Counter-complementarity control of the weak exchange interaction in a bent {Ni(ii)3 complex with a μ-phenoxide-μ-carboxylate double bridge

Article abstract of DOI:10.1039/c9nj03574e

We have prepared and structurally characterized a novel {Ni₃} bent complex bearing a double μ-phenoxide-μ-carboxylate bridge. Both terminal Ni(ii) sites are symmetry related, offering a simplified exchange interaction scheme. DC magnetic data i

Full text of DOI:10.1039/c9nj03574e

NJC
View Article Online
View Journal | View Issue
PAPER
Counter-complementarity control of the weak
exchange interaction in a bent {Ni(II) } complex
3
Cite this: New J. Chem., 2019,
with a l-phenoxide-l-carboxylate double bridge†
43, 16218
a
b
b
a
Guillermo Fiorini, Luca Carrella, Eva Rentschler
and Pablo Albores
*
´
3
We have prepared and structurally characterized a novel {Ni } bent complex bearing a double
m-phenoxide-m-carboxylate bridge. Both terminal Ni(II) sites are symmetry related, offering a simplified
exchange interaction scheme. DC magnetic data is consistent with a weak antiferromagnetic interaction
between the central and terminal Ni(II) ions. As expected for a Ni(II) system, local zero-field splitting is
observed, which can be experimentally established. Broken symmetry quantum chemical calculations, as
well as ab initio CASSCF-SA-SOC computations that support the magnetic experimental data, were also
performed. From the analysis of other reported closely related Ni(II) systems, a counter-complementarity
effect exerted by the carboxylate bridge is proposed, which might explain the weaker exchange
interactions compared to those observed in double m-phenoxide bridged Ni(II) compounds.
Received 11th July 2019,
Accepted 18th September 2019
DOI: 10.1039/c9nj03574e
rsc.li/njc
3
order to achieve molecular nanomagnets. There are different
alternatives for the development of ferromagnetic exchange
Introduction
The research field of molecular magnetism is envisaged as interactions, the main ones being orthogonality of the magnetic
being a key area in the exploration of the diversity of systems, orbitals, spin-polarization or counter-complementarity of the
4
with the aim of understanding the origins and the degree of bridging ligands. In this sense, cumulated experimental infor-
1
strength of the spin exchange interaction. Regarding this mation indicates that there is not a definitive recipe to find a
fundamental issue, several magneto-structural correlations bridging ligand that unequivocally promotes the ferromagnetic
2
(both theoretical and experimental) have been established
exchange interaction independent of the metal ion and the
and owing to these studies, a broad range of molecule-based nature of additional bridges.
magnetic materials including single-molecule-magnets, have
It is well established that the counter-complementarity
been reported in recent years. With respect to the exchange effects of a second bridging ligand (L) play a fundamental
interaction in the polynuclear compounds of paramagnetic role in governing the magnetic properties of heterobridged
metal ions, the ferromagnetic type is significantly rarer than m-hydroxo/alkoxo/phenoxo-m-L Cu(II) dinuclear complexes
4
the anti-ferromagnetic one.
(L = azide, thiocyanate, pyrazolate, carboxylato, etc.). Although
The preparation of coordination compounds exhibiting Cu(II) has one magnetic orbital, other 3d metal ions (e.g. Ni(II),
ferromagnetic exchange interactions constitutes a hard task Mn(II) and Fe(II)) have two or more magnetic orbitals. Hence, it
for synthetic chemists, particularly when dealing with systems becomes very challenging to explore counter-complementarity
with a high degree of nuclearity. The importance of succeeding effects in these systems owing to the many possible combina-
in this enterprise relies not only on the scarcity of systems with tions of magnetic orbitals. The Ni(II) ion with only two unpaired
dominant ferromagnetic exchanges, but also in the fact that a electrons (two magnetic orbitals) appears to be the simplest
high spin is required, together with a large axial anisotropy, in case after the Cu(II) ones. In this sense, focusing on the
dinuclear or trinuclear compounds (with high symmetry) of
a
Ni(II) offers the possibility of looking into the nature of the
exchange interaction mediated by the hetero-bridged motifs.
In this work, we report a novel Ni(II) trinuclear complex
arranged in a bent configuration with symmetry related Ni(II)
Departamento de Qu ´ı mica Inorg a´ nica, Anal ´ı tica y Qu ´ı mica F ´ı sica/INQUIMAE
CONICET), Facultad de Ciencias Exactas y Naturales Universidad de Buenos
(
Aires, Pabell o´ n 2, Ciudad Universitaria, C1428EHA Buenos Aires, Argentina.
