Synthesis, Crystal Structures, Magnetic Properties, and Fluorescence of Two Heptanuclear CoIII4LnIII3 Compounds (Ln = GdIII, DyIII): Multiple Relaxation Dynamics in the DyIII Analogue

Article abstract of DOI:10.1002/ejic.201900383

This report describes the synthesis, crystal structures, magnetic and spectroscopic properties of two heptanuclear Co^III₄Ln^III₃ systems of composition [Co^III₄Gd^III₃L₄(μ₄-O)₂(μ-OMe)₂(μ_1,3-OAc)₄(H₂O)₂(NO₃)₂]·NO₃·2CH₃OH·2H₂O (1) and [Co^III₄Dy^III₃L₄(μ₄-O)₂(μ-OMe)₂(μ_1,3-OAc)₄(H₂O)₂(NO₃)₂]·NO₃·3CH₃OH·1.5H₂O (2), where H₂L is [1+1] condensation product of 3-methoxysalicylaldehyde and 2-amino-2-methyl-1-propanol. The heptametallic cores in 1 and 2 may be taken as consisted of two symmetry related trinuclear Co^IIILn^IIICo^III moieties and a central lanthanide center in such a way that each of the three metal ions of each of the two trinuclear moieties are bridged with the central lanthanide through bridging oxo/alkoxo/acetate ligands. DC magnetic measurements of 1 and 2 reveal very weak Gd^III···Gd^III antiferromagnetic interaction (J = –0.01 cm^–1) in 1 and significant anisotropy in 2. Variable-temperature and variable-frequency ac susceptibility measurements of 2 reveal that it is a zero-field SMM with multiple relaxation dynamics. Above 5 K, two well resolved relaxation channels are observed; U_eff and τ₀ are 51.4 cm^–1 and 9.40 × 10^–7 s for the slow relaxation channel and 24.3 cm^–1 and 6.91 × 10^–7 s for the fast relaxation channel. The characteristic Dy^III based transitions, ⁴F_9/2→⁶H_15/2, ⁴F_9/2→⁶H_13/2 and ⁴F_9/2→⁶H_11/2, can be well assigned in the fluorescence spectrum of 2 at solid state. All in all, compound 2 is a bifunctional molecular material.

Full text of DOI:10.1002/ejic.201900383

A Journal of
Accepted Article
Title: Synthesis, Crystal Structures, Magnetic and Fluorescence
Properties of Two Heptanuclear CoIII4LnIII3 Compounds (Ln=
GdIII, DyIII): Multiple Relaxation Dynamics in DyIII Analogue
Authors: Shuvayan Roy, Nairita Hari, and Sasankasekhar Mohanta
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.201900383
Link to VoR: http://dx.doi.org/10.1002/ejic.201900383

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
Synthesis, Crystal Structures, Magnetic and Fluorescence Properties of
III
III
III
III
Two Heptanuclear Co Ln
Compounds (Ln= Gd , Dy ): Multiple
4
3
III
Relaxation Dynamics in Dy Analogue
Shuvayan Roy, Nairita Hari and Sasankasekhar Mohanta*
S. Roy, N. Hari and Prof. S. Mohanta
Department of Chemistry, Inorganic Chemistry Section, University of Calcutta
9
2 A. P. C Road, Kolkata 700 009, India
E-mail: sm_cu_chem@yahoo.co.in.
Fax: 91-33-23519755
http://www.caluniv.ac.in/academic/department/Chemistry/Sasankasekhar-Mohanta.pdf
۩ Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejic.201900383
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
This report describes the synthesis, crystal structures, magnetic and spectroscopic
III
III
III
III
properties of two heptanuclear Co Ln systems of composition [Co Gd L (µ -O) (µ-
4
3
4
3
4
4
2
III
III
OMe) (µ -OAc) (H O) (NO ) ]·NO ·2CH OH·2H O (1) and [Co Dy L (µ -O) (µ-
2
1,3
4
2
2
3 2
3
3
2
4
3
4
4
2
OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2), where H2L is [1+1]
2
1,3
4
2
2
3 2
3
3
2
condensation product of 3-methoxysalicylaldehyde and 2-amino-2-methyl-1-propanol. The
heptametallic cores in 1 and 2 may be taken as consisted of two symmetry related trinuclear
III
III
III
Co Ln Co moieties and a central lanthanide center in such a way that each of the three
metal ions of each of the two trinuclear moieties are bridged with the central lanthanide
through bridging oxo/alkoxo/acetate ligands. DC magnetic measurements of 1 and 2 reveal
III
III
‒1
very weak Gd ···Gd antiferromagnetic interaction (J = ‒0.01 cm ) in 1 and significant
anisotropy in 2. Variable-temperature and variable-frequency ac susceptibility measurements
of 2 reveal that it is a zero-field SMM with multiple relaxation dynamics. Above 5 K, two
‒
1
‒7
well resolved relaxation channels are observed; U and τ0 are 51.4 cm and 9.40 × 10 s for
eff
‒
1
‒7
the slow relaxation channel and 24.3 cm and 6.91 × 10 s for the fast relaxation channel.
III
4
6
4
6
4
6
The characteristic Dy based transitions, F → H , F → H and F → H11/2, can be
9
/2
15/2
9/2
13/2
9/2
well assigned in the fluorescence spectrum of 2 at solid state. All in all, compound 2 is a
bifunctional molecular material.
Introduction
III
IV
It was observed in the early 1990s that a dodecanuclear Mn Mn cluster (Mn12),
8
4
[
1]
the first reported single molecule magnet (SMM), exhibits slow relaxation of magnetization
after the removal of the applied magnetic field. Magnetic bistability and sufficient energy
barrier (U ) for magnetization reversal in this molecule take place due to overall uniaxial
eff
anisotropy (negative D value), which, in turn, arises because of inherent single-ion zero-field
III
splitting of Mn . In contrast to the bulk magnets where magnetism is a bulk phenomenon,
the magnetic memory of SMMs originate solely from the molecule itself, and, therefore such
magnetic systems have potential applications in advanced technological areas such as
[
2]
[3]
molecular spintronics,
ultrahigh density magnetic information storage,
quantum
[
4]
[5]
[6]
computing, magneto-chiral dichroism and magneto-electric effects, etc. Notably,
blocking temperature (T ) below which the magnetic hysteresis takes place and U are the
B
eff
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
3
two parameters to govern the efficiency of an SMM; the larger the values of these two
parameters the better is the SMM.
