipped with graphite-monochromatized Mo Kα radiation (λ = 0.71073 Å) at 180 K. Structures were solved in Olex2 with SHELXT using intrinsic phasing and were refined with SHELXL using least squares minimization.31–33 All nonhydrogen atoms were refined anisotropically. All hydrogen atom positions were calculated geometrically and refined using the riding model. The crystal data have been deposited at the Cambridge Structural Database (CCDC-2078246-2078249, 2078252-2078253), which can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif. The magnetic measurements were explored using a Quantum Design MPMS XL-7 SQUID magnetometer (Quantum Design, Inc., San Diego, CA) equipped with a 7 T magnet. Variable-temperature direct current (dc) magnetic susceptibility data under a 1000 Oe field and in the 2–300 K range were collected on powdered crystalline samples. The alternating current (ac) measurements were investigated at different frequencies from 1 to 1488 Hz under a 0 dc field and a 3.0 Oe ac oscillating field. Note that all complexes can rapidly lose solvent (DCM and H2O) from the lattice at room temperature. Therefore, all measurements were performed after the sample had been exposed to air for several days at room temperature. Computational details The geometries of 1R, 1S, 2R, 2S, 3R, and 3S were extracted from the respective crystal structures. Solvent molecules and counterions not directly coordinated to the Dy ions were removed from the structures. The positions of the hydrogen atoms were optimized using density functional theory (DFT) while the positions of heavier atoms were frozen to their crystal-structure coordinates. The optimization of the hydrogen positions was carried out using the Amsterdam Density Functional (ADF) code version 2019.103.34,35 The pure Perdew–Burke–Ernzerhof (PBE) exchange-correlation functional36,37 was used in conjunction with Grimme’s empirical DFT-D3 dispersion correction38 utilizing the Becke–Johnson damping function.39 Scalar relativistic effects were introduced using the zeroth-order regular approximation (ZORA) as implemented in ADF.40–42 The default ZORA-optimized all-electron Slater-type basis sets of valence triple-ζ-quality with two sets of polarization functions were used throughout.43 The effects of static electron correlation within the 4f shell were simulated by distributing the nine 4f electrons evenly over the seven orbitals. In practice, this means that the occupations of the seven (in the case of 1R, 1S, 2R, and 2S) or fourteen (in the case of 3R and 3S) highest occupied β-electron orbitals were set to 0.285714 each. The geometry convergence thresholds were increased to 10−4, 10−4, 10−3, and 10−1 atomic units for energy, gradient, bond length step size, and bond angle step size, respectively. A set of state-averaged (SA) complete active space self-consistent field (CASSCF) calculations were then carried out on each Dy ion in 1R, 1S, 2R, 2S, 3R, and 3S.44–47 In the case of bimetallic complexes 3R and 3S, they were carried out by performing two separate sets of calculations where the other Dy(III) ion was replaced by the diamagnetic Y(III) ion with a similar ionic radius. The active space consisted of the nine 4f electrons in the seven 4f orbitals. All 21 sextet, 224 doublet, and 490 quartet states were solved in three different SA calculations. Spin–orbit coupling (SOC) was then introduced using the well-established spin–orbit restricted active space state interaction (SO-RASSI) approach where the SOC is introduced in a basis of the SA-CASSCF eigenstates.48 All 21 sextets, lowest 128 doublets, and 130 quartets corresponding to an energy cut-off of 50,000 cm−1 were included in the SO-RASSI calculation. The SA-CASSCF calculations were carried out using OpenMolcas code version 20.10.49,50 Scalar relativistic effects were introduced using the exact two-component transformation.51–53 Relativistically contracted atomic natural orbital basis sets were used throughout.54–57 A polarized valence triple-ζ basis was used for the Dy ions; a valence double-ζ basis without polarization functions was used for the H atoms; and a polarized double-ζ basis was used for the remaining atoms. The SOC operator was constructed using the atomic mean field integral formalism.48,58 The single-ion static magnetic properties ( g-tensors, ab initio crystal field decompositions, local effective barriers for the relaxation of magnetization, and the magnetic susceptibilities of 1R, 1S, 2R, and 2S) were calculated using the SINGLE_ANISO module of OpenMolcas.59–62 The exchange interaction in 3R and 3S was modeled using the