E-mail: albores@qi.fcen.uba.ar; Fax: +54-11-4576-3341
b
Johannes Gutenberg University Mainz Chemistry, Institute of Inorganic and
Analytical Chemistry, Duesbergweg 10-12, 55128 Mainz, Germany
Electronic supplementary information (ESI) available. CCDC 1938756. For ESI
terminal sites, bearing a m-phenoxo-m-carboxylate bridge, with
II
3
the formula [Ni
(L
2
)(piv)
2
(H
2
O)
4
] (1), H
2
L = o-aminophenol/
†
o-vanillin Schiff base (Scheme 1), Hpiv = pivalic acid. The ligand
L has been previously employed to fabricate different Ni(II)
and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c9nj03574e
H
2
16218 | New J. Chem., 2019, 43, 16218--16225 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
Paper
NJC
strictly equivalent. The trinuclear Ni(II) complex shows a
bent arrangement with a Ni–Ni–Ni angle of 146.27(4)1. All of
the Ni(II) centres display distorted octahedral coordination
environments with O N and O ligands for the terminal and
5
6
central Ni(II) centres, respectively. Terminal Ni(II) sites are
2
ꢀ
chelated by the L ligand, and doubly bridged to the central
2
ꢀ
Ni(II) centre through the phenoxide O of the L ligand and a
syn–syn pivalate ligand. Two water ligands complete their
coordination sphere. The Ni(II)–O bond distances involving
water ligands are the longest, 2.088(4) and 2.129(5) Å; with
Scheme 1 The Schiff-base type H
complex 1.
2
L ligand employed for the synthesis of
complexes, as well as Ni(II)/Ln(III) compounds with a variety of the remaining Ni(II)–O/Ni(II)–N bond distances ranging from
5
nuclearities.
2.031(4)–2.050(4) Å. These metric data suggest an axial distor-
Through a combined experimental and theoretical approach, tion running along the O–Ni–O axis (both O atoms from water
we attempt to understand the weak exchange interactions ligands). On the other hand, the central Ni(II) site is chelated
2
ꢀ
observed in complex 1, close to the boundary between a ferro/ twice by the phenoxide-methoxy motif of the L ligand in a
antiferromagnetic nature.
cis arrangement. The coordination sphere is completed by
both bridging pivalate ligands. Owing to the imposed crystal
symmetry, the central Ni(II) Ni–O bond distances are grouped in
three equivalent pairs corresponding to the contiguous bonds:
Results and discussion
Synthesis and structural characterization
2.014(4), 2.030(3) and 2.110(4) Å. The longest pair of bond
distances, as expected, corresponds to the Ni–O bond involving
The reaction of stoichiometric amounts of a dinuclear Ni(II)
2
ꢀ
the L ligand methoxy group. This bonding pattern indicates a
trigonal distortion instead of an axial one. Owing to the planar
pivalate precursor with the Schiff-base type H L ligand in
2
acetonitrile under mild conditions affords single crystal crops
with high reproducibility. X-ray data show that these crystals
correspond to the trinuclear Ni(II) complex 1 (Fig. 1).
It packs into the monoclinic C2/c space group with one
acetonitrile solvent molecule and two pivalic acid molecules
per molecule of complex 1 (Fig. S1, ESI†). A crystallographic
2
ꢀ
arrangement of the L ligands, the elongated O–Ni–O axis (the
one involving both water ligands) of the terminal Ni(II) sites,
makes an angle of 70.8(2) degrees (estimated through the
0
0
O(4)–Ni(2)–Ni(2) –O(4) torsion angle), running close to a perpendi-
cular arrangement.
Surprisingly, few structural characterized {Ni } complexes
imposed C rotation axis makes both the terminal Ni(II) units
3
2
sharing the m-OR/m-syn, syn-carboxylato bridge, have been
6
reported so far. The Cambridge Structural Database (CSD) only
affords six structures, most of them built upon rather similar
7
Schiff-base like ligands as H
2
L (see Table 1). The structural
3
similarity of these complexes, all showing the bent {Ni } motif,
affords a range of Ni–Ni–Ni angles from ca. 130 to 1501.
The coordinated water ligands are involved in hydrogen
bridge interactions that are intra-molecular and inter-
molecular in nature, with m-pivalate ligands in the case of
the former and free pivalic acid molecules together with the
2
ꢀ
O-phenoxide atoms of the L
ligand of the neighbouring
complex in the latter (Fig. S2, S3 and Table S2, ESI†). The
free pivalic acid molecule in the packing is also held by
hydrogen interactions with the water ligand and the non-
bridging phenoxide moiety of the L ligand (Fig. S2, ESI†).
On the other hand, the acetonitrile solvent molecules show
Fig. 1 A ball and stick molecular representation of the crystal structure of
complex 1. H atoms have been removed for clarity. Colour code: Green:
Ni; red: O; blue: N; grey: C.
2
ꢀ
3
Table 1 Structural data for {Ni } complexes with a bent arrangement and a m-OR/m-syn, syn-carboxylato bridge
CSD code
name
Ni–Ni–Ni
angle/degrees
Mean Ni–Ophenolate
bond distance/Å
Ni–Ophenolate–Ni
angle/degrees
Mean Ni–Ocarboxylate
bond distance/Å
RCOO–NiOphNi
torsion angle/degrees
Ref.
BENWEX
BENWIB
GAHTIU
KOMREL
PEMREH
ZEKREP
1
131
132
134
146
135
149
146
2.066
2.063
2.068
2.031
2.109
2.012
2.040
124
123–125
118
118
123
118
117
2.048
2.037
2.040
2.025
2.078
2.019
2.029
30
32
9
19
29
33
24
7a
7a
7b
7c
7d
7e
This work
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16218--16225 | 16219

View Article Online
NJC
Paper
non-covalent short contact interactions with the free pivalic
2
ꢀ
acid tert-butyl moieties and the methyl moieties of the L
ligand (Fig. S4, ESI†).