As the anisotropy is the key factor for magnetization reversal, lanthanides are much
[
7‒9]
more promising
than 3d metal ions for getting better SMMs because most of the
lanthanides have large unquenched orbital angular momentum, which gives rise to large
anisotropy. However, although lanthanide containing alloys (e. g., SmCo and Nd Fe B) are
5
2
14
[
7b,c,10]
known
since long ago to produce strong bulk magnets, magnetism of lanthanides in
molecular level was invented only before 15 years when, in 2003, Ishikawa and coworkers
[
11]
III
reported slow magnetic relaxation of a mononuclear Tb compound, tetrabutylammonium
–
1 [11]
bis(phthalocyanine)terbiumate(III). The U of this system is 230 cm , significantly larger
eff
[
12‒14]
than any SMM of non-lanthanide metal ions reported so far.
Since then, main focus of
molecule based magnetic materials was shifted to lanthanide containing systems and, in the
[
7‒9, 15‒20]
last fifteen years, numerous mono/di/oligonuclear SMMs having only 4f metal ions
[21‒26]
or both 3d and 4f metal ions
as the paramagnetic centers, have been reported.
Apparently, a diamagnetic metal ion should have no role in the SMM behavior of a
II
II
heteronuclear system containing a lanthanide and a diamagnetic metal ion, such as Zn , Mg ,
II
I
III
[27c,f]
Ba , K , Co , etc.
However, interestingly, it has been observed and theoretically
III
III
explained that M‒Ln compounds (M = a diamagnetic metal ion), particularly M‒Dy
[27]
compounds are better SMMs than the analogous 3d-4f or 4f compounds.
Therefore,
combination of an anisotropic lanthanide and a diamagnetic metal ion has been proposed as
III
III
an approach to get more efficient SMMs. Therefore, isolation and studies of Co Dy
systems deserve attention, which is the major focus of this investigation.
Luminescence of lanthanides is generally characterized with narrow line-like
emissions, long excited-state lifetimes and high quantum yields and some of Ln(III) ions
[
28]
exhibit characteristic emission from visible to near-IR region.
As f-f transitions are
Laporte-forbidden, Ln(III) ions need to be sensitized indirectly via the intervention of an
organic ligand that efficiently transfers the energy to the Ln(III) ion, the so called “antenna
[
29]
effect”.
Luminescent lanthanide complexes are promising candidates for their potential
[
28a,30]
III
application in biological systems, optoelectronic devices and as biological sensors.
Dy
based luminescent complexes deserve special attention as they can behave as a SIM/SMM
III
too due to the intrinsically anisotropic nature of Dy center. Hence, combination of dual
III
properties, magnetic and optical on a single Dy based entity leads to the development of
multifunctional molecular material and this has become a burgeoning interest of researchers
in very recent times. Moreover, there are few numbers of luminescent magnets (i.e;
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
4
combining both magnetic and luminescent properties on a single molecule). Those systems
III
[31a‒e]
II
III
[31f]
III
III
are either purely Dy based
or Zn Dy based
while Co Dy based luminescent
[
31g,h]
magnets are significantly scarce.
Obviously, this area needs much exploration.
The family of Schiff base ligands which are obtained on [1+1] condensation of 3-
methoxysalicylaldehyde or 3-ethoxysalicylaldehyde and an aminoalcohol or aminophenol,
[26c, 32, 33]
can stabilize 3d-4f systems,
owing to the presence of two types of compartments (as
in Scheme 1); O(alcohol/phenol)N(imine)O(phenol) and O(ether)O(phenol), where the
former is more potential to occupy a 3d metal ion while the latter is more potential to bind a
III
III
[32i,33a]
lanthanide. Some Co Ln compounds derived from such ligands are also known.
However, condensation product (H L; Scheme 1) of 3-methoxysalicylaldehyde and 2-amino-
2
III
III
2
-methyl-1-propanol was not previously utilized to prepare a Co Ln compound. As slight
change in the ligand periphery may result in the generation of drastically different type of
structure as well, we reacted H L with cobalt(II) acetate tetrahydrate and hydrated Ln(NO )
2
3 3
III
III
(
Ln = Gd and Dy) and isolated two compounds of composition [Co Gd L (µ -O) (µ-
4 3 4 4 2
III
III
OMe) (µ -OAc) (H O) (NO ) ]·NO ·2CH OH·2H O (1) and [Co Dy L (µ -O) (µ-
2
1,3
4
2
2
3 2
3
3
2
4
3
4
4
2
OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2), which represent a new type of
2
1,3
4
2
2
3 2
3
3
2
structure indeed. Herein, we report the syntheses, crystal structures and dc and ac magnetic
properties of 1 and 2.
Scheme 1. Chemical structure of the ligand.
Results and discussion
Syntheses
H L solution. A solution of 2-amino-2-methyl-1-propanol (0.045 g, 0.5 mmol) in 5 mL
2
methanol was added dropwise to a 10 mL methanol solution of 3-methoxy salicylaldehyde
(0.075 g, 0.5 mmol) under warming condition. The reaction mixture was refluxed for 2 h.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
5
After cooling, the volume of the solution was diluted to 25 mL in a volumetric flask. The
resulting orange coloured ‘H L solution’ was considered to contain 0.5 mmol of the ligand
2
H L and was utilized for subsequent reactions without further purification.
2
Description of the crystal structures of 1 and 2.
III
III
Compounds [Co Gd L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·2CH OH·2H O
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
III
III
(
1) and [Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2)
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
crystallize in the same crystal system, monoclinic, and same space group, C2/c, with closely
similar values of unit cell parameters (Table 1), revealing that these two compounds are
isomorphous. One half of each structure is symmetry related to the another half due to the
presence of inversion center. The crystal structures of 1 and 2 are shown, respectively, in
III
III
Figure 1 and Figure 2. The structures show that both compounds are heptanuclear Co Ln
4
3
systems (Ln = Gd or Dy) having similar core structure; the compositions are different only in
terms of solvents of crystallization, 2CH OH and 2H O in 1 while 3CH OH and 1.5H O in 2.
3
2
3
2
A simplified and general schematic presentation of the heptametallic cores is shown in
Scheme 2.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
6
Table 1. Crystallographic data for 1 and 2.
