Synthesis, characterization, magnetism and theoretical analysis of hetero-metallic [Ni2Ln2] partial di-cubane assemblies

Article abstract of DOI:10.1039/d1dt00510c

A family of four isostructural [Ln₂Ni₂(L)₂(μ₃-OCH₃)₂(μ_1,3-PhCO₂)₂(PhCO₂)₂(CH₃OH)₄]·2CH₃OH [where Ln = Gd (1), Tb (2), Dy (3) and Ho (4)] complexes has been synthesized using Schiff base ligand 2-[{(2-hydroxybenzyl)imino}methyl]-6-methoxyphenol (H₂L). All the complexes possess a partial di-cubane core structure where the growth of the core is contingent upon the ligand anions and solvent generated μ₃-OCH₃groups. DC magnetic analysis revealed dominating ferromagnetic interactions between the metal ions, however, we find no slow relaxation characteristics in the AC susceptibility. Further insight into the magnetic behavior of the reported complexes was achieved using DFT and CASSCF theoretical calculations, leading to the comprehension of the fast relaxation characteristics observed by magnetometry.

Full text of DOI:10.1039/d1dt00510c

Dalton
Transactions
View Article Online
View Journal
PAPER
Synthesis, characterization, magnetism and
theoretical analysis of hetero-metallic [Ni2Ln2]
Cite this: DOI: 10.1039/d1dt00510c
partial di-cubane assemblies†
Avik Bhanja, a Radovan Herchel, b Eufemio Moreno-Pineda,c Anjan Khara,a
a
Wolfgang Wernsdorfer d,e and Debashis Ray
*
A family of four isostructural [Ln2Ni2(L)2(μ3-OCH3)2(μ1,3-PhCO2)2(PhCO2)2(CH3OH)4]·2CH3OH [where Ln =
Gd (1), Tb (2), Dy (3) and Ho (4)] complexes has been synthesized using Schiff base ligand 2-[{(2-hydroxy-
benzyl)imino}methyl]-6-methoxyphenol (H2L). All the complexes possess a partial di-cubane core struc-
ture where the growth of the core is contingent upon the ligand anions and solvent generated μ3-OCH3
groups. DC magnetic analysis revealed dominating ferromagnetic interactions between the metal ions,
however, we find no slow relaxation characteristics in the AC susceptibility. Further insight into the mag-
netic behavior of the reported complexes was achieved using DFT and CASSCF theoretical calculations,
leading to the comprehension of the fast relaxation characteristics observed by magnetometry.
Received 15th February 2021,
Accepted 26th July 2021
DOI: 10.1039/d1dt00510c
rsc.li/dalton
expected potential for use in high-density information
storage,8 quantum computing,9 luminescence,10 and mole-
Introduction
Advances in strategies for the synthesis of oxido–hydroxido cular spintronics.11 Using lanthanoid ions within the aggre-
mixed transition metal-lanthanoid aggregates during the last gates is highly advantageous because of its large magnetic
three decades have resulted in plentiful crystalline materials moment (J) and considerable magnetic anisotropy, though the
that exhibit attractive structures and interesting properties. Ln⋯Ln interaction often increases the rate of quantum tunnel-
The choice, design and synthesis of new ligand systems have ing, leading to an apparent decrease in Ueff values.12 The
attracted considerable attention in the exploratory synthesis of single-ion magnetic anisotropy of these metal ions is known to
newer 3d–4f complexes.1–5 Synthesis of such multinuclear be modulated by changing their coordination environments
coordination clusters from newer ligand systems having pre- and resulting ligand field.13 Whereas the molecular anisotropy
defined coordination pockets for different metal ions, unfold- depends on factors like ligand field,14 the relative orientation
ing the synthetic potential and versatility of the reaction con- of the individual single-ion easy axes,15,16 magnetic coupling,17
ditions, has contributed to the understanding of the struc- and the structural topology of the newly found magnetic
ture–property relationship for Single Molecule Magnets cores.18–20 New types of such magnetic cores were discovered
(SMMs).6,7 Such newly synthesized molecular magnets have by the incorporation of varying numbers of 3d and 4f ions
bound to the multiple numbers of ligand anions and ancillary
bridges.21,22 Such heterometallic molecular aggregates,
aDepartment of Chemistry, Indian Institute of Technology, Kharagpur 721302, India.
bearing multiple paramagnetic centers (3d and 4f ions) in
E-mail: dray@chem.iitkgp.ernet.in
close proximity, can influence the effective energy barriers23
(Ueff) and magnetic blocking temperatures (TB).24 DyIII, TbIII
and HoIII ions are usually known to display a large magnetic
anisotropy steaming from the unquenched spin–orbit coupling
and magnetic moment. In practice, the presence of ligand
anion binding 3d ions in these aggregates induces exchange
interactions which can stabilize the bistable ground state and
suppresses the quantum tunneling of magnetization.25 The
ligand design strategy for the synthesis of 3d–4f aggregates is
guided by the hard–soft acid–base (HSAB) principle where
imine donor atoms bind 3d ions and oxygens are utilized for
coordinating/bridging 4f ions. Such a strategy is fruitful in pro-
bDepartment of Inorganic Chemistry, Faculty of Science, Palacky University, 17. listo-
padu12, CZ-771 46 Olomouc, Czech Republic
cDepto. de Química-Física, Escuela de Química, Facultad de Ciencias Naturales,
Exactas y Tecnología, Universidad de Panamá, Panamá, Panamá
dInstitute for Quantum Materials and Technology (IQMT), Karlsruhe Institute of
Technology (KIT), Hermann-von-Helmholtz-Platz 1, D-76344 Eggenstein-
Leopoldshafen, Germany
ePhysikalisches Institut, Karlsruhe Institute of Technology, D-76131 Karlsruhe,
Germany
†Electronic supplementary information (ESI) available: SHAPE analysis, crystal
data, PXRD curves, selected bond lengths and angles, and CASSCF calculations
are described in Fig. S1–S11 and Tables S1–S15 in the text. CCDC
2062614–2062617. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/d1dt00510c
This journal is © The Royal Society of Chemistry 2021
Dalton Trans.

View Article Online
Paper
Dalton Transactions
viding many Ni-4f complexes of different nuclearity of individ- (S. D. Fine Chemicals, Mumbai, India); Gd(NO3)3·6H2O, Tb
ual 3d and 4f ions having
topology.26–29
a
unique magnetic core (NO3)3·5H2O, Dy(NO3)3·5H2O and Ho(NO3)3·5H2O (Alfa Aesar,
India); and o-vanillin (Spectrochem Pvt. Ltd Mumbai).
Tetranuclear 3d–4f butterfly complexes are interesting
owing to their synthetic accessibility and structural unique-
ness, stimulating to the magnetochemists because of their
defective partial di-cubane magnetic core and the changes in
magnetic properties upon changing the metal ion position.30
Within this metal-oxo core, either the 3d ions can be posi-
tioned at the tip of the core with a good range separation
(inverse butterfly, type-I) or they can be located at the vertices
of the body (butterfly, type-II). From the earlier reports, 3d–4f
complexes with Cr(III),31 Co(III)32 and Mn(III)33 ions provided
structures with the type-I core possessing SMM properties with
a high anisotropic barrier; whereas the type-II core having
partial di-cubane structures are mainly found with NiII 34 ions,
though reports of type-II core systems employing other 3d
metal ions are also there showing fascinating SMM
behavior.35,36 The magnetization relaxation pathways of these
“Butterfly” complexes could be well elucidated following their
magnetostructural correlation data.37,38 Keeping all these in
mind, we have synthesized the ligand H2L, 2-[{(2-hydroxyben-
zyl)imino}methyl]-6-methoxyphenol, for self-aggregating reac-
tions with Ni(NO3)2·6H2O and selected LnIII metal ions.