When looking in detail at the crystal structure packing, it is
observed that the molecules of complex 1 run along the c-axis
held by the previously described H-interaction. On the other
hand, the acetonitrile and pivalic acid free molecules fill the
channels in between (Fig. S5, ESI†). The closest intra-molecular
Ni–Ni distance along this c-axis direction is 5.345(2) Å, longer
than the intra-molecular Ni–Ni distance of 3.4800(13) Å.
Scheme 2 Exchange interactions considered in the spin Hamiltonian of
eqn (1).
Magnetic properties: experimental and quantum chemical
calculations
In order to evaluate the spin ground state, as well as the
We performed a simultaneous data fitting of susceptibility
8
magnitude and nature of the Ni(II)–Ni(II) exchange interactions, and magnetization employing the PHI package. In order to
we performed DC magnetic measurements for complex 1. The reach a satisfactory agreement, the following spin Hamiltonian
temperature dependence of magnetic susceptibility in the range was required:
X
ꢀ
ꢁ
X
2
–300 K shows a profile compatible with the dominant antiferro-
2
H^ ¼ gbB
Si ꢀ 2J S1S2 þ S2S3 þ D
^
^ ^
^ ^
Sz;i
^
(1)
magnetic exchange interactions between the Ni(II) centres (Fig. 2).
At 300 K, the wT value of 4.204 cm K mol is consistent
with the expected value for three isolated Ni(II) ions (S = 1) with
i
i
3
ꢀ1
It includes a unique exchange interaction between the
a g value close to 2.3 of ca. 3.9 cm K mol . This value smoothly central and terminal Ni(II) ions (Scheme 2) (in agreement with
drops to reach 3.783 cm K mol at 30 K where it abruptly the symmetry of complex 1) as well as a unique axial local ZFS
diminishes to a final value of 1.719 cm K mol at 2 K. Of term (as including different ZFS terms for central and terminal
course, the single ion zero field splitting (ZFS) contributions sites would certainly result in over-parameterization problems).
3
ꢀ1
3
ꢀ1
3
ꢀ1
from the Ni(II) sites are probably also contributing at low
This model proved to be the simplest, as well as the
temperature in addition to the Ni(II)–Ni(II) antiferromagnetic minimum required to provide a good agreement with the
exchange interaction. When looking at the reduced magnetization experimental data (Fig. 2 and Fig. S6, ESI†). Two sets of best
ꢀ
1
plots achieved in the range 2–10 K and up to a 70 kOe magnetic fitting parameters were obtained: g = 2.31; J = ꢀ0.32 cm
;
ꢀ
1
ꢀ1
ꢀ1
field, neither saturation is observed nor isotherms superposition D = ꢀ8.4 cm and g = 2.31; J = ꢀ0.48 cm ; D = 4.2 cm . A TIP
Fig. 2). This is in line with possible ZFS contributions and/or low- (temperature independent paramagnetism) contribution of
(
ꢀ
6
3
ꢀ1
lying excited spin multiplets owing to weak exchange interactions. 600 ꢁ 10 cm K mol was also included. The data fitting
residuals are lower for a negative D value, whereas at the same
time a strong correlation is found between the J and D para-
meters (Fig. S7, ESI†). Despite the uncertainty around these
parameter values, there is no doubt that ZFS dominates over
the exchange interactions, precluding the strong exchange regime
and making it harder to precisely determine the exchange inter-
ꢀ
1
action parameters. With a J value of around ꢀ0.5 cm , a triplet
ꢀ
1
ground state is found with a singlet at only 1 cm and the entire
ꢀ
1
spin ladder closely packed in just 5 cm
.
In order to support the experimental results, we performed
quantum chemical calculations of the spin Hamiltonian para-
meters (see Quantum chemical computations section). For the
exchange interaction parameter we relied on the widely used
broken-symmetry (BS) approach which we have successfully
2g,9
employed in related systems.
In the case of the computed
g and ZFS tensors we performed CASSCF-SA-SOC (state averaged
8
spin–orbit coupled) calculations based on all d configuration
microstates, over the crystallographic individual Ni(II) sites
(central and terminal positions).
The calculated ZFS parameters, as well as the main values of
Fig. 2 Top: wT versus T data for complex 1 under a 1000 Oe magnetic
field. Bottom: Reduced magnetization data for complex 1 in the range of
the g-tensors show an excellent agreement with the experimental
ones (Table 2), suggesting a negative value for D which was also
presumed from the DC magnetic data fit residuals. The support
from the quantum calculations of a sizeable D value, provides
2
–10 K under magnetic fields of up to 70 kOe. Open symbols: experi-
mental data; full lines: simulations employing best fitting parameters with a
negative D value, as described in the text.
1
6220 | New J. Chem., 2019, 43, 16218--16225 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
Paper
NJC
3
Table 2 Magnetic data for {Ni } complexes with a bent arrangement and a m-OR/m-syn, syn-carboxylate bridge
ꢀ1
ꢀ1
ꢀ1
2
CSD code name Experimental SH parameters J, D/cm Computed SH parameters J, D/cm Computed J/cm in the model OH–OH complex
BENWEX
BENWIB
GAHTIU
—
—
—
—
—
—
—
g = 2.08
J = ꢀ0.85
g = 2.01
J— = ꢀ1.28
g = 2.31
J = 0.51
J = ꢀ0.72/ꢀ0.57
KOMREL
J = ꢀ2.85
—
PEMREH
ZEKREP
—
J = 0.49
—
J = ꢀ0.54
1
g = 2.31
g
g
x,y,z = (2.31 2.32 2.38) Ni(1)
x,y,z = (2.37 2.32 2.26) Ni(2)
J = ꢀ2.5
J = ꢀ0.32 (ꢀ0.48)
D = ꢀ8.4 (4.2)
J = ꢀ0.95
D = ꢀ10.0 Ni(1)
|
E/D| = 0.02 Ni(1)
D = ꢀ11.2 Ni(2)
|
E/D| = 0.32 Ni(2)
robustness to the J exchange parameter value extracted, which is not which the exchange interaction is ferromagnetic. Luckily, this com-
a trivial task when the strong coupled regime is not valid ( J { D). pound is quite similar from a structural point of view to complex 1
For computation of the isotropic exchange interaction para- offering us two examples with opposite J natures to compare.