1
2
Empirical formula
C H N O Co Gd
C H N O Co Dy
58 78 7 35 4 3
5
8
78
7
35
4
3
Formula weight
Crystal system
Space group
a[Å]
2140.74
Monoclinic
C2/c
17.505(6)
19.784(7)
24.106(8)
90
2156.49
Monoclinic
C2/c
17.531(6)
20.038(7)
24.325(8)
90
b[Å]
c[Å]
α
[°]
[°]
β
103.413(4)
90
102.979(4)
90
γ
[°]
3
V[Å ]
Z
8121(5)
4
8326(5)
4
T[K]
296(2)
296(2)
2
θ[°]
3.156−51.798
3.297
3.132−51.000
3.518
–
1
µ
[mm ]
–
3
ρcalcd[g cm ]
F(000)
1.751
1.720
4220
4244
Absorption-correction
Index ranges
Multi-scan
−20 ≤ h ≤ 21
Multi-scan
−21 ≤ h ≤ 21
−23 ≤ k ≤ 23
−29 ≤ l ≤ 29
28955
−
23 ≤ k ≤ 24
−
29 ≤ l ≤ 29
Reflections collected
29203
Independent reflections
7767 (0.0616)
7610 (0.0688)
(
R )
int
a
b
R ,wR [I > 2σ(I)]
0.0605, 0.1595
0.0910, 0.1723
0.0546, 0.1404
0.0892, 0.1544
1.003
1
2
b
a
2
R ,wR [for all F ]
1
2
o
2
Goodness-of-fit on F , S
1.090
a
b
2
2 2
4 1/2
R = [∑||F | – |F ||/∑|F |]. wR = [∑w(F – F ) /∑wF ]
1
o
c
o
2
o
c
o
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
7
III
III
Figure 1. Crystal structure of [Co Gd L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·
4
3
4
4
2
2
1,3
4
2
2
3 2
NO ·2CH OH·2H O (1). All the Hydrogen atoms and counter nitrate ion have been omitted
3
3
2
for clarity. Symmetry code: D = 1–x, y, 1.5–z.
III
III
Figure 2. Crystal structure of [Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·
4
3
4
4
2
2
1,3
4
2
2
3 2
NO ·3CH OH·1.5H O (2). All the Hydrogen atoms and counter nitrate ion have been omitted
3
3
2
for clarity. Symmetry code: D = 1–x, y, 1.5–z.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
8
III
III
Scheme 2. Simplified illustration of the heptanuclear [Co Ln L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
+
‒
OAc) (H O) (NO ) ] NO part in 1 and 2. All solvent of crystallization molecules have been
omitted for clarity. Ln = Gd for 1 and Ln = Dy for 2, respectively.
4
2
2
3 2
3
III
III
III
III
III
The heptametallic core in 1 and 2 consists of the following species: (i) Four Co ; (ii)
III
Three Ln ; (iii) Four deprotonated O(methoxy)O(phenoxo)N(imine)O(alkoxo) Schiff base
2
–
ligands, [L] , where both phenoxo and alkoxo moieties of H L are deprotonated; (iv) Two
2
µ -oxo ligands; (v) Two µ-methoxide ligands (vi) Four µ -acetate ligands; (vii) Two
4
1,3
chelating nitrate ligands; (viii) Two water ligands. Bond Valence Sum (BVS) calculations on
[
34a,b]
oxygen atoms
in 1 and 2 (Table S1) confirm their appropriate protonation level on
alkoxo, oxo, methoxide and water oxygen atoms.
It may be simpler to describe the structures by fragment analyses. The heptametallic
III
III
III
cores in 1 and 2 may be taken as consisted of two symmetry related trinuclear Co Ln Co
moieties (Co1Ln1Co2 and Co1DLn1DCo2D) and a central lanthanide center (Ln2) in such a
way that each of the three metal ions of each of the two trinuclear moieties are bridged with
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
9
the central lanthanide through bridging oxo/alkoxo/acetate ligands. Each symmetry related
2
–
trinuclear core contains two [L] . The O(phenoxo)N(imine)O(alkoxo) compartment of each
2
–
III
of the two [L] is occupied by a Co ion (Co1 in O1N1O2 site and Co2 in O4N2O5
2
–
compartment)
O(phenoxo)O(methoxy)O(methoxy)O(phenoxo) site (O1O3O6O4) is generated by their
second compartments (i. e. O(phenoxo)O(methoxy) compartments). The
O(phenoxo)O(methoxy)O(methoxy)O(phenoxo) site thus generated is occupied by a Ln ion
Ln1) to result in the formation of the trinuclear core, which contains a number of other
and
the
two
[L]
are
so
placed
that
a
III
(
bridging/chelating/monodentate ligands to stabilize it and those are as follows: (i) One µ-
methoxide ligand (containing O12) which bridges Co1 and Co2; (ii) A µ -oxo ligand
4
(containing O11) which bridges Co1, Ln1 and Co2 in the trinuclear core as well as the central
lanthanide ion, Ln2; (iii) A µ -acetate ligand which bridges Co1 and Ln1 (through O7 and
1,3
O8); (iv) A µ -acetate ligand which bridges Co2 of the trinuclear core and the central
1
,3
lanthanide ion, Ln2 (through O9 and O10); (iv) A bidentate nitrate ligand which chelates Ln1
through O13 and O14; (v) A water ligand (containing O16) which coordinates Ln1. It has
already been mentioned that the central lanthanide (Ln2) is bridged with Co2 by a µ -acetate
1
,3
ligand and all the three metal ions (Co1, Ln1 and Co2) of a trinuclear core by a µ -oxo ligand.
4
There are two more linkings between a trinuclear core and the central lanthanide and those
are the two alkoxo oxygen atoms (O2 and O5) which bridge the central lanthanide with Co1
and Co2, respectively, of a trinuclear core. Thus, the central lanthanide is bridged with each
of the two trinuclear cores by one µ -oxo, two µ-alkoxo and one µ -acetate ligands. The
4
1,3
III
III
III
disposition of two symmetry related trinuclear Co Ln Co core with the central lanthanide
ion leads to the generation of ‘sand clock’ like (Figure 3) topology in 1 and 2.
III
III
Figure 3. ‘Sand Clock’ like topology of the Co Ln core in (1) (Ln= Gd) and (2) (Ln=
4
3
III
III
Dy); color code: Navy blue for Ln , olive green for Co and red for oxygen.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
0
It is relevant to summarize the nature of bridging ligands between two metal ions in
different bridged pairs: (i) Between Co1 and Co2 ‒ one µ -oxo and one µ -methoxide; (ii)
4
4
Between Co1 and Ln1 ‒ one µ-phenoxo, one µ -oxo and one µ -acetate; (iii) Between Co2
4
1,3
and Ln1 ‒ one µ-phenoxo and one µ -oxo; (iv) Between Co1 and Ln2 ‒ one µ -oxo and one
4
4
µ-alkoxo; (iv) Between Co2 and Ln2 ‒ one µ -oxo, one µ-alkoxo and one µ -acetate; : (v)
4
1,3
III
Between Ln1 and Ln2 ‒ one µ -oxo. Both Co centers (Co1 and Co2) are hexacoordinated,
4
Ln1 center is 9-coordinated and Ln2 center is 8-coordinated. The bond distances involving
the metal ions and the metal‒O‒metal bridge angles in 1 and 2 are listed in Table 2, 3, 4 and
5
. The coordination environment of Co1 and Co2 are similarly and slightly distorted
III
octahedral. The six Co‒O/N bond distances for both Co centers lie within a narrow range of
ca. 1.86‒1.92 Å, whereas that involving µ -acetate/µ-methoxide moiety is the longest and
1
,3
that involving µ-alkoxo is the shortest. This also indicates +3 oxidation state of this metal ion
[34c,d]
and also supported by BVS calculations (Table S2).