Lines model as implemented in the POLY-ANISO module of OpenMolcas.60,63,64 The lines parameter was determined by scanning the value of a single effective exchange parameter in increasingly shorter increments. The parameter was initially scanned between values of J = −4.0 and +4.0 cm−1 in 0.02 cm−1 increments. Then the increments were decreased to 0.01, 0.001, and 0.0001 cm−1, and the scan was repeated near the observed minimum in the error evaluated as the root-mean-square difference between the calculated and measured product of temperature and magnetic susceptibility (χMT). The reported magnetic susceptibilities of 3R and 3S were calculated using the POLY_ANISO module and the obtained lines exchange parameters. Results and Discussion Synthesis and molecular structures All complexes were synthesized using the same procedure: (1) in situ formation of a Dy(III)-macrocycle precursor with Cl– or H2O as axial ligands and (2) replacement of the above axial ligands with stronger anionic donors. Their molecular structures were confirmed by single-crystal X-ray diffraction analysis ( Supporting Information Tables S2–S10). In all structures, the charges of the complex cation were balanced by uncoordinated BPh4− anions. Thermogravimetric analysis shows excellent thermal stability of the complexes ( Supporting Information Figures S1–S3). The enantiomeric nature of the R-/S-isomers was confirmed by completely inverse circular dichroism (CD) spectra in the solid state and CH3OH solution (Figure 1 and Supporting Information Figure S4). In 1R or 1S, the spectra are dominated by two bands centered at ∼221 and 235 nm corresponding to the π→π* transitions of the naphthalenic unit. 2R/ 2S and 3R/ 3S exhibited similar CD spectra, because they contain the same type of equatorial macrocycle featuring two absorption bands at 298 and 324 nm. In the structures of 1R or 1S (Figure 2), the five neutral N atoms of the macrocycle are almost coplanar, and the Dy(III) ion is slightly (0.132 Å) tilted out of this plane ( Supporting Information Figure S5). The coordination geometry around the Dy(III) ion can be approximated to a distorted pentagonal bipyramid with point-group symmetry D5h ( Supporting Information Table S11). For 1R, the axial Dy–O bond lengths are 2.184(5) and 2.150(5) Å, while the equatorial Dy–N distances are obviously longer, averaging to 2.450 Å. Note that it is the intramolecular offset face-to-face π–π stacking interaction between the equatorial pyridine ring and one axial naphthalenic unit that leads to a difference of 0.03 Å in the axial Dy–O distance ( Supporting Information Figure S6). The axial O–Dy–O angle is 166.89(19)°, deviating far from a linear arrangement, probably as a result of the unique axial ligand.65 In 2R, the axial positions are occupied by an achiral ligand, Ph3SiO−, whereas the equatorial ligand is optically pure. The geometry around the Dy(III) ion is a pseudohexagonal bipyramid ( Supporting Information Table S12) with axial Dy–O distances of 2.139(3) and 2.123(3) Å and an average Dy–N bond length up to 2.647 Å. The axial O–Dy–O angle is 178.06(16)°, which is almost linear. The above structural characteristics imply that 2R/ 2S possess higher anisotropy barriers than 1R/ 1S due to their stronger axial CF. Figure 1 | CD spectra of R- (red) S- (blue) isomers for 1R/1S (left), 2R/2S (middle) and 3R/3S (right) in CH3OH (c = 1 × 10−5 M). Download figure Download PowerPoint Figure 2 | Structure of the cation (top) and equatorial plane (bottom) in 1R/1S (left) and 2R/2S (right). Orange, Dy(III); cyan, Si; blue, N; red, O; gray and green, C. Hydrogen atoms are omitted for clarity. Download figure Download PowerPoint In 3R (Figure 3) both Dy(III) ions are nine-coordinate with a hula-hoop coordination geometry of a local approximate C2v symmetry ( Supporting Information Table S13). Two hydroxide anions reside between two [DyIIILN6] units with Dy–O distances of 2.2790(3) and 2.2878(4) Å for Dy1 and 2.2568(3) and 2.2869(4) Å for Dy2 as well as Dy1–O–Dy2 angles of 114.413(13)° and 113.590(13)°. The outer axial positions are occupied by deprotonated 4-(methylthio)phenol with short Dy–O bond lengths of 2.153(6) Å for Dy1 and 2.186(6) Å for Dy2. The average Dy–N distances are 2.686 Å in the upper plane and 2.663 Å in the lower plane. The intramolecular Dy(III)–Dy(III) distance is 3.820 Å, implying the presence of a strong dipolar interaction. Initially, in addition to pursuing the polar space group, we also hoped to introduce an exchange bias effect in 3R by maintaining the local D6h symmetry of Dy(III) to observe open hysteresis loops at