Herein we report, the synthesis and characterization of four
Synthesis of 2-[{(2-hydroxybenzyl)imino}methyl]-6-
methoxyphenol (H2L)
The Schiff base ligand was synthesized by following a modified
reported procedure.42
General protocol for synthesizing complexes 1–4
All the complexes reported in this work were synthesized fol-
lowing a similar reaction protocol. H2L (0.1 mmol, 0.025 g)
was dissolved in 10 mL of MeOH–DCM (2 : 1) solution. 5 mL
methanolic solution of Ln(NO3)3·nH2O (0.10 mmol) [n = 6 for
Ln3+ = Gd (1) and n = 5 for Ln3+ = Tb (2), Dy (3) and Ho (4),
Scheme 1] was dissolved into the ligand solution, followed by
stirring for 1 h. Solid Ni(NO3)2·6H2O (0.10 mmol, 0.03 g),
sodium benzoate (0.20 mmol, 0.029 g) and NEt3 (0.20 mmol,
0.02 g) were further added to the preceding pale yellow solu-
tion and the reaction mixture was stirred for a further 7 h at
room temperature. The clear bright green solution was then fil-
tered and left undisturbed for slow evaporation. Green block
shaped air stable crystals were acquired after a week, suitable
for single crystal analysis. The characterization data of the
complexes are given below.
3d–4f isostructural aggregates, [Ln2Ni2(L)2(μ3-OCH3)2(μ1,3
-
[Gd2Ni2(L)2(μ3-OCH3)2(μ1,3
-
PhCO2)2(PhCO2)2(MeOH)4]·2CH3OH [where Ln = Gd (1), Tb (2),
Dy(3) and Ho (4)]. To date, to the best of our knowledge, only
one report of the Dy6 coordination cluster is known with the
present ligand system39 but no 3d–4f complex has been
reported.
PhCO2)2(PhCO2)2(MeOH)4]·2CH3OH (1). Yield: 0.035 g, 42%
(based on Gd). Anal. Calcd for C66H76Gd2N2Ni2O22 (1681.19):
C, 47.19; H, 4.56; N, 1.67. Found: C, 47.33; H, 4.57; N, 1.61.
Selected FT-IR data (KBr) cm−1: 1636 (w), 1593 (m), 1556 (m),
1474 (s), 1454 (m) 1397 (s), 1307 (m), 1276 (s), 1220 (s), 1169
(s), 1077 (s), 1025 (s), 848 (m), 718 (s), 589 (w), 471 (m), 419 (s).
[Tb2Ni2(L)2(μ3-OCH3)2(μ1,3
-
PhCO2)2(PhCO2)2(MeOH)4]·2CH3OH (2). Yield: 0.031 g, 37%
(based on Tb). Anal. Calcd for C66H76Tb2N2Ni2O22 (1684.54):
C, 47.06; H, 4.55; N, 1.66. Found: C, 47.27; H, 4.59; N, 1.59.
Experimental section
Reagents and starting materials
All solvents were purified according to the literature pro- Selected FT-IR data (KBr) cm−1: 1637 (w), 1592 (m), 1556 (m),
cedures,40 while the remaining starting materials were used as 1472 (s), 1454 (m), 1397 (s), 1305 (m), 1275 (s), 1219 (s), 1170
obtained without further purification. 2-(Aminomethyl)phenol (s), 1074 (s), 1020 (s), 853 (m), 718 (s), 590 (w), 467 (m), 416 (s).
was synthesized following a reported procedure.41 The follow-
[Dy2Ni2(L)2(μ3-OCH3)2(μ1,3-
ing chemicals were used as received: Ni(NO3)2·6H2O and NEt3 PhCO2)2(PhCO2)2(MeOH)4]·2CH3OH (3). Yield: 0.037 g, 44%
Scheme 1 Schematic diagram for complexes 1–4.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
(based on Dy). Anal. Calcd for C66H76Dy2N2Ni2O22 (1691.69): C, hydrogen atom positions were optimized using the B97M-D3BJ
46.86; H, 4.53; N, 1.66. Found: C, 46.73; H, 4.51; N, 1.63. functional and DKH-def2-TZVP for Ni, N, and O atoms;
Selected FT-IR data (KBr) cm−1: 1634 (w), 1593 (m), 1560 (m), SARC2-DKH-QZV for Ln atoms; and DKH-def2-SVP for C and
1475 (s), 1454 (m), 1394 (s), 1309 (m), 1275 (s), 1220 (s), 1173 H atoms.47,48 The Douglas–Kroll–Hess Hamiltonian was used
(s), 1075 (s), 1019 (s), 851 (m), 720 (s), 592 (w), 468 (m), 419 (s). to treat relativistic effects49,50 together with the Gaussian finite
[Ho2Ni2(L)2(μ3-OCH3)2(μ1,3
-
nucleus model51 and the increased radial integration accuracy
PhCO2)2(PhCO2)2(MeOH)4]·2CH3OH (4). Yield: 0.033 g, 39% for metal atoms was also set. The auxiliary basis sets were gen-
(based on Ho). Anal. Calcd for C66H76Ho2N2Ni2O22 (1696.55): erated using the AutoAux52 and the chain-of-spheres
C, 46.72; H, 4.52; N, 1.65. Found: C, 46.53; H, 4.50; N, 1.61. (RIJCOSX) approximation to the exact exchange was also
Selected FT-IR data (KBr) cm−1: 1632 (w), 1594 (m), 1561 (m), used.53,54 The isotropic exchange parameters were calculated
1475 (s), 1456 (m), 1394 (s), 1309 (m), 1275 (s), 1219 (s), 1173 using the B3LYP hybrid functional.55 The state average com-
(s), 1073 (s), 1020 (s), 852 (m), 719 (s), 592 (w), 469 (m), 415 (s). plete active space self-consistent field (SA-CASSCF) wave func-
tion calculations were done for NiZnLu2 or Zn2LnLu complexes
Physical measurements
1–4, in which only one paramagnetic metal ion was preserved.
For the elemental analysis (C, H, N) of the powder samples, a The active space is defined by five d-orbitals/seven f-orbitals
PerkinElmer model 240C elemental analyzer was used. A and the respective number of electrons for nickel/lanthanide
PerkinElmer model RX1 FT-IR spectrometer fitted with KBr ions. The number of the involved multiplets was as follows:
disks is used for FT-IR spectral measurements. Powder X-ray NiII – 10 triplets and 15 singlets; TbIII – 7 septets, 140 quintets,
diffraction was carried out using a BRUKER D2 PHASER diffr- 588 triplets, and 490 singlets; DyIII – 21 sextets, 224 quartets,
actometer (30 kV/10 mA) using Cu-Kα radiation (λ = 1.5418 Å) and 490 doublets; and HoIII – 35 quintets, 210 triplets, and
within the 5–50° (2θ) range.