The overall (for Co1 and Co2)
ranges of the cisoid and transoid angles in 1 and 2 are 82.3‒95.2º and 174.9‒178.0º. All these
indicate slight distortion from octahedral geometry, which is further supported by SHAPE
[
35]
analyses (Table S3).
Table 2. The values of bond lengths (Å) and angles (º) around the Co1 center in 1 and 2.
Coordination environment Bond Lengths
1
2
of Co1 and geometry
(Å)
Co1−N1
Co1−O1
Co1−O2
Co1−O7
Co1−O11
Co1−O12
1.899(8)
1.890(6)
1.861(5)
1.921(6)
1.900(5)
1.921(6)
Bond Angles (º)
175.1(3)
176.5(2)
176.9(3)
95.2(3)
1.885(7)
1.903(6)
1.864(5)
1.927(6)
1.901(6)
1.918(6)
N1−Co1−O11
O7−Co1−O12
O1−Co1−O2
O1−Co1−N1
O1−Co1−O7
O1−Co1−O11
O1−Co1−O12
O2−Co1−N1
O2−Co1−O7
O2−Co1−O11
O2−Co1−O12
O7−Co1−N1
175.5(3)
176.3(3)
177.4(3)
94.9(3)
89.9(3)
88.5(3)
92.3(2)
86.6(3)
88.1(3)
90.2(2)
89.7(2)
88.3(3)
90.0(3)
88.6(3)
92.2(2)
86.2(3)
87.3(3)
90.1(2)
90.4(2)
88.1(3)
Octahedron
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
1
O7−Co1−O11
O11−Co1−O12
O12−Co1−N1
94.9(3)
82.3(2)
94.4(3)
94.6(2)
82.5(2)
94.4(3)
Transoid angle range
1
1
75.1(3) −
76.9(3)
Cisoid angle range
2.3(2) −
5.2(3)
175.5(3) −
177.4(3)
8
82.5(2) −
94.9(3)
9
Table 3. The values of bond lengths (Å) and angles (º) around the Co2 center in 1 and 2.
Coordination environment
of Co2 and geometry
Bond Lengths
(Å)
1
2
Co2−N2
Co2−O4
Co2−O5
Co2−O9
Co2−O11
Co2−O12
1.896(7)
1.895(5)
1.863(5)
1.918(6)
1.904(5)
1.892(6)
Bond Angles (º)
174.9(3)
176.5(3)
178.0(3)
95.1(3)
1.889(7)
1.890(5)
1.857(5)
1.920(6)
1.914(5)
1.902(6)
O9−Co2−O12
N2−Co2−O11
O4−Co2−O5
O4−Co2−N2
O4−Co2−O9
O4−Co2−O11
O4−Co2−O12
O5−Co2−N2
O5−Co2−O9
O5−Co2−O11
O5−Co2−O12
O9−Co2−N2
O9−Co2−O11
O11−Co2−O12
O12−Co2−N2
175.0(3)
177.0(3)
177.4(3)
95.2(3)
85.6(3)
87.4(2)
92.4(3)
86.1(3)
92.2(2)
91.4(2)
89.7(2)
88.9(3)
92.8(2)
82.6(2)
95.8(3)
86.1(3)
87.9(2)
91.7(3)
86.1(3)
92.4(3)
90.9(2)
89.6(3)
89.7(3)
Octahedron
92.3(2)
83.0(2)
95.0(3)
Transoid angle range
1
1
74.9(3) −
78.0(3)
Cisoid angle range
3.0(2) −
5.1(3)
175.0(3) −
177.4(3)
8
82.6(2) –
95.8(3)
9
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
2
Table 4. The values of bond lengths (Å) and bond angles (º) around the Ln1 center in 1 and
2
.
Coordination environment of
Ln1 and geometry
Bond Lengths (Å)
1
2
Ln1−O1
Ln1−O3
Ln1−O4
Ln1−O6
Ln1−O8
Ln1−O11
Ln1−O13
Ln1−O14
Ln1−O16
2.331(7)
2.614(8)
2.318(6)
2.757(6)
2.347(7)
2.298(5)
2.516(8)
2.446(8)
2.400(7)
Bond Angles (º)
62.5(2)
2.309(7)
2.611(7)
2.303(5)
2.771(6)
2.320(7)
2.272(5)
2.498(7)
2.451(7)
2.352(7)
O1−Ln1−O3
O1−Ln1−O4
O1−Ln1−O6
O1−Ln1−O8
O1−Ln1−O11
O1−Ln1−O13
O1−Ln1−O14
O1−Ln1−O16
O3−Ln1−O4
O3−Ln1−O6
O3−Ln1−O8
O3−Ln1−O11
O3−Ln1−O13
O3−Ln1−O14
O3−Ln1−O16
O4−Ln1−O6
O4−Ln1−O8
O4−Ln1−O11
O4−Ln1−O13
O4−Ln1−O14
O4−Ln1−O16
O6−Ln1−O8
O6−Ln1−O11
O6−Ln1−O13
O6−Ln1−O14
O6−Ln1−O16
O8−Ln1−O11
O8−Ln1−O13
O8−Ln1−O14
O8−Ln1−O16
O11−Ln1−O13
O11−Ln1−O14
O11−Ln1−O16
O13−Ln1−O14
O13−Ln1−O16
O14−Ln1−O16
63.0(2)
75.0(2)
73.9(2)
89.8(2)
72.6(2)
87.4(2)
73.9(2)
70.8(2)
148.5(2)
137.8(2)
135.5(2)
110.2(2)
65.3(2)
69.76(19)
149.1(2)
140.2(2)
133.7(2)
107.7(2)
65.7(2)
72.4(3)
71.8(2)
130.2(2)
118.0(2)
77.8(3)
130.5(3)
59.34(19)
141.5(2)
69.67(18)
77.2(2)
120.2(2)
121.6(2)
138.0(2)
128.63(19)
65.9(2)
131.1(2)
116.1(2)
74.9(3)
129.4(2)
59.58(19)
143.1(2)
70.12(18)
76.5(2)
119.4(2)
119.9(2)
137.1(2)
128.89(19)
66.4(2)
Spherical tricapped
trigonal prism
71.6(2)
136.4(2)
81.3(2)
138.2(3)
97.8(3)
71.4(2)
137.0(2)
81.3(2)
137.5(3)
97.0(3)
72.2(3)
72.6(3)
109.9(2)
148.6(2)
75.9(2)
51.6(3)
72.0(2)
111.4(2)
150.0(2)
75.9(2)
51.5(2)
71.8(2)
74.0(3)
75.0(2)
Angle range
5
1.6(3) −
49.1(2)
51.5(2) −
150.0(2)
1
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
3
Table 5. The values of bond lengths (Å) and bond angles (º) around the Ln2 center in 1 and
2
.