meter, J, we employed two different BS states, obtained using
No correlation was found between the J values and
the subsequent spin-flipping of the central and terminal Ni(II) Ni–Ophenolate distances or the Ni–Ophenolate–Ni angles in all
spins (Fig. S8, ESI†). However, owing to the observed symmetry complexes shown in Table 2. Thus, we decided to test the
that makes both the terminal Ni(II) groups equivalent, it would carboxylate role in the exchange interaction parameter as the
have been sufficient with a unique BS state energy calculation. second bridging ligand. Clearly, it cannot be mediating an
Hence, we obtained a mean calculated value (from both BS state additional antiferromagnetic pathway, therefore it is feasible
ꢀ
1
energies), J = ꢀ0.95 cm , which is in good agreement with the that it is exerting a counter-complementary effect. This effect is
experimental result (Table 2). Among the surveyed structural well known in Cu(II) complexes with the same combination of
1
0
related bent Ni(II) complexes (Table 1), a small number were bridging ligands, but there are no reports for Ni(II) systems
found in which the magnetic properties have been reported with the m-phenoxo-m-carboxylate bridge. In order to verify this
(Table 2). All of them also show an anti-ferromagnetic exchange hypothesis, we performed BS calculations on the two complexes
interaction with a unique exception (KOMREL) that exhibits a (1 and KOMREL, Table 2) substituting the bridging carboxylate
ferromagnetic positive J value. Among the Ni(II) systems bearing a ligand with aqua and hydroxo ligands placed at the same
phenolate bridge, only the bis-m -phenolate ones have been deeply Ni–O bond distances (Fig. S9, ESI†). Using this strategy the
2
analysed through magneto-structural correlations. In these unique bridging ligand is split into two independent ones
systems it has been shown that the Ni–O–Ni angle dominates disassembling any counter-complementary effect. Both of the
the nature and magnitude of the exchange interaction. The critical calculated J parameters (Table 2) afford negative values
angle for moving from a ferromagnetic to an antiferromagnetic supporting the feasibility of the counter-complementarity
2i,j,l,n,o
interaction is close to 941.
In all complexes related to our hypothesis, as the exchange coupling parameter becomes more
reported compound 1 (Table 1), the Ni–O–Ni angle was well above negative when the carboxylate bridge is removed. Hence, the
this threshold, indicating an antiferromagnetic interaction. It almost fixed Ni–Ophenolate–Ni angle at ca. 117–1181 observed in
must be stressed at this point that these complexes are not bis- complexes in Table 1 provides an antiferromagnetic pathway
2
n
m -phenolate, but single m -phenolate ones. In fact, if the values (according to the double m-phenolate existing correlations )
2
2
for the J magnitude between these two types of Ni(II) complex while the carboxylate bridge suppresses this contribution
families are compared, a factor of ten is observed in favour of the through a counter-complementarity effect. In the case of the
bis-phenolate systems. Nevertheless, it seems reasonable to expect complex which has an overall ferromagnetic interaction, this
a similar Ni–O–Ni angle correlation for the single phenolate suppression must be improved in comparison with the other
systems. Following this reasoning, all anti-ferromagnetic J para- complexes showing anti-ferromagnetic interactions. This counter-
meters should be observed in the complexes shown in Table 2, complementarity role of the carboxylate bridge has also recently
however there is one case in which this is not true.
In order to gain a deeper insight into this singular instance systems.
we performed BS calculations over crystal structure geometries From the inspection of the magnetic orbitals arising from
been observed in Cr(III) double m-alkoxide-m-carboxylate dinuclear
11
for all complexes in Table 2 for which the experimental the BS calculations (Fig. 3) it can be observed that the torsion
magnetic data was available, at the same theory level employed for angle between the plane containing the carboxylate –COO
complex 1. The obtained computed J parameters agree with the group and the plane containing the Ni–Ophenolate–Ni group, as
experimental ones. Thus, it is further confirmed the one case in well as the mean Ni–Ocarboxylate bond distance, are the critical
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16218--16225 | 16221

View Article Online
NJC
Paper
Fig. 3 Magnetic orbital iso-surface contours (0.02 a.u.) arising from COT for both BS calculations in complex 1. The different colour combinations
correspond to a–b pairs with overlapping that is significantly different from unity.