196 singlets. These calculations employed slightly different
basis sets: SARC2-DKH-QZVP for paramagnetic Ln or DKH-
def2-TZVP for Ni, DKH-def2-TZVP for all donor atoms attached
Magnetic measurements
The magnetic susceptibility data for all complexes were col- to paramagnetic metal, and DKH-def2-SVP for all other atoms,
lected using a Quantum Design MPMS®3 magnetometer on a except for Lu for which SARC-DKH-TZVP was used. In the case
polycrystalline material in the temperature range 2–300 K of complexes 2–4, analogous CASSCF calculations were done
under an applied DC magnetic field (H). The magnetization using OpenMOLCAS 19.11.56 The RASSCF method was used in
data were collected between 2 and 5 K and fields ranging from the CASSCF calculations with the following numbers of multi-
0 to 7 T. The data were corrected for diamagnetic contributions plets: 7 septets, 140 quintets, 113 triplets, and 123 singlets for
from the eicosane and core diamagnetism employing Pascal’s TbIII; 21 sextets, 224 quartets, and 490 doublets for DyIII; 35
constants.43
quintets, 210 triplets, and 196 singlets for HoIII; and 10 triplets
and 15 singlets for NiII. While all multiplets of TbIII, HoIII and
NiII were included in the spin–orbit RASSI–SO procedure, the
X-ray crystallography
Single crystal X-ray structural studies of complexes 1–4 were number of states for DyIII was limited as follows: 21 sextets, 128
performed in a Bruker SMART APEX-II CCD X-ray diffract- quartets, and 130 doublets. The following basis sets were used:
ometer equipped with a graphite-monochromated Mo-Kα (λ = Tb,·ANO-RCC-VQZP; Dy,·ANO-RCC-VQZP; Ho,·ANO-RCC-VQZP;
0.71073 Å) radiation source using the ω scan (width of 0.3° per Lu,·ANO-RCC-VDZP; Zn,·ANO-RCC-VDZP; O,·ANO-RCC-VDZP;
frame) at 293 K with a scan rate of 5 s per frame. For indexing, N,·ANO-RCC-VDZP; C,·ANO-RCC-VDZ; and H,·ANO-RCC-VDZ.57
integration and space group determination, SAINT, SMART The calculated spin density and tensor axis were visualized
and XPREP software programs have been used. Direct methods using the VESTA 3 program.58
using SHELXS-2014 were used to solve crystal structures and
refined by the full-matrix least squares technique using
SHELXL (2014/7) programs. The empirical absorption correc-
tions were done by the Multi-scan method of the SADABS
program and the H atoms were incorporated using the riding
Results and discussion
General synthetic procedures
model. The DIAMOND software was used for presenting mole- In our previous work, we have shown that the presence of a
cular structures. The cell parameters for complexes 1–4 are flexible amine arm in the ligand backbone does facilitate the
summarized in Table S1 in the ESI.† Crystallographic data coordination driven aggregation during the synthesis of a new
(excluding structure factors) have been deposited with the family of Ni–Ln complexes.59 Earlier, while using 2-{[(2-
Cambridge Crystallographic Data Centre as supplementary hydroxy-3-methoxybenzyl)imino]methyl}phenol (H2L′) as
a
publications CCDC 2062614–2062617 (1–4).†
ligand, we have seen that the flexible –CH2 moiety plays an
important role in trapping LnIII metal ions within the hexanuc-
lear cluster.60 In continuation of our previous study, herein we
Theoretical calculations
First, the ORCA 4.2 software44–46 was used for the theoretical have synthesized H2L by reacting o-vanillin and 2-(amino-
calculations. All calculations were based on the experimental methyl)phenol (Scheme 2). The overall molecular structure of
X-ray structures of Ni2Ln2 complexes of 1–4, in which only H2L′ and H2L are the same, while the flexibility around the
This journal is © The Royal Society of Chemistry 2021
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Scheme 2 Synthesis of H2L.
imine nitrogen donor has transferred from o-vanillin to the
salicylaldehyde part. The used ligand skeleton, 2-[{(2-hydroxy-
benzyl)imino}methyl]-6-methoxyphenol, has shown to be an
excellent coordination support for synthesizing high nuclearity
4f-complexes by Zhang et al.,39 but never explored for synthe-
sizing 3d–4f complexes. Thus, we have explored the efficacy of
H2L for the coordination induced self-assembly reactions with
LnIII nitrate salts in the presence of Ni(NO3)2·6H2O, PhCO2Na
and NEt3 at a ratio of 1 : 1 : 1 : 2 : 2 in MeOH–DCM (2 : 1)
solvent medium. The reaction resulted in the formation of
[Ln2Ni2(L)2(μ3-OCH3)2(μ1,3-PhCO2)2(PhCO2)2(MeOH)4]·2CH3OH
(where Ln = Gd (1), Tb (2), Dy(3) and Ho (4)). These electroneu-
tral complexes 1–4 were synthesized as per eqn (1).
Fig. 1 Earlier reported Ni2Dy2 complexes with a butterfly topology: (a),
(b),34 (c), (d),62 (e)61 and (f).63
2H2L þ 2LnðNO3Þ ꢀ 6H2O þ 2NiðNO3Þ2 ꢀ 6H2O
þ 4PhCO2Na þ38CH3OH ! ½Ln2Ni2ðLÞ ðμ ꢁ OCH3Þ
3
2
2
ð1Þ
ðμ1;3 ꢁ PhCO2Þ2ðPhCO2Þ2ðCH3OHÞ4ꢂ ꢀ 2CH3OH
þ 10NO3ꢁ þ 6Hþ þ 4Naþ þ 24H2O
on several M2Ln2 complexes using HO2CtBu as a ligand back-
bone where Ni2Dy2 and Ni2Er2 complexes show slow relaxation
of magnetization.63 Zhang et al. have synthesized an Ni2Dy2
complex using a 2-(((2-hydroxy-3-methoxyphenyl)methylene)
amino)-2-(hydroxymethyl)-1,3-propanediol ligand and Ni2Dy2
and Ni2Tb2 butterfly complexes with a 2-(2,3-dihydroxpropyl-
iminomethyl)-6-methoxyphenol ligand.64 Later on in 2018,
Mohanta et al. also reported a series of Ni2Ln2 complexes
using an ((E)-2-(2-hydroxy-3-ethoxybenzylideneamino)phenol
ligand.65 In all these butterfly shaped Ni2Ln2 complexes, either
the carboxylato backbone or Schiff base ligand anion frame is
exploited to get the coordination aggregates bearing 3d and 4f
ions. In the present case, we have been successful in incorpor-
ating both Schiff base ligand anions and benzoato groups for
the growth of the new family.
The initial characterization of all the obtained complexes
was done from FTIR spectroscopic and PXRD analyses.
Complex 1 has been taken as a representative for the series for
the discussion. The imine nitrogen bound to the NiII center
recorded the characteristic stretching frequency for the NiII
bound CvN bond at 1636 cm−1 for 1 in comparison to the
free ligand value of 1648 cm−1. The two types of benzoato
group are recorded as sharp asymmetric stretches at 1593 and
1556 cm−1 and the symmetric stretches appear at 1454 and
1397 cm−1 (Fig. S2 in the ESI†). The differences of (ν
−
ˉasymm
ν
) 102 cm−1 and 196 cm−1 in stretching frequency values
ˉsymm
unequivocally confirm the presence of both the µ1,3-bridged
and terminal benzoato groups, respectively, within the mole-
cular aggregates. Experimental PXRD data of the synthesized
complexes are in good agreement with the simulated ones
derived from the single crystal X-ray diffraction data for all the
complexes (Fig. S3 in the ESI†).