Coordination
environment of Ln2 and
geometry
Bond Lengths (Å)
1
2
Ln2−O2
Ln2−O5
Ln2−O10
Ln2−O11
2.350(5)
2.300(5)
2.409(6)
2.344(5)
2.290(5)
2.404(5)
2.425(5)
2.431(5)
Bond Angles (º)
101.8(3)
O2−Ln2−O2D
O2−Ln2−O5
101.8(3)
79.5(2)
78.9(2)
O2−Ln2−O5D
O2−Ln2−O10
O2−Ln2−O10D
O2−Ln2−O11
O2−Ln2−O11D
O5−Ln2−O5D
O5−Ln2−O10
O5−Ln2−O10D
O5−Ln2−O11
O5−Ln2−O11D
O10−Ln2−O10D
O10−Ln2−O11
O10−Ln2−O11D
O11−Ln2−O11D
75.94(18)
138.96(18)
103.2(2)
67.67(18)
144.74(18)
139.6(3)
75.68(18)
139.61(18)
102.5(2)
67.95(18)
145.16(17)
140.1(3)
75.97(19)
140.49(18)
69.77(18)
125.91(19)
78.2(3)
73.63(19)
74.56(18)
138.6(3)
76.3(2)
140.77(19)
69.08(18)
126.58(18)
77.6(3)
73.10(19)
75.59(18)
139.5(3)
Triangular dodecahedron
Angle range
6
7.67(18) −
67.95(18) −
145.16(17)
1
44.74(18)
The Ln‒O bond distances in the 9-coordinated Ln1 environment lie in the ranges
III
III
2
.298‒2.757 Å in 1 (Gd analogue) and 2.272‒2.498 Å in 2 (Dy analogue), where a clear
order of the bond distances involving the following ligand moieties is observed:
methoxy>nitrate>water>µ -acetate>µ-phenoxo>µ-oxo. The overall ranges of O‒Ln1‒O
1
,3
bond angles in 1 and 2 are similar, ca. 51.5‒150.0º. The Ln‒O bond distances in the 8-
III
coordinated Ln2 environment lie in the similar ranges 2.300‒2.431 Å in 1 (Gd analogue)
III
and 2.290‒2.425 Å in 2 (Dy analogue), where a clear order of the bond distances involving
the following ligand moieties is observed: oxo>µ -acetate>µ-alkoxo. The overall ranges of
1
,3
O‒Ln2‒O bond angles in 1 and 2 are similar, 67.67‒144.74º in 1 and 67.95‒145.16º in 2.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
4
[35]
SHAPE analyses (Table S4 and S5) reveal that spherical tricapped trigonal prism
TCTPR-9) and triangular dodecahedron (TDD-8) are the ‘most ideal’ geometries of the Ln1
(
and Ln2 centers, respectively. The general illustrations of the TCTPR-9 and TDD-8
geometries of Ln1 and Ln2 centers, respectively, are shown in Figure 4.
Figure 4. General representation of the co-ordination environment showing a) Octahedral
Co1; b) Octahedral Co2; c) Spherical tricapped trigonal prism (D ) (TCTPR-9) geometry for
3
h
Ln1 center and d) Triangular dodecahedron (D ) (TDD-8) geometry of Ln2 center in 1 and
2
d
2
.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
5
Magnetic Properties
Variable temperature (2–300 K) dc magnetic susceptibilities of powdered samples of 1 and 2
were measured at an applied dc field of 0.1 T and has been portrayed in Figure 5 and 6,
respectively, in terms of χ T versus T plots. The χ T values at 300 K for 1 and 2 are,
M
M
3
–1
respectively, 23.38 and 39.90 cm K mol . The observed χ T values are close enough to
M
3
–1
III
those (23.63 and 42.51 cm K mol , respectively) expected for three noninteracting Gd ions
8
III
6
(
S , L = 0, S = 7/2, g = 2) for 1 and three noninteracting Dy ions ( H15/2, L = 5, S = 5/2, g
7/2
=
4/3) for 2. For 1, on lowering the temperature, the χ T product remains apparently constant
M
3
–1
down to 10 K followed by rapid drop to a minimum value of 20.90 cm K mol at 2 K. This
III
III
feature could be attributable to weak antiferromagnetic Gd ···Gd exchange interactions.
The dc magnetic susceptibility of complex 2 (Figure 6) features a very gradual decrease of
χMT product upon lowering of temperature and finally reaches the minimum value of 32.17
3
–1
III
cm K mol at 2 K. This is indicative of depopulation of the M sublevels of Dy ions or
J
III
III
weak Dy ···Dy antiferromagnetic exchange interactions or both.
III
III
Figure 5. Temperature dependence of the dc magnetic susceptibility for [Co Gd L (µ -
4
3
4
4
O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·2CH OH·2H O (1) in the temperature region
2
2
1,3
4
2
2
3 2
3
3
2
2
‒300 K. The pink colored squares are experimental data, while the solid line represents the
calculated curve.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
6
III
III
Figure 6. Temperature dependence of the dc magnetic susceptibility for [Co Dy L (µ -
4
3
4
4
O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) in the temperature region
2
2
1,3
4
2
2
3 2
3
3
2
2
‒300 K.
The magnetization (M) data up to 5 T and at 2, 4, 6, 8 and 10 K for 1 (Figure 7) and 2
Figure S1) have been collected. The magnetization curve of 1 displays a steady increase that
finally reaches the value of 21.54 Nµ at 2 K and 5 T. This value is close to the saturation
(
β
III
8
magnetisation of 21 Nµ for three noninteracting Gd ions ( S , g = 2), indicative of weak
β
7/2
nature of the possible exchange interactions. Moreover, M versus H/T plot (Figure S2) of 1
shows that all magnetisation isotherms merge on the same master curve establishing the
III
isotropic nature of Gd ions. The M versus H data (Figure S1) of 2 reveals that the
magnetisation value (15.33 Nµ ) even at 2 K and 5 T is significantly smaller than the
β
III
6
saturation value (30 Nµ ) for three noninteracting Dy ions ( H15/2, g = 4/3), revealing the
β
existence of appreciable anisotropy which is further strengthened by the non-superimposable
nature of the M versus H/T profiles with a master curve (Figure 8).
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
7
III
III
Figure 7. Magnetizations versus field plots for [Co Gd L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·2CH OH·2H O (1), collected at the indicated temperatures. The
4
2
2
3 2
3
3
2
different coloured circles are experimental data, while the solid lines represent the calculated
curves.
Figure 8. Magnetizations vs field/temperature plots of the experimental data for
III
III
[
Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O
(2),
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
collected at the indicated temperatures.