zero field in contrast to 2R ( Supporting Information Figure S13). However, the change of coordination geometry of the Dy(III) ion from hexagonal bipyramid to hula-hoop greatly weakens the single-ion anisotropy. The molecular structures of the S isomers are very similar to the respective R isomers. Figure 3 | Structure of the cation (top) and equatorial plane (bottom) in 3R/3S. Orange, Dy(III); yellow, S; blue, N; red, O; gray and green, C. Hydrogen atoms are omitted for clarity. Download figure Download PowerPoint Magnetic property measurements By analyzing the crystal packing of the above complexes, the shortest intermolecular Dy(III)···Dy(III) distances in 2R/2S and 3R/3S are in the range of 11.991–12.122 Å ( Supporting Information Figures S11–S14), suggesting that the complexes are discrete cations with negligible magnetic intermolecular interactions; however, for 1R/1S, the C–H⋯O intra- and intermolecular interactions create a 1D foldable structure along the a-axis, giving Dy(III)⋯Dy(III) distances of 8.744 and 8.715 Å ( Supporting Information Figures S7–S10). The strong CF splitting existing in 1R/ 1S and 2R/ 2S is evident from the slight decrease of χMT values upon cooling ( Supporting Information Figure S15) and the large separation between the ground and excited states (vide infra). For 1R/1S, a slight increase of χMT values below 10 K was observed, which was attributed to the presence of weak ferromagnetic intermolecular interactions,66,67 whereas a rapid increase of χMT values at low temperatures for 3R/3S indicated the existence of intramolecular ferromagnetic coupling. Ab initio studies show that the exchange interaction is dominated by ferromagnetic dipolar coupling due to the short intramolecular Dy(III)⋯Dy(III) distance and the collinear alignment of the principal magnetic axes ( Supporting Information Table S14 and Figure S29). Ac susceptibility measurements were performed under a zero external dc field to investigate their dynamic magnetic properties. As shown in Figure 4, compared with 1R, 2R exhibited a higher temperature of 84 K for peaks in χ″(ν) signals and a slower relaxation rate at low temperatures, indicating a higher anisotropy barrier (Ueff). An almost closed magnetic hysteresis loop for 1R at 1.9 K was observed with an average sweep rate of 31 Oe/s ( Supporting Information Figure S16), while 2R displayed waist-restricted hysteresis loops that remained open up to 5 K at the same sweep rate ( Supporting Information Figure S18). The magnetic behavior of 2R was also confirmed by field cooling and zero field cooling measurements, which showed divergence at ∼13 K, with a peak in the zero-field cooling (ZFC) measurement observed at about 5.1 K, consistent with the magnetic hysteresis study ( Supporting Information Figure S20). In the case of 3R, the χ″(ν) plots display well-defined maxima at temperatures of 1.9–16 K. However, the intramolecular ferromagnetic interaction cannot compensate for the weakening of axial CF,68 consequently, no open magnetic hysteresis was observed for 3R at 1.9 K using the same field sweep rate as 2R ( Supporting Information Figure S21). The temperature-dependent relaxation times and distribution parameters from ac susceptibility data were fitted to a generalized Debye model using the CC-FIT2 code.69 These data were fit to eq (1) for 1R and 2R ln ( τ ) = − ln [ CT n + τ 0 − 1 exp ( − U eff / k B T ) + 1 / τ QTM ] (1) ln ( τ ) = − ln [ CT n + τ 0 − 1 exp ( − U eff / k B T ) ] (2)with Ueff = 280(3) cm−1, τ0 = 2.99(10) × 10−9 s, C = 0.79(9) s−1 K−n, n = 2.64(4), τQTM = 1.03(4) × 10−2 s and Ueff = 1011(26) cm−1, τ0 = 3.10(19) × 10−12 s, C = 2.29(6) × 10−2 s−1 K−n, n = 2.51(7), and τQTM = 9.68(80) × 10−2 s, respectively (Figure 4). For 3R, Raman relaxation is the dominant process. The temperature-independent quantum tunneling of magnetization (QTM) process was not observed in ln(τ) versus T−1. Therefore, relaxation times were fitted using eq (2), yielding C = 5.45(25) s−1 K−n, n = 2.51(3), Ueff = 104(3) cm−1, and τ0 = 7.85(55) × 10−8 s. The fitted parameters of S isomers are very close to the respective R isomers due to their almost identical first coordination sphere of the Dy(III) center ( Supporting Information Figures S17, S19, and S22–S28). Figure 4 | χ″(ν) plots at zero applied dc field (top) and temperature dependent relaxation rate (bottom) for 1R (left, 1.9–40 K), 2R (middle, 6–90 K)and 3R (right, 1.9–20 K). The blue line in ln(τ) vs T−1 corresponds to a fit of the data in the respective temperature range. Download figure Download PowerPoint Ab initio calculations Ab initio calculations