Structural description of complexes 1–4
Diffraction quality single crystals of complexes 1–4 were
Several other Ni2Ln2 complexes capped by several classes of obtained from the MeOH–CHCl3 solution mixture. All the four
ligand anions are reported earlier showing interesting mag- crystals are isostructural and found to crystalize in the triclinic
ˉ
netic behavior. In 2014, Shanmugam et al. reported a series of P1 space group with Z = 1. To describe the structural features,
bent Ni2Ln2 complexes showing zero field SMM properties.61 complex 1 has been taken as a representative of the whole
Zhao et al. reported a linear Ni2Dy2 complex showing slow series with identical atom numbering schemes for all struc-
relaxation of magnetization with an appreciable anisotropic tures for easy comparison of the structural parameters. In all
barrier.62 The Ni2Ln2 complexes with an Ni2Ln2O6 butterfly the four cases an asymmetric unit contains half of the whole
core topology have also been proved as excellent candidates for molecule, viz., [NiLn(L)(μ3-OCH3)(μ1,3-PhCO2)(PhCO2)(MeOH)2]
single molecule magnets (Fig. 1). In 2011, Powell et al. (where Ln = Gd (1), Tb (2), Dy (3) and Ho (4), Fig. S4 in the
reported two types of Ni2Dy2 and Ni2Tb2 butterfly complexes34 ESI†). The crystal refinement parameters are presented in
using an ((E)-2-(2-hydroxy-3-methoxybenzylideneamino)phenol Table S1† whereas the bond distances and bond angles are
ligand and in 2015 Winpenny et al. reported a systematic study presented in Tables S4–S11 in the ESI.†
Dalton Trans.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Fig. 2 (a) Molecular structure of complex 1. H atoms and solvent molecules are removed for clarity. Colour code: C, grey; O, red; N, blue; Ni, green
and Gd, yellow. (b) Distorted partial dicubane core of complex 1.
The X-ray structure analysis revealed that the metal ion = 0.751, Table S2 in the ESI†) and the coordination sites are
centers of the tetranuclear core are well supported by two L2− occupied by ligand L2−, two μ3-OCH3 groups and one benzoate
ligand anions and two μ3-OCH3 groups, derived from solvent group (Fig. 3). Around Ni1, the N1–Ni1–O4 axis is the longest
MeOH used in the reaction medium. A detailed insight into one at 4.181 Å with bond distances N1–Ni1, 2.064(6) Å and
the structure revealed that each imine N atom of the ligand O4–Ni1, 2.124(5) Å. The remaining Ni–O distances lie within
anion specifically binds NiII centers with chelating support of the range 2.009(5)–2.070(5) Å. Both the NiII centers along with
two adjacent phenolato O donors (Fig. 2a). Both the phenolate two μ3-methoxo groups formed a rhombus where the Ni1–O4–
oxygen atoms are used for bridging NiII and its adjacent GdIII Ni1* angle is 97.4(2)° and the O4–Ni1–O4* angle is 82.6(2)°.
metal centers. The bidentate O,OMe donor of each ligand pre- The formation of this Ni2O2 rhombus is important for the for-
ferentially traps the oxophillic GdIII ions, making each ligand mation of the self-assembled Ni2Gd2 butterfly core.
susceptible to coordinate two GdIII and one NiII through its
The eight coordinated environments around the Gd1/Gd1*
OONO donor atoms. The Ni1 and Gd1 bound to pocket I and centers are occupied by the eight O atoms of four different
pocket II of L2− and further bridged by the μ1,3-PhCO2 moiety, types: three O atoms from two L2− ligands, two O from MeOH,
making a shorter intrametallic distance of 3.397(3) Å, whereas two O from the –PhCO2 moiety and one O from the μ3-OCH3
the other part bound to Gd1* possesses no bridging benzoate group. SHAPE 2.1 analysis confirmed the distorted trigonal
group and records a distance of 3.485(3) Å from Ni1. This dodecahedral environment around the GdIII center (Fig. 3,
Ni2Gd2O6 butterfly-shaped partial di-cubane core is mainly TDD-8 = 1.193, Table S3 in the ESI†). The Gd–O bond distances
built over two μ3-methoxo groups, providing a Ni1⋯Ni1* separ- lie within the range 2.297(5)–2.646(6) Å, where the longest
ation of 3.142(3) Å where both the NiII metal ions present at bond corresponds to Gd1–O1. Within the partial di-cubane
the vertices of the body position and two GdIII metal centers core a total of four different Ni–O–Gd angles is present in the
are at the tip of the wing position of the butterfly core. The NiII range of 98.15(19)°–105.2(2)° due to the presence of two
metal centers are in a distorted octahedral environment (OC-6 different types of O donors; where the smallest angle encom-
passed over the participation of the μ3-methoxido oxygen atom
and the largest one for the μ-phenoxido oxygen atom. The
mean plane analysis suggests all four metal centers are co-
planar and the μ3-methoxo oxygen atoms are displaced by
1.020 Å from the plane. The search for a hydrogen bonding
network showed that the molecule has no intramolecular
H-bonding interaction present but a 1D intermolecular weak
H-bonding connection propagates through the lattice MeOH
(Fig. S2, ESI†).
Magnetic studies
Several 3d–4f butterfly complexes have been shown to behave
Fig. 3 Distorted octahedral geometry of NiII metal ions (left) and dis-
torted trigonal dodecahedral geometry of GdIII ions (right).
as SMMs, a characteristic mainly arising from the lanthanide
complex.33,36,64 Interestingly, it has been shown that the relax-
This journal is © The Royal Society of Chemistry 2021
Dalton Trans.

View Article Online
Paper
Dalton Transactions
ation characteristics are contingent upon the neighboring 3d The lack of saturation in these systems indicates strong an-
metal, which could in principle be a consequence of the inter- isotropy in the systems. We also explore the dynamic character-
action between the 3d–4f pairs and the relative orientation of istics of the complexes via AC magnetic susceptibility;
the easy axes between the ions. To investigate the magnetic however, slow relaxation characteristics were not found for any
characteristics of the complexes herein studied, DC SQUID of the butterfly complexes without and with an applied HDC
measurements were carried out for all systems. The room field, and thus none of them shows SMM properties.
temperature χMT value for all complexes is very close to that
In the case of compound 1 containing GdIII, the spin
expected for the non-interacting systems, i.e., 17.8, 27.1, 29.9 Hamiltonian formalism is adequate for describing the mag-
and 30.3 cm3 K mol−1 for Ni2Gd2, Ni2Tb2, Ni2Dy2 and Ni2Ho2, netic properties. Based on the molecular structure of 1, the fol-
respectively.