As Gd1 and Gd2 as well as Gd1D and Gd2 are bridged by diamagnetic atoms and
there is no bridging ligand between Gd1 and Gd1D, the magnetic properties of 1 can be
modelled by Hamiltonian, H = –2J(S ⋅S + S ⋅S ), where S = S = S = 7/2 (Figure 9). With
1
2
2
3
1
2
3
this model Hamiltonian, χ T versus T data and M versus H data were simulated
M
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
8
[
36]
simultaneously to a satisfactory level using PHI software with the following converging
−
1
parameters: J = –0.010 cm and gGd = 1.99. This clearly reflects the very weak
III
III
antiferromagnetic Gd ···Gd interaction present in 1.
III
III
Figure 9. Simple representation of the core structure of [Co Gd L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·2CH OH·2H O (1) showing a) only the metal atoms and b)
4
2
2
3 2
3
3
2
III
accounting the magnetic interaction present among Gd ions. Here, J and J both are equal.
1
2
Symmetry code: D = 1–x, y, 1.5–z.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
1
9
The presence of significant magnetic anisotropy may lead to SMM behavior in
complex 2. To explore this aspect, variable-temperature (2–20 K) ac magnetic susceptibility
(Figure S3 and 10) at different fixed frequency (1–951 Hz) as well as variable-frequency (1–
9
51 Hz) ac magnetic susceptibility (Figure S4 and 11) at different fixed temperatures (2–20
K) were measured on powdered sample under zero-dc field with 3.5 Oe oscillating field. Both
in-phase (χ ') and out-of-phase (χ '') ac magnetic susceptibility of both types (variable-
M
M
temperature and variable-frequency) data exhibit significant temperature and frequency
dependence under zero-dc field, indicating that cluster 2 exhibits slow relaxation of
magnetisation and is a zero-field SMM.
Figure 10. Temperature dependence of the out-of-phase ac molar susceptibility (χM'') for
III
III
[
Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) at the
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
indicated frequencies under zero-dc field. Solid lines are guide for eyes only.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
0
Figure 11. Frequency dependence of the out-of-phase ac molar susceptibility (χM'') for
III
III
[
Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) under
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
zero-dc field. Inset shows the magnifying view that displays the clearer presence of two peak
maxima in χM'' versus Frequency plot. The solid lines corresponding to each temperature are
visual guide only.
Temperature-dependent out-of-phase (χ '') signals (Figure 10) reveal the presence of
M
two clear peak maxima which are separated well; one at higher temperature region, ca. 14 K,
and the other at lower temperature region, ca. 7 K. At further lower temperature (below 5 K),
χM'' rises rapidly. This complex pattern suggests that multiple relaxation mechanism is
operative with quantum-tunneling mechanism prevailing in the low temperature region as one
[
37]
of the relaxation processes. From the frequency-dependent χ '' data (Figure 11), presence
M
of two peak maxima are more clearly evident; the presence of the higher frequency peak
maxima in the data in the lower temperature region, 5–10 K, and the lower frequency peak
maxima in the data in the higher temperature region, above 10 K, corroborates the operation
III
[27e,38]
of two separate relaxation channels, as observed in some other Dy based SMMs/SIMs.
Obviousloy, the higher frequency peak is the fast relaxation (FR) channel while lower
frequency peak is the slow relaxation (SR) channel.
The Cole-Cole plots (Figure 12) corresponding to the frequency-dependent in-phase
(
χ ') (Figure S4) and out-of-phase (χ '') (Figure 11) data were constructed in the temperature
M
M
region 5–11 K. It consists of distorted semicircles, indicative of the presence of two fused
semicircles. Thus, from Cole-Cole plots, the presence of two separate relaxation processes is
clarified (Figure 12 and S5); left and right semicircles represent the fast relaxation (FR) and
slow relaxation (SR) processes, respectively. It is well documented from Figure 12, S5 and
S6 that, with increasing temperature from 5 K, a slow collapsing of two relaxation processes
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
1
into a single one takes place. These data were nicely fitted by using the sum of two modified
[
39]
Debye functions, yielding two sets (α and α ) of relaxation distribution parameters (Table
1
2
S6); α lies in the range 0.00–0.22 and corresponds to the fast relaxation channel, α lies in the
1
2
range of 0.00–0.10 and corresponds to the slow relaxation channel. The values of α represent
2
an overall narrow distribution of relaxation time for slow relaxation process. On the other
hand, for the fast relaxation process, the values of α take a wider distribution in the low
1
temperature region (α = 0.17–0.22 in the temperature range 5–9 K) and switch over to
1
extremely narrower distribution (α ≈ 0 at T > 9 K) in the higher temperature region, above 9
1
K. The relaxation times (τ) at different temperatures corresponding to the temperature-
dependent out-of-phase data (in Figure 10) could be fitted well (Figure 13) with the
‒
1
‒7
Arrhenius equation (ln τ = ln τ0 + U /k T), giving U = 51.4 cm and τ0 = 9.40 × 10 s for
eff B eff
‒
1
‒7
the slow relaxation (SR) channel and Ueff = 24.3 cm and τ0 = 6.91 × 10 s for the fast
relaxation (FR) channel. Linear fitting above 9.2 K (9.2‒14.3 K) for the slow relaxation
channel and above 5.4 K (5.4‒6.5 K) for the fast relaxation channel suggests that both are
thermally activated Orbach processes. Appreciable deviation from linearity of the data below
9
.2 K and 5.4 K, respectively, reveals a transition from thermally activated to a quantum-
tunneling mechanism.
III
III
Figure 12. Cole-Cole plots for compound [Co Dy L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) measured at the indicated temperatures
4
2
2
3 2
3
3
2
under zero-dc field. The black solid lines at each temperature represents the best fits to the
experimental data obtained with the sum of two modified Debye functions.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
2
Figure 13. Plot of ln(
τ) versus reciprocal temperature (Arrhenius plot) in the temperature
III
III
region 5.30‒14.30 K under zero-dc field for [Co Dy L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2). The pink solid line corresponds to the linear
4
2
2
3 2
3
3
2
fits for both (Fast Relaxation and Slow Relaxation) the thermally activated regime.
To get further insight into the magnetic relaxation dynamics of complex 2, we have
collected the ac magnetic susceptibility data under an applied 2000 Oe dc field. Variable-
temperature (2–20 K) ac magnetic susceptibility (Figure S7 and 14) at different fixed
frequencies (1–951 Hz) and variable-frequency (1–951 Hz) ac magnetic susceptibility (Figure
S8 and 15) at different fixed temperatures (2–20 K) were measured on powdered sample
under 2000 Oe dc field with 3.5 Oe oscillating field. Temperature-dependent out-of-phase
signals (χ ''; Figure 14) display, compared to zero-dc field (Figure 10), further well resolved
M
fast and slow relaxation processes making the fast relaxation process more readily observable
with clear presence of fast relaxation peaks (under 2000 Oe dc field) compared to tailing
(under zero-dc field). Below 4 K, with increasing frequency, there are upturns in the out-of-
phase (χ '') signals (Figure 14). This suggests the onset of quantum-tunneling mechanism, a
M
third relaxation process that has been alleviated upon application of 2000 Oe dc field.