Scheme 3 Structural parameters influencing the counter-complementarity effect of the bridging carboxylate ligand.
parameters for the counter-complementarity contribution that the considerable single ion ZFS contribution of Ni(II) makes
(Table 1 and Scheme 3).
determination of the precise exchange interaction parameters a very
In fact, these two parameters reasonably correlate with the challenging task. As the currently reported complexes feeding these
observed tendency of the J values: the shorter Ni–Ocarboxylate results are still very scarce, more related novel complexes are needed
bonds, as well as the larger torsion angles indicate a ferromagnetic to further test this preliminary hypothesis.
overall interaction (Fig. S10, ESI†). On the other hand, the anti-
ferromagnetic contribution through the phenoxide bridge is
explained in terms of a favourable overlap of at least one pair of Experimental section
magnetic orbitals. Nevertheless, many more examples with
theoretical and experimental data must be accumulated before
a magneto-structural correlation can be determined.
Material and physical measurements
The complex [Ni (m-OH )(m-piv) (piv) (Hpiv) ], piv = trimethylacetate,
2
2
2
2
4
12
was prepared following a previously reported procedure. The
ligand H L was prepared by standard reported procedures for this
2
type of Schiff-base compound, by refluxing equimolar amounts of
the starting materials o-vanillin and o-aminophenol in methanol
and crystallizing the product by cooling.
Conclusions
3
We have prepared and structurally characterized a novel {Ni }
All other chemicals were reagent grade and were used as
received without further purification. Elemental analysis for C,
H and N was performed using a Carlo Erba 1108 analyser.
Magnetic measurements were performed using a Quantum
Design MPMS XL-7 SQUID magnetometer. All experimental
magnetic data were corrected for the diamagnetism of the
sample holders and for the constituent atoms (Pascal’s tables).
DC measurements were conducted from 2 to 300 K at 1 kOe and
at 2 K in the range 1–70 kOe.
bent complex, bearing the double m-phenoxide-m-carboxylate
bridge. Notably, very few reported examples sharing this motif
can be found, and even less of them have been studied from the
magnetic behaviour aspect; therefore, we decided to make a
step forward in this direction. Combining DC magnetic experi-
mental data, as well as quantum computations, we have shown
that the nature of the weak exchange interaction between the
Ni(II) ions seems to be established by counter-complementarity
effects. Although the phenoxide ligand mediates a modest
magnetic orbital overlap and promotes an antiferromagnetic
pathway, the carboxylate bridge exerts counter-complementarity
Preparation of complex Ni
3
(L
2
)(piv)
2
(H
2
O)
4
3
ꢂ2HpivꢂCH CN (1)
and favours ferromagnetic exchange. BS calculations support this 0.1 g (0.105 mmol) of [Ni (m-OH )(m-piv) (piv) (Hpiv) ] was
2
2
2
2
4
idea and the variation of the main structural parameters (torsion dissolved at room temperature under an open atmosphere in
angle and bond distances) around the carboxylate bridge suggest 10 ml of acetonitrile. After complete dissolution resulting in a
correlation with the nature of the J parameter. It is worth remarking green solution, 0.025 g (0.105 mmol) of solid H L was added
2
1
6222 | New J. Chem., 2019, 43, 16218--16225 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
Paper
NJC
and stirring was continued for one hour. The solution colour molecule appears to be disordered around two positions related by a
changed to brownish yellow. It was filtered to remove small crystallographic C axis. It was then refined with 0.5 occupancy
amounts of undissolved materials and left standing undisturbed numbers.
in a closed vial. After a week, yellow crystals suitable for X-ray Final crystallographic data and values of R and wR are listed
2
1
diffraction (XRD) measurements were obtained. After picking one in Table 3. CCDC 1938756 contains the supplementary crystallo-
for single crystal XRD collection, the remaining crystals were graphic data for this paper.†
filtered and washed with acetonitrile. After air drying, 0.034 g
Quantum chemical calculations
were obtained. Yield: 53%.
ꢀ1
71 3 3
Anal. calcd for C50H N Ni O18 (1219.25 g mol ) C: 51.0, For the computation of the exchange interaction J parameters,
1
7
H: 6.1, N: 3.6 found: C: 51.1, H: 5.9, N: 4.4.
the ORCA program package was employed. Single point
calculations for the high-spin state (HS) and the broken-
symmetry (BS) states at the X-ray geometry were carried out at
DC magnetic measurements
X-ray structure determination. The crystal structure of compound the B3LYP level of density functional theory (DFT) employing
1
was determined with an Oxford Xcalibur, Eos, Gemini CCD area- the def2-TZVP Ahlrichs basis set for all atoms and taking
detector diffractometer using graphite-monochromated Mo-Ka advantage of the RI (resolution of identity) approximation.
radiation (l = 0.71069 Å) at 298 K. Crystals were directly obtained The self-consistent field (SCF) calculations were of the spin-
ꢀ7
from the synthetic procedure. Data was corrected for absorption polarized type and were tightly converged (10 Eh in energy,
ꢀ
6
ꢀ6
with CrysAlisPro, Oxford Diffraction Ltd, Version 1.171.33.66, apply- 10 in the density change and 10 in the maximum element
ing an empirical absorption correction using spherical harmonics, of the DIIS error vector).