Upon lowering the temperature, χMT remains practically
lowing spin Hamiltonian was employed:
~
~
~
~
~
~
~
~
ˆ
H ¼ ꢁ JGdꢁNiðSGd ꢀ SNi þ SGd ꢀ SNi* þ SGd* ꢀ SNi þ SGd* ꢀ SNi*
Þ
constant until about 20 K where it sharply increases to χMT
values of 34.6 cm3 K mol−1 for Ni2Gd2. For Ni2Tb2, Ni2Dy2 and
Ni2Ho2 the χMT also increases below 20 K reaching maximum
4
4
X
X
2
DkðSz;k ꢁ Sk2=3Þ þ μBB
gkSa;k
~
~
ˆ
ˆ
ˆ
ꢁ JNiꢁNiðSNi ꢀ SNi*Þ þ
k¼1
k¼1
χMT values of 55.3 (at 3 K), 47.7 (at 2.5 K) and 40.8 cm3
K
ð2Þ
mol−1 (at 2.5 K) before decreasing to 53.7, 46.8 and 40.0 cm3 K
mol−1, respectively (Fig. 4). The upsurge observed in the χMT
data indicates the existence of ferromagnetic interaction occur-
ring between the 3d–4f pairs. Further magnetization investi-
gations at variable fields and temperatures result in a satur-
ation value of 18.0 µB for Ni2Gd2 at 7 T and 2 K, while for
Ni2Tb2, Ni2Dy2 and Ni2Ho2 the magnetization values at the
lowest temperature (2 K) and the highest field (7 T) are 14.5,
15.4, and 16.4 µB, respectively, without reaching saturation.
where the first two terms describe the isotropic exchange
among Gd–Ni and Ni–Ni ions, then the single-ion zero-field
splitting term is added and finally there is a Zeeman term
defined for the a-direction of the magnetic field as Ba = B(sin
(θ)cos(φ), sin(θ)sin(φ), cos(θ)).66 Usually, it holds that D(GdIII)
is very small, and therefore we assumed D(GdIII) = E(GdIII) = 0.
With the aim to minimize the number of free parameters, ZFS
parameters of NiII were fixed to values calculated using the
Fig. 4 Temperature dependence of the χT product and the isothermal magnetizations measured at T = 2, 3, 4, and 5 K for (a) Ni2Gd2 (1). The empty
circles represent the experimental data points while the full lines in panel (a) represent the best fits calculated using eqn (2) and JGd–Ni = 0.715 cm−1
JNi–Ni = 6.49 cm−1, DNi = 4.35 cm−1, g = 1.98 and zj = −0.469 cm−1; (b) Ni2Tb2; (c) Ni2Dy2; and (d) Ni2Ho2.
,
Dalton Trans.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
Table 1 The CASSCF calculated D-tensor and g-tensor values of NiII in
1–4 performed using ORCA
interaction between the inner 4f orbitals and the 3d orbitals of
nickel, which often induces ferromagnetic interaction between
3d–4f metal pairs.63,70 Moreover, the ferromagnetic interaction
is in line with the magnetostructural correlation of Ni–Gd
exchange and the respective Ni–O–Gd angle.71
Compound
D(Ni)
E/D(Ni)
g-factors
1
2
3
4
5.43
4.69
4.56
4.52
0.173
0.147
0.165
0.133
2.274, 2.304, 2.316
2.273, 2.299, 2.309
2.259, 2.284, 2.295
2.261, 2.287, 2.295
Theoretical studies
The magnetic interactions and magnetic anisotropy are key
points in understanding the magnetic behavior of polynuclear
metal complexes. The complexity of such interactions is
usually hard to grasp from the experimental data alone, and
Density Functional Theory (DFT) and multireference calcu-
lations such as Complete Active Space Self-Consistent Field
(CASSCF) Methods are essential for better understanding the
behavior of this class of compounds. First, magnetic inter-
actions were investigated using DFT calculations with the
B3LYP hybrid functional using the ORCA 4.2 software.45 Here,
the broken-symmetry approach was applied to evaluate JNi–Ni
and JGd–Ni in the whole series 1–4, for which the respective
molecular structures were extracted from X-ray experimental
data. For complex 1, the following spin Hamiltonian was
considered:
CASSCF method (Table 1) and we also assumed gNi = gGd = g,
and thus, we are left with three parameters (JGd–Ni, JNi–Ni, and
g). Here, all experimental magnetic data were treated simul-
taneously using the POLYMAGNET program67 and the best-
fitted parameters were found as JGd–Ni = 0.685 cm−1, JNi–Ni
=
7.16 cm−1, and g = 1.98 providing good agreement with the
experiment – Fig. 4a. Obviously, both exchange interactions
are ferromagnetic in nature. The obtained ferromagnetic JNi–Ni
interaction also bodes well with the small Ni1–O4–Ni1* angle
in the complex, i.e., 97.4(2) Å, which has been found to induce
ferromagnetic interactions when the angle is small (<98°) and
antiferromagnetic interactions for wider angles.68,69 The ferro-
magnetic JGd–Ni interaction can likewise be related to the small
~
~
~
~
~
~
~
~
ˆ
H ¼ ꢁ JGdꢁNiðSGd ꢀ SNi þ SGd* ꢀ SNi*Þ ꢁ JGd*ꢁNiðSGd ꢀ SNi* þ SGd* ꢀ SNi
Þ
~
~
ꢁ JNiꢁNiðSNi ꢀ SNi*Þ:
ð3Þ
The comparison of energies of the high-spin state (HS) and
several broken–symmetry spin states (BS_NiNi*, BS_NiGd,
BS_NiGd*) was utilized for the calculations of the isotropic
exchange parameter J by Ruiz’s approach72
εBS NiNi* ꢁ εHS ¼ 16JGdꢁNi þ 16JGd*ꢁNi
37
2
37
εBS NiGd ꢁ εHS
¼
JGd*ꢁNi þ 3JNiꢁNi
ð4Þ
Fig. 5 The magnetostructural correlation of JNi–Ni* calculated using the
B3LYP functional in a series of 1–4 compounds.
εBS NiGd* ꢁ εHS
¼
JGdꢁNi þ 3JNiꢁNi
2
Fig. 6 The output of the ORCA CASSCF calculations with CAS(8,5) for Ni ions in complexes 1–4. The plot of the d-orbital splitting calculated using
ab initio ligand field theory (AILFT) (left), low-lying ligand-field terms (middle), and ligand-field multiplets showing splitting of the ground triplet state
(right).
This journal is © The Royal Society of Chemistry 2021
Dalton Trans.