Frequency-dependent out-of-phase (χ '') signals (Figure 15) also exhibit more clear peaks
M
than zero-dc field and as the temperature is lowered the position of peak maxima changes
from higher frequency to lower frequency. From this frequency dependent out-of-phase (χ '')
M
profile, it is apparent that apart from the observed peaks in the investigated frequency range
(1–951 Hz), multiple peaks would be visible if we go beyond either 1 Hz or 951 Hz. But this
was the hardware limit.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
3
Figure 14. Temperature dependence of the out-of-phase ac molar susceptibility (χM'') for
III
III
[
Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) at the
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
indicated frequencies under 2000 Oe dc field. Solid lines are guide for eyes only.
Figure 15. Frequency dependence of the out-of-phase ac molar susceptibility (χM'') for
III
III
[
Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) under
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
2
000 Oe dc field. The solid lines corresponding to each temperature are visual guide only.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
4
Cole-Cole plots in the temperature region, 6‒9 K (Figure 16), were attempted to fit by using
[40]
generalized Debye function but it was unsatisfactory. Hence, we tried to fit the data by
[39]
employing the sum of two modified Debye functions.
It gives satisfactory fitting and
yields α value in the range of 0.20–0.26 while the value of α ≈ 0. Such a high value of
1
2
relaxation width (α = 0.20–0.26) for fast relaxation (FR) indicates wider distribution of
1
relaxation time and extremely low value of relaxation width (α ≈ 0) for slow relaxation (SR)
2
suggests extremely narrower distribution of relaxation time. We have quantified the value of
effective energy barrier (U ) and relaxation time (τ0) as well by taking the relaxation times
eff
(
τ) at different temperatures corresponding to the temperature-dependent out-of-phase ( M
χ '')
data in Figure 14. Linear fitting (Figure S9) in the temperature region 10.1‒12.8 K for the
‒
1
‒7
slow relaxation (SR) gives U = 51.5 cm and τ0 = 4.85 × 10 s and linear fitting in the
eff
‒
1
temperature region 5.0‒7.1 K for the fast relaxation (FR) gives U = 28.4 cm and τ0 = 8.62
eff
‒
7
×
10 s. Hence, operation of Orbach processes at thermally activated regime for both slow
and fast relaxation processes under zero-dc field is again corroborated by the applied field
2000 Oe) magnetic relaxation dynamics of compound 2.
(
III
III
Figure 16. Cole-Cole plots for compound [Co Dy L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) measured in the temperature region 6K‒9K
4
2
2
3 2
3
3
2
under 2000 Oe dc field. The black solid lines at each temperature represents the best fits to
the experimental data obtained with the sum of two modified Debye functions.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
5
[
41]
A MAGELLAN
analysis (Figure 17) of compound 2 reveals the disposition of
III
local anisotropy axes on the Dy centers (Dy1, Dy1D and Dy2) are noncolinear, that
possibly regulate the relaxation of magnetisation through multiple pathways. In weakly
III
exchange coupled Dy containing compounds, slow relaxation of magnetisation arises not
III
only due to intrinsic magnetic anisotropy of the Dy centers but also due to ligand field
[
16a,42]
effects associated with the coordination geometry.
Again, the presence of diamagnetic
III
Co ions in 2 can potentially influence the magnetic behavior that, in turn, increases the
III
III
[27c]
effective energy barrier as observed in previously reported Co Dy systems.
In case of
the 9-coordinated Dy1/Dy1D centers having DyO coordination environment, the oxide
9
oxygen atom O11/O11D is closest (2.298 Å) to Dy1/Dy1D while for the 8-coordinated Dy2
center with DyO coordination environment, alkoxo oxygen atom O5 is closest (2.272 Å) to
8
Dy2 among other ligand atoms. Now, the anisotropy axes on Dy1 and Dy1D are closely
parallel to Dy1‒O11 and Dy1D‒O11D bonds, respectively, while the anisotropy axis on Dy2
is not at all parallel to Dy2‒O5 bond but the anisotropy axis is closely parallel to Dy2‒O2D
bond, where O2D is the second closest (2.350 Å) atom to Dy2 among the ligating atoms.
Hence, it is difficult to assign the slow relaxation and fast relaxation processes corresponding
to the Dy centers.
III
Figure 17. The direction of the local anisotropy axes (yellow coloured dotted lines) of Dy
III
III
ions in [Co Dy L (µ -O) (µ-OMe) (µ -OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2)
4
3
4
4
2
2
1,3
4
2
2
3 2
3
3
2
showing only the metal atoms with their coordination environment.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
6
Absorption and emission spectra
Diffuse reflectance spectra (Figure S10) of 1 and 2 were recorded at room temperature. Both
compounds exhibit three bands at approximately the same positions; 422 nm (sharp), 559 nm
(broad) and 710 nm (broad) for 1 and 422 nm (sharp), 547 nm (broad) and 700 nm (broad)
for 2.
−
6
Absorption spectra of 5 × 10 (M) methanol solutions of 1 and 2 are shown in Figure
S10. Four absorptions occur in each case. One among them is a sharp peak while the other
–
1
–
three are broad shoulders. The band maxima (λmax in nm) and their corresponding ε (M cm
1
)
values are as follows: 240 and 111708 (sharp), 269 and 55955 (broad), 324 and 25078
broad) and 378 and 10011 (broad) for 1; 241 and 110358 (sharp), 269 and 54506 (broad),
(
3
29 and 23292 (broad) and 378 and 10854 (broad) for 2.
In order to check the emissive property of compound 2, we performed the steady state
fluorescence spectroscopy at room temperature for both solid and solution phases. In solid
state, upon excitation in the visible region at 422 nm, the compound exhibits three
III
characteristic bands of Dy in addition to a ligand based band (Figure 18). The emission
III
maxima of the bands at 506 nm, around 569 nm and 668 nm could be assigned to Dy based
4
6
4
6
4
6
F → H , F → H
and F → H
transitions, respectively. Besides, the emergence
9
/2
15/2
9/2
13/2
9/2
11/2
of a broad band at around 468 nm could be ascribed to ligand based emission. The split
III
4
6
pattern observed in the Dy based band centerd at around 569 nm ( F → H transition)
9
/2
13/2
into two components, one at 562 nm and the other at 576 nm, is attributable to the presence of
III
4
6
two crystallographically inequivalent Dy centers in 2 because the F → H transition in
9
/2
13/2
III
III
[43]
Dy based systems is highly sensitive to Dy geometry. Again, the emission band at 506
4
6
nm ( F → H transition) is significantly overlapped with the ligand based emission band.
9
/2
15/2
In solid state emission spectrum, presence of significant residual ligand based emission
2
‒
III
[16b]
reveals that photoinduced energy transfer from the ligand L to Dy is not complete.