13
implemented in the SCALE3 ABSPACK scaling algorithm. The
The methodology applied here relies on the BS formalism,
structures were solved using direct methods with SHELXT and originally developed by Noodleman for SCF methods, which
refined by full-matrix least-squares on F with SHELXL-2014 under involves a variational treatment within the restrictions of a
the WinGX platform. Hydrogen atoms were added geometrically single spin-unrestricted Slater determinant built upon using
14
18
2
15
16
and refined as riding atoms with a uniform value of U , with the different orbitals for different spins. This approach has been
iso
1
9
exception of aqua ligands and the pivalic acid solvent molecule, later applied within the frame of DFT. The HS and BS energies
whose H atoms were located in the difference map and further were then combined to estimate the exchange coupling para-
refined with O–H constrained distances. The solvent acetonitrile meter J involved in the widespread used Heisenberg–Dirac–van
Table 3 Crystallographic data for complex 1
1
Empirical formula
Formula weight
T (K)
Crystal system
Space group
a (Å)
b (Å)
c (Å)
a (1)
50 71 3 3 18
C H N Ni O
1178.22
298 (2)
Monoclinic
C
2/c
17.893(3)
16.511(2)
21.476(4)
90
b (1)
110.56(2)
90
5941(2)
4
1.317
1.008
g (1)
V (Å )
3
Z
D
ꢀ3
calc (mg m
)
ꢀ1
Absorption coefficient (mm
F(000)
)
2480
l (Å)
0.71069
3.64–27.0
y range data collection (1)
Index ranges
ꢀ22 r h r 19
ꢀ
19 r k r 20
27 r l r 26
ꢀ
Reflections collected/unique
13 797/5905
0.0761
R
int
Observed reflections [I 4 2s(I)]
Completeness (%)
3591
93.2
Maximum/minimum transmission
Data/restraints/parameters2
Goodness-of-fit (GOF) on F
Final R-index [I 4 2s(I)]/all data
wR index [I 4 2s(I)]/all data
1.000/0.361
5905/22/364
1.023
0.0761/0.2568
0.2030/0.2568
1.166 and ꢀ0.459
ꢀ3
Largest peak and hole (e A
)
2
2
2
2
2
Weights, w
o o c
1/[s (F ) + (0.1399P) + 0.6906P] where P = (F + 2F )/3
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16218--16225 | 16223

View Article Online
NJC
Paper
Vleck Hamiltonian. We used the method proposed by Ruiz and
co-workers, in which the following equation is applied:
and S. Mohanta, Inorg. Chem., 2011, 50, 7257–7267;
(d) N. Berg, T. Rajeshkumar, S. M. Taylor, E. K. Brechin,
G. Rajaraman and L. F. Jones, Chem. – Eur. J., 2012, 18,
2
0
E
BS ꢀ EHS = 2J12(2S + S ), with S o S
1
S
2
1
2
1
(2)
5906–5918; (e) W. P. Barros, R. Inglis, G. S. Nichol,
T. Rajeshkumar, G. Rajaraman, S. Piligkos, H. O. Stumpf
and E. K. Brechin, Dalton Trans., 2013, 42, 16510; ( f ) J. P. S.
Walsh, S. Sproules, N. F. Chilton, A. Barra, G. A. Timco,
D. Collison, E. J. L. McInnes and R. E. P. Winpenny, Inorg.
Chem., 2014, 53, 8464–8472; (g) A. V. Funes, L. Carrella,
L. Sorace, E. Rentschler and P. Albor ´e s, Dalton Trans., 2015,
We have calculated the different spin topologies for the BS
nature (Fig. S8, ESI†) by alternately flipping spin on the
different metal sites. The exchange coupling constants J were
i
obtained after considering the individual pair-like components
spin interactions involved in the description of the different BS
states by solving a set of linear equations. In order to visualize
the magnetic orbitals involved in the exchange interactions we
performed the corresponding orbital transformations (COT)
4
4, 2390–2400; (h) E. A. Suturina, D. Maganas, E. Bill,
M. Atanasov and F. Neese, Inorg. Chem., 2015, 54,
948–9961; (i) I. Oyarzabal, J. Ruiz, A. J. Mota, A. Rodr ´ı guez-
Di ´e guez, J. M. Seco and E. Colacio, Dalton Trans., 2015, 44,
825–6838; ( j) F. Tori ´c , G. Pavlovi ´c , D. Paji ´c , M. Cindri ´c and
2
1
9
over the BS solutions as implemented in ORCA.
To compute the local g and ZFS tensors, the MOLCAS
2
2
6
package (as the MOLCAS@UU version) was employed. The
calculations of these parameters for the two different Ni(II) sites
in the trinuclear complex were performed by replacing the
remaining Ni(II) sites with the diamagnetic Zn(II) cation. We
performed these computations at the complete active space
self-consistent field (CASSCF) level, and the spin–orbit coupling
K. Zadro, CrystEngComm, 2018, 20, 3917–3927; (k) M. K. Singh
and G. Rajaraman, Inorg. Chem., 2019, 58, 3175–3188;
(
l) F. Tori ´c , G. Pavlovi ´c , D. Paji ´c , T. Hrenar, K. Zadro and
M. Cindri ´c , Inorg. Chim. Acta, 2019, 484, 457–463;
m) S. M. Gorun and S. J. Lippard, Inorg. Chem., 1991, 30,
625–1630; (n) K. K. Nanda, L. K. Thompson, J. N. Bridson and
K. Nag, J. Chem. Soc., Chem. Commun., 1994, 8, 1337–1338;
o) M. A. Halcrow, J. S. Sun, J. C. Huffman and G. Christou,
(
1
(SOC) was introduced in a second step using the spin orbital
restricted-active-space state-interaction (SO-RASSI) method. We
8
(
performed a states-average approach including the full d
Inorg. Chem., 1995, 34, 4167–4177; (p) H. Weihe and
H. U. G u¨ del, J. Am. Chem. Soc., 1998, 120, 2870–2879;
microstates, 10 triplets and 15 singlets in the SO-RASSI step.