View Article Online
Paper
Dalton Transactions
which resulted in JNi–Ni* = 5.89 cm−1, JGd–Ni = 0.968 cm−1 and spin–orbit coupling and the ligand-field. Therefore, CASSCF
JGd*–Ni = 0.948 cm−1. Thus, both types of isotropic exchange calculations were performed using ORCA, which enabled us to
are ferromagnetic, which is in accordance with fitted values evaluate the energies of d- or f-orbitals with ab initio ligand
from the magnetic data. The respective plots of the calculated field theory (AILFT)73,74 and respective ligand-field terms and
spin densities of 1 are shown in Fig. S9.† For the rest of com- ligand-field multiplets. The details of these calculations are
pounds 2–4, two paramagnetic lanthanide ions were replaced stated in the Experimental section. The splitting of d-orbitals
by the diamagnetic Lu analogue and calculations of BS_Ni shows a typical pattern for pseudooctahedral complexes and
were used to evaluate JNi–Ni* in these complexes as presented 10Dq is reaching values of ≈8000 cm−1 (Fig. 6). The lowest
in Table S12.† Generally, it holds that the value of JNi–Ni* corre- triplet ligand field term is well separated from the excited
lates with the Ni–Ni* distance as well as with the average <Ni– terms, and therefore, relatively small zero-field splitting of the
O–Ni* angle – Fig. 5. Next, we were interested in the zero-field (2S + 1) ground state is expected and found as plotted in Fig. 6
splitting of Ni and Ln ions (Tb, Dy, Ho) in 1–4 induced by the (right), reaching D ≈ 5.4–4.5 cm−1 with E/D values close to
0.13–0.17 (Table 1), where the values of DNi are in the expected
range for the hexacoordinate NiII ion.75,76 Interestingly, D is
constantly decreasing in 1–4 with an increase in the atomic
number of Ln and mostly correlates with a decrease in the
SHAPE index of OC-6 geometry (Table S2†). Similar CASSCF
calculations for Tb, Dy, and Ho in compounds 2–4 are shown
in Fig. 7.
As expected, the impact of the ligand-field on f-orbitals is
much smaller, f-orbitals are split in the range up to 800 cm−1
7
6
5
and the splitting of the respective F6, H15/2 and I8 states are
shown in Fig. 7 (right) and the respective energy levels are pre-
sented in Table S13.† In the case of DyIII, the ground state
Kramers doublet has large axial magnetic anisotropy (calcu-
lated g-factors: 0.061, 0.135, 19.342) and the axes of this
g-tensor are shown in Fig. S10.† The first excited state is
located at 108 cm−1, which makes a good predisposition for
SMM properties. Moreover, these calculations also suggest an
antiferromagnetic dipole–dipole interaction between two Dy
Fig. 7 The output of the ORCA CASSCF calculations with CAS(8,7) for
Tb in 2, with CAS(9,7) for Dy in 3 and with CAS(10,7) for Ho in 4. The plot
of the f-orbital splitting calculated using ab initio ligand field theory
(AILFT) (left), and ligand-field multiplets showing splitting of ground atoms, because the angle between the orientation of the mag-
states (7F6, 6H15/2 and 5I8) (right).
netic moment of the ground Kramers doublet with respect to
Fig. 8 Magnetization blocking barrier of DyIII in 3 calculated using SINGLE_ANISO (left) and the magnetization blocking barrier of Ni2Dy2 in 3 calcu-
lated using POLY_ANISO with JDy–Ni = 0.873 cm−1 and JNi–Ni = 7.96 cm−1 (right), zoomed to the lowest energy levels. The numbers presented in the
plot represent the corresponding matrix element of the transversal magnetic moment (for values larger than 0.1 an efficient relaxation mechanism
is expected), and Δtun shows the tunnelling gap of the indicated pseudo-doublets. Dashed lines refer to (temperature assisted) quantum tunnelling
(blue), Orbach/Raman mechanisms (red) and direct/Raman mechanisms (green).81
Dalton Trans.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
the Dy⋯Dy connecting line is 87°, which is more than the 1.25 cm−1, JDy–Ni = 0.873 cm−1 and JHo–Ni = 1.42 cm−1) and the
border value of 54.7°.77 However, the relatively large distance respective magnetic anisotropy of the axial symmetry for DyIII
between two ions means that this contribution to the whole ions in 3. The ferromagnetic interaction between the Ni⋯Ni
magnetic behavior will be minimal.
couple can be understood by the relatively narrow Ni–O–Ni
To better understand the magnetic behavior of 2–4, the angle. Furthermore, the lack of SMM characteristics can be
magnetic exchange interactions need to be determined. explained by a complex interplay of magnetic interactions in
Therefore, the OpenMOLCAS program was employed, because the compounds leading to low lying energy levels with zero
with SINGLE_ANISO and POLY_ANISO modules we are able to magnetic moments and thus diminishing the effective energy
calculate the dipolar and exchange magnetic interactions and barrier for the Orbach relaxation mechanism.
on top of that also the magnetization reversal barrier for whole
complexes. Thus, analogous CASSCF calculations were per-
formed for NiII and LnIII ions in 2–4 using OpenMOLCAS with Conflicts of interest
the SINGLE_ANISO module resulting in similar parameters of
The authors declare no conflict of interest.
the zero-field splitting of these ions – Tables S14 and S15.†
Evidently, the tunneling should be very fast in TbIII of 2 and
HoIII of 4, because there is a large energy gap of the first
Acknowledgements
pseudo-doublet, Δtun = 2.09 cm−1 for TbIII and Δtun = 4.14 cm−1
for HoIII, which clarify why compounds 2 and 4 lack slow relax-
ation of the magnetization. In contrast, the DyIII ion possesses
large axial anisotropy of the first Kramers doublet, and the
tunneling probability is small as shown in the magnetization
blocking diagram in Fig. 8.
AB is thankful to DST-INSPIRE for the research fellowship. For
the single-crystal XRD facility at the Department of Chemistry,
IIT Kharagpur, we acknowledge the FIST program of DST, New
Delhi. RH acknowledges the financial support from the insti-
tutional sources of the Department of Inorganic Chemistry,
Palacký University Olomouc, Czech Republic. EMP thanks the
Panamanian National Systems of Investigators (SNI, SENACYT)
for support. WW thanks the A. v. Humboldt foundation and
the ERC grant MoQuOS Nr. 741276.
Subsequently, a homemade routine was used to fit experi-
mental dc magnetic data of 2–4, both temperature and field
dependent in cooperation with the POLY_ANISO module.78–80
Within these calculations, the six lowest states for each LnIII
ions and triplet ground states for NiII ions were included in
the exchange interaction. During fitting, JNi–Ni was fixed to a
value found using DFT (Table S12†), and therefore we were left
with fitting just the JDy–Ni parameter and scaling coefficient.
The best agreement was obtained with JTb–Ni = 1.25 cm−1 for 2,
JDy–Ni = 0.873 cm−1 for 3 and JHo–Ni = 1.42 cm−1 for 4 –
Fig. S11.† Interestingly, all magnetic exchanges in 1–4 are
ferromagnetic in nature. To understand also the dynamic mag-
netic properties of 3, the magnetization reversal barrier for
Ni2Dy2 of 3 is plotted in Fig. 8 (right) showing dense splitting
of low energy levels, and a low-lying first pseudo-doublet at
∼2.5 cm−1 with zero magnetic moment, which most likely pre-
cludes slow relaxation of the magnetization in this compound.
References
1 C. Aronica, G. Pilet, G. Chastanet, W. Wernsdorfer,
J.-F. Jacquot and D. Luneau, Angew. Chem., Int. Ed., 2006,
45, 4659–4662.
2 N. F. Chilton, S. K. Langley, B. Moubaraki and K. S. Murray,
Chem. Commun., 2010, 46, 7787–7789.
3 Y.-Z. Zheng, M. Evangelisti, F. Tuna and R. E. P. Winpenny,
J. Am. Chem. Soc., 2012, 134, 1057–1065.
4 S. Hossain, S. Das, A. Chakraborty, F. Lloret, J. Cano,
E. Pardob and V. Chandrasekhar, Dalton Trans., 2014, 43,
10164–10174.