The
2
‒
III
overall emissive nature in solid state for 2 indicates that L is acting as a sensitizer for Dy .
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
7
III
III
Figure 18. (a) Solid state fluorescence spectra of [Co Dy L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) at room temperature; λ = 422 nm. (b)
4
2
2
3 2
3
3
2
ex
−
6
III
III
Fluorescence spectra of 5 × 10 (M) MeOH solution of [Co Dy L (µ -O) (µ-OMe) (µ -
4
3
4
4
2
2
1,3
OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) at room temperature; λ = 378 nm.
4
2
2
3 2
3
3
2
ex
In contrary to the solid state, the emissive nature of 2 in solution phase is totally
diminished with subsequent presence of only ligand based emission band at 431 nm and that
−
6
is observable under excitation at 378 nm for 5 × 10 (M) MeOH solution. The nonemissive
nature of 2 in solution phase may be rationalised in terms of the interaction with OH
oscillators which favour thermal dissipation of energy that actually inhibits the emission.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
8
Comparison of the structures and magnetic properties of 1 and 2 with the related
systems.
III
III
A number of compounds containing both Co and Dy have been reported previously (Table
[44]
III
III [27h,43a]
III
III [44b‒d]
S7).
In terms of nuclearity, those systems include Co Dy ,
Co Dy ,
2
III
III [44e‒k]
III
III [42a]
III
III [44l]
III
III [44m]
3
III
III [44n‒p]
Co Dy ,
Co Dy ,
Co Dy ,
Co Dy
and Co Dy .
2 4
2
2
4
2
3
3
III
III
III
III
However, no compound of nuclearity Co Dy (or Co Ln , Ln = Ce‒Yb) was reported
4
3
4
3
previously. As already discussed, heptanuclear compounds 1 and 2 consist of two trinuclear
III
III
III
III
Co Ln Co fragments and a central Ln center. Interestingly, such an assembly of metal
III
III
ions in 3d-4f family is not known. Few of the known Co Dy systems were derived from
[
32i,33a]
[32i,33a]
ligands similar to H L.
In terms of nuclearity, this is only {Co Ln }.
It is worth
2
2
4
II
III
mentioning that there are few compounds having Co and Ln where the metal ions are
[
45]
II
III
[45b]
bridged by µ -oxo moiety.
Among them, there are only three Co Ln compounds
4
III
III
having two µ -oxo moiety but there is no Co Ln compound having µ -oxo moiety.
4
4
Notably, the heptamellic core can be compared with a ‘sand clock’ like topology (Figure 3),
II III
III
not reported previously in Co / Ln systems. Clearly, the title compounds represent some
new aspects from structural point of view.
[
46a‒g]
III
There are a number of homonuclear compounds
of Dy as well as heteronuclear
[
46h]
III
compounds
of Dy and a diamagnetic metal ion of different nuclearity that show double
relaxation channels. Some structural and SMM aspects of some of such systems and 2 are
compared in Table S8. The reason of two relaxations in most cases is the difference of the
III
surrounding of Dy centers in terms of ligand atoms, coordination geometry, ligand field,
etc. So, in 2, two relaxation channels at higher temperature (T > 5 K) region arise most
III
probably due to the presence of two distinct Dy centers, viz, nine coordinated Dy1/Dy1D
centers with TCTPR as the ‘most ideal’ geometry and eight coordinated Dy2 center with
TDD as the ‘most ideal’ geometry. As listed in Table S7 and S8, the double-relaxation
[
44f,g,k, 46a,c,e,h]
channels in most cases are not well resolved in zero-dc field
and also even under
[
27h, 42a, 44d]
an applied dc field.
Only in very few cases, the two processes are well resolved in
[
44h, 46b]
zero-dc field.
Clearly, compound 2 is among some few examples where the two
channels are well resolved in zero-dc field and the Ueff values of both channels could be
III
III
-1
determined. In the Co Dy SMMs, the range of U values is ca. 23‒88 cm and τ0 = 6.0 ×
eff
‒
8
‒7 [27h]
-1
‒7
1
0 ‒1.0 × 10 s.
Thus, 2 with U (SR) of 51.4 cm and τ0 = 9.40 × 10 s is among the
eff
III
III
good Co Dy SMMs.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900383
2
9
III
Although there are several examples in which Dy containing compounds show
[
31a‒e]
[44]
fluorescence behavior,
emission properties of the reported heteronuclear clusters
III
III
III
III
having Co and Dy were not reported. Clearly, 2 is the sole example of a Co Dy cluster
for which emission behavior is studied.
Conclusions
III
III
The
title
compounds
[Co Gd L (µ -O) (µ-OMe) (µ -
4 3 4 4 2 2 1,3
III
III
OAc) (H O) (NO ) ]·NO ·2CH OH·2H O (1) and [Co Dy L (µ -O) (µ-OMe) (µ -
4
2
2
3 2
3
3
2
4
3
4
4
2
2
1,3
III
III
OAc) (H O) (NO ) ]·NO ·3CH OH·1.5H O (2) are the sole examples of Co Ln clusters.
4
2
2
3 2
3
3
2
4
3
III
III
The compounds 1 and 2 have two µ -oxo moieties; there are no Co Ln cluster having at
4
III
III
[32i,33a]
least one µ -oxo bridge. Some Co Ln compounds were reported
previously from
ligands which are similar to H L but those compounds have nuclearity of the type, only
4
2
[
32i,33a]
{
Co Ln }.
Obviously, slight modification in the ligand periphery gives rise to a new
2
4
type of 3d-4f cluster in the form of 1 and 2.
III
The variable-temperature and variable-field magnetic data of the Gd analogue 1
could be simultaneously simulated; it exhibits very weak antiferromagnetic interactions with
−
1
III
J = –0.010 cm . The variable-temperature and variable-field dc data of the Dy analogue 2
reveal that it has significant anisotropy. Variable-temperature and variable-frequency ac data
reveal that 2 is a zero-field SMM with multiple relaxation dynamics in the investigated
temperature region, 2‒20 K. In the temperature region, T > 5 K, two relaxation channels (fast
III
and slow) are well-resolved. Notably, 2 is among only a few examples of Dy containing
SIMs/SMMs which exhibit two well-resolved relaxation channels at zero field. The presence
III
of two different types of Dy centers (eight coordinated ‘TDD’ and nine coordinated
‘TCTPR’) is the reason of the appearance of two channels.
2
‒
III
III
L
is a sensitizer for Dy because the characteristic Dy based transitions,
4
6
4
6
4
6
F → H , F → H and F → H appear in the fluorescence spectrum of 2 at solid
9
/2
15/2
9/2
13/2
9/2
11/2
state. Therefore, compound 2 is a bi-functional molecular material, combining SMM and
luminescent properties.
This article is protected by copyright. All rights reserved.