For the CASSCF calculation, the active space contained eight
electrons in five orbitals (the five 3d orbitals). Relativistic
atomic natural orbital (ANO-RCC) basis sets were used with
the following contraction scheme: 7s6p4d3f2g1h (VQZP) for Ni
and Zn, 4s3p2d (VTZ) for O and N, 3s2p (VDZ) for C and 2s
(q) T. K. Karmakar, B. K. Ghosh, A. Usman, H. K. Fun,
E. Rivi `e re, T. Mallah, G. Arom ´ı and S. K. Chandra, Inorg.
Chem., 2005, 44, 2391–2399; (r) S. Triki, C. J. G ´o mez-Garc ´ı a,
E. Ruiz and J. Sala-Pala, Inorg. Chem., 2005, 44, 5501–5508;
(s) R. Inglis, L. F. Jones, C. J. Milios, S. Datta, A. Collins,
(VDZ) for H. Final g and ZFS tensors were obtained through the
S. Parsons, W. Wernsdorfer, S. Hill, S. P. Perlepes, S. Piligkos
and E. K. Brechin, Dalton Trans., 2009, 3403; (t) R. Costa, I. de,
P. R. Moreira, S. Youngme, K. Siriwong, N. Wannarit and
F. Illas, Inorg. Chem., 2010, 49, 285–294.
SINGLE-ANISO MOLCAS routine.
Conflicts of interest
3
4
5
D. Gatteschi, R. Sessoli and J. Villain, Molecular Nanomagnets,
Oxford University Press, 2006.
O. Kahn, Molecular Magnetism, Wiley-VCH Verlag GmbH &
Co. KGaA, 1993.
(a) K. C. Mondal, G. E. Kostakis, Y. Lan, W. Wernsdorfer,
C. E. Anson and A. K. Powell, Inorg. Chem., 2011, 50,
There are no conflicts of interest to declare.
Acknowledgements
We gratefully acknowledge UBA, ANPCYT and CONICET for
funding resources. PA is a staff member of CONICET. The
authors gratefully acknowledge computing time granted on
the supercomputer Mogon at Johannes Gutenberg University
Mainz (hpc.uni-mainz.de).
11604–11611; (b) H. Ke, L. Zhao, Y. Guo and J. Tang, Inorg.
Chem., 2012, 51, 2699–2705; (c) S. Bag, P. K. Bhaumik,
S. Jana, M. Das, P. Bhowmik and S. Chattopadhyay, Poly-
hedron, 2013, 65, 229–237; (d) S. Saha, S. Pal, C. J. G o´ mez-
Garc ´ı a, J. M. Clemente-Juan, K. Harms and H. P. Nayek,
Polyhedron, 2014, 74, 1–5; (e) H. Ke, S. Zhang, W. Zhu, G. Xie
and S. Chen, J. Coord. Chem., 2015, 68, 808–822;
( f ) K. Griffiths, C. W. D. Gallop, A. Abdul-Sada, A. Vargas,
O. Navarro and G. E. Kostakis, Chem. – Eur. J., 2015, 21,
6358–6361; (g) K. Griffiths, P. Kumar, J. D. Mattock,
A. Abdul-Sada, M. B. Pitak, S. J. Coles, O. Navarro,
A. Vargas and G. E. Kostakis, Inorg. Chem., 2016, 55,
6988–6994; (h) K. Griffiths, V. N. Dokorou, J. Spencer,
A. Abdul-Sada, A. Vargas and G. E. Kostakis, CrystEngComm,
2016, 18, 704–713; (i) S. Saha, S. Jana, S. Gupta, A. Ghosh
and H. P. Nayek, Polyhedron, 2016, 107, 183–189;
References
1
J. Ferrando-Soria, J. Vallejo, M. Castellano, J. Mart ´ı nez-Lillo,
E. Pardo, J. Cano, I. Castro, F. Lloret, R. Ruiz-Garc ´ı a and
M. Julve, Coord. Chem. Rev., 2017, 339, 17–103.
2
(a) V. H. Crawford, H. W. Richardson, J. R. Wasson,
D. J. Hodgson and W. E. Hatfield, Inorg. Chem., 1976, 15,
2107–2110; (b) J. Glerup, D. J. Hodgson, E. Pedersen,
A. Haaland, B. E. R. Schilling, R. Seip and K. Taugbøl, Acta
Chem. Scand., Ser. A, 1983, 37, 161–164; (c) S. Sasmal,
S. Hazra, P. Kundu, S. Dutta, G. Rajaraman, E. C. Sa n˜ udo
1
6224 | New J. Chem., 2019, 43, 16218--16225 This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
Paper
NJC
(
2
j) K. Griffiths, A. Escuer and G. E. Kostakis, Struct. Chem.,
016, 27, 1703–1714.
R. A. Porter, J. F. Richardson, D. N. Hendrickson and
R. M. Buchanan, Inorg. Chem., 2002, 35, 3325–3334.
6
7
C. R. Groom, I. J. Bruno, M. P. Lightfoot and S. C. Ward, Acta 11 H. W. L. Fraser, L. Smythe, S. Dey, G. S. Nichol, S. Piligkos,
Crystallogr., Sect. B: Struct. Sci., Cryst. Eng. Mater., 2016, 72,
71–179.
G. Rajaraman and E. K. Brechin, Dalton Trans., 2018, 47,
8100–8109.
1
(a) H. Adams, D. E. Fenton and P. E. McHugh, Inorg. Chem. 12 G. Chaboussant, R. Basler, H. G u¨ del, S. Ochsenbein, A. Parkin,
Commun., 2004, 7, 147–150; (b) Y. Y. Hui, H. M. Shu,
S. Parsons, G. Rajaraman, A. Sieber, A. A. Smith, G. A. Timco
H. M. Hu, J. Song, H. L. Yao, X. Le Yang, Q. R. Wu,
and R. E. P. Winpenny, Dalton Trans., 2004, 2758–2766.
M. L. Yang and G. L. Xue, Inorg. Chim. Acta, 2010, 363, 13 SCALE3 ABSPACK: empirical absorption correction, CrysAlis –
238–3243; (c) X. Qin, Y. Ji, Y. Gao, L. Yan, S. Ding, Y. Wang Software package, 2006.
and Z. Liu, Z. Anorg. Allg. Chem., 2014, 640, 462–468; (d) 14 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Adv., 2015,
C.-Y. Li, Y.-C. Su, C.-H. Lin, H.-Y. Huang, C.-Y. Tsai, T.-Y. Lee 71, 3–8.
and B.-T. Ko, Dalton Trans., 2017, 46, 15399–15406; 15 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
3
(e) M. Liu, H. Yu, Y. Wang and Z. Liu, Inorg. Chem. Commun.,
2008, 64, 112–122.
2017, 86, 281–284.
16 L. J. Farrugia, J. Appl. Crystallogr., 2012, 45, 849–854.
8
9
N. F. Chilton, R. P. Anderson, L. D. Turner, A. Soncini and 17 F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2,
K. S. Murray, J. Comput. Chem., 2013, 34, 1164–1175. 73–78.
(a) P. Albor ´e s, J. Seeman and E. Rentschler, Dalton Trans., 18 L. Noodleman, J. Chem. Phys., 1981, 74, 5737–5743.
009, 7660–7668; (b) P. Albor ´e s and E. Rentschler, Polyhe- 19 L. Noodleman and E. J. Baerends, J. Am. Chem. Soc., 1984,
dron, 2009, 28, 1912–1916; (c) P. Albor ´e s and E. Rentschler,
106, 2316–2327.
Inorg. Chem., 2010, 49, 8953–8961; (d) P. Albor ´e s, C. Plenk 20 (a) E. Ruiz, J. Cano, S. Alvarez and P. Alemany, J. Comput.
2
and E. Rentschler, Inorg. Chem., 2012, 51, 8373–8384; (e) I. C.
Lazzarini, L. Carrella, E. Rentschler and P. Albor ´e s, Polyhedron,
Chem., 1999, 20, 1391–1400; (b) E. Ruiz, A. Rodr ´ı guez-Fortea,
J. Cano, S. Alvarez and P. Alemany, J. Comput. Chem., 2003,
24, 982–989.
2
012, 31, 779–788; ( f ) P. Albor ´e s and E. Rentschler, Dalton
Trans., 2013, 42, 9621–9627; (g) I. C. Lazzarini, A. V. Funes, 21 F. Neese, J. Phys. Chem. Solids, 2004, 65, 781–785.
L. Carrella, L. Sorace, E. Rentschler and P. Albor ´e s, Eur. 22 F. Aquilante, J. Autschbach, R. K. Carlson, L. F. Chibotaru,
J. Inorg. Chem., 2014, 2561–2568; (h) C. Sarto, M. Rouzi `e res,
J. L. Liu, H. Bamberger, J. Van Slageren, R. Cl ´e rac and
P. Albor ´e s, Dalton Trans., 2018, 47, 17055–17066.
0 (a) G. Christou, E. Libby, K. Folting, J. C. Huffman,
R. J. Webb, D. N. Hendrickson and S. P. Perlepes, Inorg.
Chem., 1990, 29, 3657–3666; (b) K. S. B u¨ rger, P. Chaudhuri
and K. Wieghardt, J. Chem. Soc., Dalton Trans., 1996, 1,
M. G. Delcey, L. De Vico, I. Fdez. Galv ´a n, N. Ferr ´e , L. M. Frutos,
L. Gagliardi, M. Garavelli, A. Giussani, C. E. Hoyer, G. Li
Manni, H. Lischka, D. Ma, P. Å. Malmqvist, T. M u¨ ller,
A. Nenov, M. Olivucci, T. B. Pedersen, D. Peng, F. Plasser,
B. Pritchard, M. Reiher, I. Rivalta, I. Schapiro, J. Segarra-Mart ´ı ,
M. Stenrup, D. G. Truhlar, L. Ungur, A. Valentini, S. Vancoillie,
V. Veryazov, V. P. Vysotskiy, O. Weingart, F. Zapata and
R. Lindh, J. Comput. Chem., 2016, 37, 506–541.
1
247–248; (c) H. Nie, S. M. J. Aubin, M. S. Mashuta,
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019 New J. Chem., 2019, 43, 16218--16225 | 16225