5 M. Savva, K. Skordi, A. D. Fournet, A. E. Thuijs,
G. Christou, S. P. Perlepes, C. Papatriantafyllopoulou and
A. J. Tasiopoulos, Inorg. Chem., 2017, 56, 5657–5668.
6 O. Kahn, Molecular Magnetism, VCH Publishers, New York,
1993.
7 R. Sessoli, D. Gatteschi, D. N. Hendrickson and
G. Christou, MRS Bull., 2000, 25, 66–71.
8 M. Affronte, J. Mater. Chem., 2009, 19, 1731–1737.
9 M. N. Leuenberger and D. Loss, Nature, 2001, 410, 789–793.
Conclusions
Four isostructural and new 3d–4f aggregates have been syn-
thesized from the coordinating support of the anionic form of
2-[{(2-hydroxybenzyl)imino}methyl]-6-methoxyphenol reacting
with Ni(NO3)2·6H2O and Ln(NO3)3·nH2O salts. The synergistic
and preferential coordination of 3d and 4f ions results in the
formation of a batch of partial di-cubane Ni2Ln2 compounds.
Typical involvement of benzoato anions from the externally 10 W. Huang, S. Huang, M. Zhang, Y. Chen, G.-L. Zhuang,
added sodium salt shapes the network within the aggregate
without giving any scope for further extended solvent water
Y. Li, M.-L. Tong, J. Yong, Y. Li and D. Wu, Chem.
Commun., 2018, 54, 4104–4107.
derived hydroxido/oxido bridges. The detailed quantitative 11 L. Bogani and W. Wernsdorfer, Nat. Mater., 2008, 7, 179–
analysis of the magnetic data supported by DFT and CASSCF 186.
calculations revealed a ferromagnetic exchange between NiII 12 R. J. Blagg, L. Ungur, F. Tuna, J. Speak, P. Comar,
ions and also between 3d–4f ions (JGd–Ni = 0.715 cm−1, JTb–Ni
=
D. Collison, W. Wernsdorfer, E. J. L. McInnes,
This journal is © The Royal Society of Chemistry 2021
Dalton Trans.

View Article Online
Paper
Dalton Transactions
L. F. Chibotaru and R. E. P. Winpenny, Nat. Chem., 2013, 5, 35 K. C. Mondal, A. Sundt, Y. Lan, G. E. Kostakis,
673–678.
O. Waldmann, L. Ungur, L. F. Chibotaru, C. E. Anson and
13 D. E. Freedman, W. H. Harman, T. D. Harris, G. J. Long,
A. K. Powell, Angew. Chem., Int. Ed., 2012, 51, 7550–7554.
C. J. Chang and J. R. Long, J. Am. Chem. Soc., 2010, 132, 36 A. Baniodeh, Y. Lan, G. Novitchi, V. Mereacre, A. Sukhanov,
1224–1225.
M. Ferbinteanu, V. Voronkova, C. E. Anson and
A. K. Powell, Dalton Trans., 2013, 42, 8926–8938.
37 K. R. Vignesh, S. K. Langley, K. S. Murray and
G. Rajaraman, Inorg. Chem., 2017, 56, 2518–2532.
14 N. Ishikawa, T. Iino and Y. Kaizu, J. Phys. Chem. A, 2002,
106, 9543–9550.
15 N. Ishikawa, Polyhedron, 2007, 26, 2147.
16 J.-L. Liu, F.-S. Guo, Z. S. Meng, Y.-Z. Zheng, J.-D. Leng, 38 Y. Peng, M. K. Singh, V. Mereacre, C. E. Anson,
M. L. Tong, L. Ungur, L. F. Chibotaru, K. J. Heroux and
G. Rajaraman and A. K. Powell, Chem. Sci., 2019, 10, 5528–
D. N. Hendrickson, Chem. Sci., 2011, 2, 1268–1272.
5538.
17 K. Kovnir, C. M. Thompson, V. O. Garlea, D. Haskel, 39 H. Zhang, R. Liu, J. Zhang, Y. Li, W. Liu and Y. Dong,
A. A. Polyanskii, H. Zhou and M. Shatruk, Phys. Rev. B:
Condens. Matter Mater. Phys., 2013, 88, 104429.
18 A. M. Ako, I. J. Hewitt, V. Mereacre, R. Clérac,
W. Wernsdorfer, C. E. Anson and A. K. Powell, Angew.
Chem., Int. Ed., 2006, 45, 4926–4929.
19 A. M. Ako, V. Mereacre, R. Clérac, W. Wernsdorfer,
I. J. Hewitt, C. E. Anson and A. K. Powell, Chem. Commun.,
2009, 544–546.
20 E. Ruiz, T. Cauchy, J. Cano, R. Costa, J. Tercero and
S. Alvarez, J. Am. Chem. Soc., 2008, 130, 7420–7426.
Chem. – Asian J., 2017, 12, 507–514.
40 B. S. Furniss, A. J. Hannaford, P. W. G. Smith and
A. R. Tatchell, Vogel’s Text book of Practical Organic
Chemistry, ELBS, Longman, London, UK, 5th edn, 1989.
41 I. Zagol-Ikapitte, V. Amarnath, M. Bala, L. J. Roberts,
J. A. Oates and O. Boutaud, Chem. Res. Toxicol., 2010, 23,
240–250.
42 Q.-L. Wang, C. Yang, L. Qi, D.-Z. Liao, G.-M. Yang and
H.-X. Ren, J. Mol. Struct., 2008, 892, 88–92.
43 G. A. Bain and J. F. J. Berry, Chem. Educ., 2008, 85, 532–536.
21 A. Mishra, W. Wernsdorfer, K. A. Abboud and G. Christou, 44 F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2,
J. Am. Chem. Soc., 2004, 126, 15648–15649. 73–78.
22 F. E. Kakaroni, D. I. Tzimopoulos, H. W. L. Fraser, 45 F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2018, 8,
M. Siczek, T. Lis, M. Evangelisti, E. K. Brechin and
C. J. Milios, Inorg. Chem., 2019, 58, 11404–11409.
23 J.-L. Liu, J.-Y. Wu, Y.-C. Chen, V. Mereacre, A. K. Powell,
e1327.
46 F. Neese, F. Wennmohs, U. Becker and C. Riplinger,
J. Chem. Phys., 2020, 152, 224108.
L. Ungur, L. F. Chibotaru, X.-M. Chen and M.-L. Tong, 47 D. Aravena, F. Neese and D. A. Pantazis, J. Chem. Theory
Angew. Chem., Int. Ed., 2014, 53, 12966–12970. Comput., 2016, 12, 1148–1156.
24 N. F. Chilton, C. A. Goodwin, D. P. Mills and 48 F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005,
R. E. Winpenny, Chem. Commun., 2015, 51, 101–103. 7, 3297–3305.
25 J. C. Oldengott, J. Schnack and T. Glaser, Eur. J. Inorg. 49 M. Douglas and N. M. Kroll, Ann. Phys., 1974, 82, 89–155.
Chem., 2020, 34, 3222–3235. 50 B. A. Hess, Phys. Rev. A, 1986, 33, 3742–3748.
26 H. Wang, H. Ke, S. Y. Lin, Y. Guo, L. Zhao, J. Tang and 51 L. Visscher and K. G. Dyall, At. Data Nucl. Data Tables,
Y.-H. Li, Dalton Trans., 2013, 42, 5298–5303. 1997, 67, 207–224.