were performed to understand the magnetic relaxation mechanisms ( Supporting Information Figures S29 and S30 and Tables S14–S46). In 1R, the transverse components of the g tensor and the angle between the principal magnetic axis and that of the ground Kramers doublet (KD) grow quickly in the excited KDs ( Supporting Information Table S15). This predicts that the effective barrier is most likely crossed already at the first excited KD, giving an effective energy barrier of 377 cm−1, which is larger than the fitted value but still in reasonable agreement considering the accuracy of the computational methods and the overlooked intermolecular magnetic interactions. In 2R, the collinear and axial structure of the doublets is retained in higher-energy KDs as expected from the more axial CF. The lowest three KDs are almost collinear whereas the principal axis of the third excited KD is almost exactly perpendicular to the principal axis of the ground KD ( Supporting Information Table S17). Similarly, the transverse components of the g tensors become very significant in the third excited KD. This suggests that the effective barrier becomes crossed at the third excited doublet giving an effective barrier of 1160 cm−1, which is close to the fitted value. In the case of both Dy(III) ions in 3R, the transverse components of the g tensors become significant already at the first excited KD, while the principal magnetic axes retain near collinearity up to the second excited KD ( Supporting Information Tables S19 and S20). This suggests that the local effective barrier is crossed already at the first excited KD, although higher-energy relaxation pathways are also possible. The relaxation pathways determined by inspecting the magnitudes of the transition magnetic dipole moment matrix elements connecting the respective states are very similar to those based on analysis of the g tensors and will not be discussed separately (Figure 5 and Supporting Information Figure S30 and Tables S39–S46).62 In the case of 3R, the fitted barrier height, 104(3) cm−1, is noticeably smaller than the calculated value of ∼330 cm−1. This “underbarrier” relaxation has been explained as a consequence of anharmonic phonons or a Raman process with an Orbach-like temperature dependence.70,71 It should be noted that ln τ versus T−1 for 3R shows a noticeable deviation from linearity throughout the temperature range indicating significant contributions from non-Orbach relaxation. Figure 5 | The predicted pathway for the relaxation of magnetization for 1R (top), 2R (middle), and 3R (bottom two). Darker arrows indicate larger transition magnetic moment matrix elements between the states. Transition magnetic moment matrix elements involving states not involved in the relaxation process have been omitted for clarity. Download figure Download PowerPoint Conclusion We have reported the design, synthesis, structures, magnetic properties, and ab initio studies for three pairs of enantiomers of Dy(III)-macrocycles. In particular, we explored for the first time how to construct air-stable SMMs with a polar space group for ME coupling. By introducing homochirality while reducing molecular symmetry, two zero-field SMMs ( 3R/ 3S) with polar space group P21 were successfully isolated. Unfortunately, ferroelectric behaviors were not observed most likely due to the large coercive field caused by the rigid homochiral cation with high rotational energy.72 This work lays the foundations for the chemical design of Dy(III)-based molecular ferroelectrics with excellent SMM properties. Studies have shown that small-size SMMs with low-energy molecular rotations would have enhanced magnetic bistability.73 Fortunately, such molecules are also expected to possess a smaller coercive field, which will boost the occurrence of ferroelectric behaviors and will be studied in future work. Supporting Information Supporting Information is available and includes crystallographic data, magnetic measurements, and the results of ab initio calculations for all compounds. Conflict of Interest There is no conflict of interest to report. Acknowledgments This work was supported by the National Natural Science Foundation of China (no. 21871247), the Key Research Program of Frontier Sciences, CAS (no. ZDBS-LY-SLH023), the Key Research Program of the Chinese Academy of Sciences (no. ZDRW-CN-2021-3-3), and the Academy of Finland (grant no. 332294). Computational resources were provided by CSC-IT Center for Science in Finland and the Finnish Grid and Cloud Infrastructure (persistent identifier urn:nbn:fi:research-infras-2016072533).