27 D. I. Alexandropoulos, L. Cunha-Silva, G. Lorusso, 52 G. L. Stoychev, A. A. Auer and F. Neese, J. Chem. Theory
M. Evangelisti, J. Tang and T. C. Stamatatos, Chem.
Commun., 2016, 52, 1693–1696.
28 A. Bhanja, R. Herchel, Z. Travnícek and D. Ray, Inorg.
Chem., 2019, 58, 12184–12198.
Comput., 2017, 13, 554–562.
53 F. Neese, F. Wennmohs, A. Hansen and U. Becker, Chem.
Phys., 2009, 356, 98–109.
54 R. Izsák and F. Neese, J. Chem. Phys., 2011, 135, 144105.
29 A. Chakraborty, J. Goura, P. Bag and V. Chandrasekhar, 55 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens.
Eur. J. Inorg. Chem., 2019, 1180–1200. Matter Mater. Phys., 1988, 37, 785–789.
30 L. R. Piquer and E. C. Sañudo, Dalton Trans., 2015, 44, 56 I. F. Galván, M. Vacher, A. Alavi, C. Angeli, F. Aquilante,
8771–8780.
J. Autschbach, J. J. Bao, S. I. Bokarev, N. A. Bogdanov,
R. K. Carlson, L. F. Chibotaru, J. Creutzberg, N. Dattani,
M. G. Delcey, S. S. Dong, A. Dreuw, L. Freitag, L. M. Frutos,
L. Gagliardi, F. Gendron, A. Giussani, L. González, G. Grell,
M. Guo, C. E. Hoyer, M. Johansson, S. Keller, S. Knecht,
G. Kovačević, E. Källman, G. L. Manni, M. Lundberg,
Y. Ma, S. Mai, J. P. Malhado, P. Å. Malmqvist,
P. Marquetand, S. A. Mewes, J. Norell, M. Olivucci,
M. Oppel, Q. M. Phung, K. Pierloot, F. Plasser, M. Reiher,
A. M. Sand, I. Schapiro, P. Sharma, C. J. Stein,
L. K. Sørensen, D. G. Truhlar, M. Ugandi, L. Ungur,
A. Valentini, S. Vancoillie, V. Veryazov, O. Weser,
31 S. K. Langley, D. P. Wielechowski, V. Vieru, N. F. Chilton,
B. Moubaraki, L. F. Chibotaru and K. S. Murray, Chem. Sci.,
2014, 5, 3246–3256.
32 A. V. Funes, L. Carrella, E. Rentschler and P. Alborés,
Dalton Trans., 2014, 43, 2361–2364.
33 P. S. Koroteev, N. N. Efimov, A. B. Ilyukhin,
Z. V. Dobrokhotova and V. M. Novotortsev, Inorg. Chim.
Acta, 2014, 418, 157–162.
34 K. C. Mondal, G. E. Kostakis, Y. Lan, W. Wernsdorfer,
C. E. Anson and A. K. Powell, Inorg. Chem., 2011, 50,
11604–11611.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
T. A. Wesołowski, P.-O. Widmark, S. Wouters, A. Zech, 68 M. A. Halcrow, J.-S. Sun, J. C. Huffman and G. Christou,
J. P. Zobel and R. Lindh, J. Chem. Theory Comput., 2019, 15,
5925–5964.
57 B. O. Roos, R. Lindh, P.-Å. Malmqvist, V. Veryazov,
Inorg. Chem., 1995, 34, 4167–4177.
69 H. L. C. Feltham, R. Clerac and S. Brooker, Dalton Trans.,
2009, 2965–2973.
P.-O. Widmark and A. C. Borin, J. Phys. Chem. A, 2008, 112, 70 E. Moreno-Pineda, F. Tuna, Y.-Z. Zheng, R. E. P. Winpenny
11431–11435. and E. J. L. McInnes, Inorg. Chem., 2013, 52, 13702–13707.
58 K. Momma and F. Izumi, J. Appl. Crystallogr., 2011, 44, 71 S. K. Singh, N. K. Tibrewal and G. Rajaraman, Dalton
1272–1276. Trans., 2011, 40, 10897–10906.
59 A. Bhanja, E. Moreno-Pineda, R. Herchel, W. Wernsdorfer 72 E. Ruiz, J. Cano, S. Alvarez and P. Alemany, J. Comput.
and D. Ray, Dalton Trans., 2020, 49, 7968–7976. Chem., 1999, 20, 1391–1400.
60 A. Bhanja, M. Schulze, R. Herchel, E. Moreno-Pineda, 73 M. Atanasov, D. Ganyushin, K. Sivalingam and F. Neese,
W. Wernsdorfer and D. Ray, Inorg. Chem., 2020, 59, 17929–
17944.
61 N. Ahmed, C. Das, S. Vaidya, S. K. Langley, K. S. Murray
Molecular Electronic Structures of Transition Metal Complexes
II, ed. D. M. P. Mingos, P. Day and J. P. Dahl, Springer
Berlin Heidelberg, Berlin, Heidelberg, 2012, pp. 149–220.
and M. Shanmugam, Chem. – Eur. J., 2014, 20, 14235– 74 S. K. Singh, J. Eng, M. Atanasov and F. Neese, Coord. Chem.
14239. Rev., 2017, 344, 2–25.
62 Y. Li, Q. Shang, Y.-Q. Zhang, E.-C. Yang and X.-J. Zhao, 75 J. Titiš and R. Boča, Inorg. Chem., 2010, 49, 3971–3973.
Chem. – Eur. J., 2016, 22, 18840–18849.
63 E. Moreno-Pineda, N. F. Chilton,
76 R. Boča, Coord. Chem. Rev., 2004, 248, 757–815.
Tuna, 77 J.-D. Leng, J.-L. Liu, W.-Q. Lin, S. Gómez-Coca, D. Aravena,
F.
R. E. P. Winpenny and E. J. L. McInnes, Inorg. Chem., 2015,
54, 5930–5941.
E. Ruiz and M.-L. Tong, Chem. Commun., 2013, 49, 9341–
9343.
64 H.-H. Zou, L.-B. Sheng, F.-P. Liang, Z.-L. Chena and 78 L. F. Chibotaru, L. Ungur and A. Soncini, Angew. Chem., Int.
Y.-Q. Zhang, Dalton Trans., 2015, 44, 18544–18552. Ed., 2008, 47, 4126–4129.
65 S. Mandal, S. Ghosh, D. Takahashi, G. Christou and 79 L. F. Chibotaru and L. Ungur, J. Chem. Phys., 2012, 137,
S. Mohanta, Eur. J. Inorg. Chem., 2018, 2793–2804. 064112.
66 R. Boča, Theoretical Foundations of Molecular Magnetism, 80 L. Ungur, M. Thewissen, J.-P. Costes, W. Wernsdorfer and
Elsevier, Amsterdam, The Netherlands, 1999. L. F. Chibotaru, Inorg. Chem., 2013, 52, 6328–6337.
67 R. Boča and R. Herchel, Program POLYMAGNET, 2006– 81 A. Dey, P. Kalita and V. Chandrasekhar, ACS Omega, 2018,
2020.
3, 9462–9475.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans.