Syntheses, structures, and magnetic properties of acetato- and diphenolato-bridged 3d-4f binuclear complexes [M(3-MeOsaltn)(MeOH) x(ac)Ln(hfac)2](M = ZnII, CuII,NiII,CoII;Ln = LaIII,GdIII, TbIII, DyIII; 3-MeOsaltn = N,N-Bis(3-methoxy-2-oxybenzylidene)-1,3

Article

pubs.acs.org/IC

Syntheses, Structures, and Magnetic Properties of Acetato-

and Diphenolato-Bridged 3d−4f Binuclear Complexes

[M(3-MeOsaltn)(MeOH)_x(ac)Ln(hfac)₂] (M = Zn^II, Cu^II, Ni^II, Co^II; Ln = La^III,

Gd^III, Tb^III, Dy^III; 3‑MeOsaltn = N,N′‑Bis(3-methoxy-2-oxybenzylidene)-

1,3-propanediaminato; ac = Acetato;

hfac = Hexaﬂuoroacetylacetonato; x = 0 or 1)

Masaaki Towatari,^†Koshiro Nishi,^†Takeshi Fujinami,^†Naohide Matsumoto,*^,†Yukinari Sunatsuki,^‡

Masaaki Kojima,^‡Naotaka Mochida,^§Takayuki Ishida,^§Nazzareno Re,^∥and Jerzy Mrozinski^⊥

^†Department of Chemistry, Faculty of Science, Kumamoto University, Kumamoto 860-8555, Japan

^‡Department of Chemistry, Faculty of Science, Okayama University, Tsushima-naka 3-1-1, Okayama 700-8530, Japan

^§Department of Engineering Science, The University of Electro-Communications, Chofu, Tokyo 182-8585, Japan

^∥Facolta di Farmacia,

̀

Universita degli Studi “G.D ’Annunzio”, I-66100 Chieti, Italy

^⊥Faculty of Chemistry, University of Wroclaw, 14, F. Joliot-Curie, 50-383 Wroclaw, Poland

S

* Supporting Information

ABSTRACT: A series of 3d−4f binuclear complexes, [M(3-

MeOsaltn)(MeOH)_x(ac)Ln(hfac)₂] (x = 0 for M = Cu^II, Zn^II;

x = 1 for M = Co^II, Ni^II; Ln = Gd^III, Tb^III, Dy^III, La^III), have been

synthesized and characterized, where 3-MeOsaltn, ac, and hfac

denote N,N′-bis(3-methoxy-2-oxybenzylidene)-1,3-propane-

diaminato, acetato, and hexaﬂuoroacetylacetonato, respectively.

The X-ray analyses demonstrated that all the complexes have

an acetato- and diphenolato-bridged M^II−Ln^IIIbinuclear structure.

The Cu^II−Ln^IIIand Zn^II−Ln^IIIcomplexes are crystallized in an

isomorphous triclinic space group P1, where the Cu^IIor Zn^II

̅

ion has square pyramidal coordination geometry with N₂O₂

donor atoms of 3-MeOsaltn at the equatorial coordination sites

and one oxygen atom of the bridging acetato ion at the axial site. The Co^II−Ln^IIIand Ni^II−Ln^IIIcomplexes are crystallized in an

isomorphous monoclinic space group P2₁/c, where the Co^IIor Ni^IIion at the high-spin state has an octahedral coordination

environment with N₂O₂donor atoms of 3-MeOsaltn at the equatorial sites, and one oxygen atom of the bridged acetato and a

methanol oxygen atom at the two axial sites. Each Ln^IIIion for all the complexes is coordinated by four oxygen atoms of two

phenolato and two methoxy oxygen atoms of “ligand-complex” M(3-MeOsaltn), four oxygen atoms of two hfac⁻, and one oxygen

atom of the bridging acetato ion; thus, the coordination number is nine. The temperature dependent magnetic susceptibilities

from 1.9 to 300 K and the ﬁeld-dependent magnetization up to 5 T at 1.9 K were measured. Due to the important orbital

contributions of the Ln^III(Tb^III, Dy^III) and to a lesser extent the M^II(Ni^II, Co^II) components, the magnetic interaction between

M^IIand Ln^IIIions were investigated by an empirical approach based on a comparison of the magnetic properties of the M^II−Ln^III,

Zn^II−Ln^III, and M^II−La^IIIcomplexes. The diﬀerences of χ_MT and M(H) values for the M^II−Ln^III, Zn^II−Ln^IIIand those

for the M^II−La^IIIcomplexes, that is, Δ(T) = (χ_MT)_MLn− (χ_MT)_ZnLn− (χ_MT)_MLa= J_MLn(T) and Δ(H) = M_MLn(H) − M_ZnLn(H) −

M

_MLa(H) = J_MLn(H), give the information of 3d−4f magnetic interaction. The magnetic interactions are ferromagnetic if M^II=

(Cu^II, Ni^II, and Co^II) and Ln = (Gd^III, Tb^III, and Dy^III). The magnitudes of the ferromagnetic interaction, J_MLn(T) and J_MLn(H), are

in the order Cu^II−Gd^III> Cu^II−Dy^III> Cu^II−Tb^III, while those are in the order of M^II−Gd^III≈ M^II−Tb^III> M^II−Dy^IIIfor M^II= Ni^II

and Co^II. Alternating current (ac) susceptibility measurements demonstrated that the Ni^II−Tb^IIIand Co^II−Tb^IIIcomplexes showed

out-of-phase signal with frequency-dependence and the Ni^II−Dy^IIIand Co^II−Dy^IIIcomplexes showed small frequency-dependence.

The energy barrier for the spin ﬂipping was estimated from the Arrhenius plot to be 14.9(6) and 17.0(4) K for the Ni^II−Tb^IIIand

Co^II−Tb^IIIcomplexes, respectively, under a dc bias ﬁeld of 1000 Oe.

Received: March 10, 2013

A

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

isotropic magnetic center and Co^IIion with highly anisotropic

magnetic center are used. Many other d−f complexes of diﬀerent

nuclearities have been reported.¹⁵All these studies demonstrate

that the d−f cluster approach is a promising pathway to SMMs,

and that SMM behavior can be driven from many combinations

of d- and f-metal ions.

INTRODUCTION

■

The discovery of single molecule magnets (SMMs) in Mn₁₂

cluster and the subsequent extensive developments are con-

sidered as one of the most important achievements of molecular

magnetism.¹The origin of the SMM behavior is the easy axis

magnetic anisotropy (D < 0), which causes the formation of an

energy barrier that prevents reversal of the molecular magnet-

ization and causes a slow relaxation of the magnetization. SMMs

can be formed as a result of the combination of a large-spin

multiplicity of the ground state and an easy-axis (or Ising-type)

magnetic anisotropy of the entire molecule. Although the con-

ditions to be SMMs are not easily achieved by d-transition metal

complexes, utilization of 4f ions, such as Tb^IIIand Dy^III, satisﬁes

these conditions much more easily, because they have large

angular momentum in the ground multiplet state and a large

Ising-type magnetic anisotropy. Ishikawa et al. reported that

Tb^III, Dy^III, and Ho^IIIphthalocyanines exhibit SMM behavior,

and a molecule with only one Tb^IIIion can indeed behave as

SMM.²The d−f polynuclear structure gives a more suitable

molecular design for SMMs, because the high-spin state is often

generated by d−f magnetic interaction and the magnetic aniso-

tropy can be derived from both f- and d-elements. Therefore, in

recent years, molecular designs of SMMs and single chain

magnets (SCMs) containing f-block elements have attracted

much attention, although it should be noted that d−f complexes

have been studied since the pioneering works of Gatteschi et al.

on Cu^II−Gd^IIIcomplexes.³Matsumoto et al. reported the ﬁrst

3d−4f cluster exhibiting frequency dependent out-of-phase ac

signals, [Cu^IILTb^III(hfac)₂]₂(H₃L = 1-(2-hydroxybenzamido)-2-

(2-hydroxy-3-methoxy-benzylideneamino)-ethylene, and Hhfac =

hexaﬂuoroacetylacetone).⁴Christou’s group⁵and Pecoraro’s

group⁶synthesized Mn^III−Dy^IIIclusters and reported their

magnetic properties. [Mn^III₁₁Dy^III₄] and [Mn^III₂Dy^III₂] clusters

reported by Christou’s group⁵are conﬁrmed to exhibit temper-

ature and sweep rate dependent hysteresis loops, a trademark of all

SMMs. Furthermore, many reports on d−f clusters, [Cu^IITb^III]

binuclear, and [Cu^IITb^III₂] tetranuclear complexes,⁷[Cu^IITb^III]

In this study, in order to look for the possible combinations

of d- and f-metal ions giving SMM behavior, according to the

reaction procedure shown in Scheme 1, we synthesized the series

of acetato- and diphenolato-bridged 3d−4f binuclear complexes,

[M(3-MeOsaltn)(MeOH)_x(ac)Ln(hfac)₂] (x = 0 for M = Cu^II,

Zn^II; x = 1 for M = Co^II, Ni^II; Ln = Gd^III, Tb^III, Dy^III, La^III), where

3-MeOsaltn, ac, and hfac denote N,N′-bis(3-methoxy-2-oxy-

benzylidene)-1,3-propanediaminato), acetate, and hexaﬂuoro-

acetylacetonato, respectively.¹⁶Due to the suitable properties of

tetradentate 3-MeOsaltn ligand, the Ni^IIand Co^IIions assume

high-spin states which are best suited to give SMMs when

assembled with Ln^IIIspecies. A series of d−f binuclear complexes

[M(3-MeOsaltn)(MeOH)_x(ac)Ln(hfac)₂] (M = Cu^II, Zn^II, Co^II,

Ni^II; Ln = Gd^III, Tb^III, Dy^III, La^III) including diamagnetic Zn^IIand

La^IIIions were obtained, and all the structures were determined.

Fortunately, the structures of the Cu^II−Ln^IIIand Zn^II−Ln^IIIseries

have an isomorphous structure as well as those of the Ni^II−Ln^III

and Co^II−Ln^IIIseries. The comparison of the magnetic prop-

erties between the d−f complex with both magnetic ions and the

corresponding reference complex with one diamagnetic species

makes it possible to evaluate the nature of the M^II−Ln^IIImagnetic

interactions. Alternating (ac) magnetic measurements under

zero and 1000 Oe bias ﬁelds were carried out to investigate the

SMM behaviors.

RESULTS AND DISCUSSION

■

Synthesis of M^II−Ln^IIIComplexes [M^II(3-MeOsaltn)-

(MeOH)_x(ac)Ln^III(hfac)₂]. The present series of 3d−4f binuclear

complexes was synthesized by mixing [M(3-MeOsaltn)(H₂O)_x]

(M = Co^II, Ni^II, Cu^II, and Zn^II; 3-MeOsaltn = N,N′-bis(3-methoxy-

2-oxybenzylidene)-1,3-propanediaminato) and [Ln^III(ac)(Hhfac)-

(hfac)₂(H₂O)₂] (Ln = Gd^III, Tb^III, Dy^IIIand La^III; hfac =

hexaﬂuoroacetylacetonato, ac = acetato) as the 3d- and 4f-

component complexes, respectively. This type of 3d−4f complex

with the Schiﬀ-base ligands has been studied by Gatteschi et al.³

and Kahn et al.,¹⁷in which Kahn et al. used [Cu(salen)] and

[Gd(hfac)₃(H₂O)₂] for the d- and f-component complexes,

respectively. Fukuhara et al. reported that [Cu(saltn)] (saltn =

N,N′-bis(salicylidene)-1,3-diaminopropane) behaves as useful

“ligand-complex” binding other d-metal ions to form tetranuclear

and trinuclear complexes such as [Cu(saltn)(μ-ac)Cu(μ-

CH₃O)₂Cu(μ-ac)(saltn)Cu] and [Zn{(μ-ac)(saltn)Cu}₂].¹⁸Gatte-

schi et al. used the ligand-complex [Cu(saltn)] to prepare the Cu^II−

Gd^IIIcomplex [Cu^II(saltn)]₂Gd^III(H₂O)(NO₃)₃·2C₂H₅NO₂.^3b

2

4

pentanuclear complex,⁸[Cu^IIDy^III₃] nonanuclear complexes,⁹

6

[Fe^IIIDy^III] binuclear complex,¹⁰and [Ni^IILn^III] binuclear

complex containing paramagnetic Ni^IIion,¹¹and [Dy^III₆Mn^III

]

12

cores¹²have appeared in the latest literature. Kajiwara et al.

reported two Cu^II−Tb^IIIbinuclear complexes, in which the

symmetry of the ligand ﬁeld around the lanthanide ion strongly

aﬀects the magnetic anisotropy to induce a drastic switching from

easy axis to easy-plane anisotropy, and one complex showed

SMM behavior while the other one did not.¹³An approach

within the framework of d−f clusters has been extended in the

recent years. Chandrasekhar et al.^14areported a linear trinuclear

Co^II−Gd^III−Co^IISMM, and Costes et al.^14b,creported trinuclear

and tetranuclear Co^II−Gd^IIISMMs, where Gd^IIIion with an

Scheme 1. Synthetic Scheme of Acetato- and Diphenolato-Bridged Binuclear 3d−4f Complex,

[M(3-MeOsaltn)(MeOH)_x(ac)Ln(hfac)₂] (x = 0 for M = Cu^II, Zn^II; x = 1 for M = Co^II, Ni^II; Ln = Gd^III, Tb^III, Dy^IIIand La^III),

by the Reaction of [M(3-MeOsaltn)(H₂O)_x] and [Ln(ac)(Hhfac)(hfac)₂(H₂O)₂]

B

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Matsumoto et al. used the modiﬁed compound [Cu(3-

MeOsaltn)] for the synthesis of a series of Cu^II−Ln^III

complexes.¹⁶He demonstrated that the d-component complexes

[M(3-MeOsaltn)(H₂O)_x] have suitable properties for the

systematic synthesis of d−f complexes with a variety of M ions

of Co^II, Ni^II, Cu^II, and Zn^IIas well as Fe^IIIand Mn^III: (1) The

d-complex like [M(3-MeOsaltn)(H₂O)_x], that contains pheno-

lato and the neighboring coordinating groups such as methoxy,

hydroxy, and carboxy oxygen atoms,¹⁹can give the “ligand-

complex”. (2) The d-complex [M(3-MeOsaltn)(H₂O)_x] can

give stable high-spin species with M = Co^IIand Ni^IIions, while on

the contrary [Ni(salen)] gives diamagnetic low-spin species. (3)

The moderate solubility in common organic solvent is helpful

for the reaction with f-component.¹⁶Recently, by the use of

[Ni(3MeOsaltn)(H₂O)_x], Wang et al. and Pasatoiu et al. re-

ported the syntheses and magnetic properties of the Ni^II−Ln^III

complexes [Ni(3-MeOsaltn)Ln(NO₃)₂], which were prepared

by the reaction of [Ni(3-MeOsaltn)(H₂O)_x] and Ln(NO₃)₃·

4H₂O.²⁰In this study, we used the f-component complexes,

[Ln^III(ac)(Hhfac)(hfac)₂(H₂O)₂] (Ln = Gd^III, Tb^III, Dy^IIIand

La^III; hfac = hexaﬂuoroacetylacetonato), which were prepared

by the reaction of lanthanide(III) acetate tetrahydrate and

hexaﬂuoroacetylacetone with 1:3 molar ratio in water, while

the f-component complex [Ln^III(hfac)₃(H₂O)₂] has been used

for a number of d−f complexes.³Due to the stronger aﬃnity of

acetato ion to Ln^IIIions, the f-component compounds with the

formula [[Ln^III(ac)(Hhfac)(hfac)₂(H₂O)₂] were obtained. By

mixing d- and f-component complexes with 1:1 molar ratio, the

3d−4f complexes were obtained as well grown crystals. The

Cu^II−Ln^IIIand Zn^II−Ln^IIIcomplexes were obtained as green and

yellow platelike crystals, respectively. The crystals of Cu^II−Ln^III

and Zn^II−Ln^IIIcomplexes do not contain the solvent molecules.

The Co^II−Ln^IIIand Ni^II−Ln^IIIcomplexes were prepared in a

mixed solution of acetone and methanol in 1/1 volume and

obtained as orange and blue plate crystals, respectively. The

crystals of Co^II−Ln^IIIand Ni^II−Ln^IIIcomplexes gradually lose the

crystal solvents and decompose. All the samples were dried

in vacuo and subjected to the elemental analyses and the physical

measurements. The elemental analyses agreed with the chemical

formula of [M(3-MeOsaltn)(MeOH)_x(ac)Ln(hfac)₂] (x = 0 for

M^II= Cu^II, Zn^II; x = 1 for M^II= Co^II, Ni^II). It should be noted that

all complexes involve one acetato ion per binuclear molecular

unit that bridges M^IIand Ln^IIIions as later conﬁrmed by the X-ray

analyses. The acetato-bridge between d- and f-ions must con-

tribute to the stabilization of the 3d−4f binuclear complexes, and

as a result a series of the d−f complexes with various d-transition

metal ions is easily and successfully synthesized.

coordination bond distances with their estimated standard

deviations in parentheses are listed in Table 2, where the same

atom numbering is taken for the six complexes. As the six Cu^II−

Ln^IIIand Zn^II−Ln^IIIcomplexes are isomorphous; only the struc-

ture of the Cu^II−Gd^IIIcomplex is described in detail. Figure 1

shows the molecular structure of binuclear Cu^II−Gd^IIIcomplex

with the selected atom numbering scheme. The complex has

acetato- and diphenolato-bridged Cu^II−Gd^IIIbinuclear structure

with the distances of Cu···Gd = 3.4373(6) Å, Cu−O(6) =

2.231(3) Å, and Gd−O(5) = 2.327(3) Å, where O(5) and O(6)

atoms of the acetate ion bridge Cu^IIand Gd^IIIions in a μ-acetato

fashion. The Cu^IIion assumes a square pyramidal coordina-

tion geometry, where the equatorial coordination sites consist

of N₂O₂donor atoms of the tetradentate Schiﬀ-base ligand (3-

MeOsaltn) with the distances of Cu−O(2) = 1.969(4) Å,

Cu−O(3) = 1.974(3) Å, Cu−N(1) = 1.980(4) Å, and Cu−N(2)

= 1.982(5) Å, and the axial position is occupied by one of two

oxygen atoms of an acetato ion with the distance of Cu−O(6) =

2.231(3) Å. The Cu^IIion is deviated by 0.231 Å from the

equatorial N₂O₂plane toward the oxygen atom of acetato ion.

The coordination number of Gd^IIIion is nine, consisting of four

oxygen atoms of two hfac⁻, two phenolato and two methoxy

oxygen atoms of 3-MeOsaltn, and one oxygen atom of acetato

ion. The Gd−O distances of two hfac ligands are Gd−O(7) =

2.443(4) Å, Gd−O(8) = 2.387(4) Å, Gd−O(9) = 2.375(4) Å,

and Gd−O(10) = 2.527(3) Å. The two phenolato and two

methoxy oxygen atoms of the tetradentate ligand of Cu^IIcomplex

coordinate to a Gd^IIIion as the bridging atoms, with the distances

of Gd−O(1) = 2.521(5) Å, Gd−O(2) = 2.388(4) Å, Gd−O(3) =

2.342(4) Å, and Gd−O(4) = 2.599(4) Å. Although the Gd−O

distances from the phenolato oxygen atoms are much shorter

than those from the methoxy oxygens, the bridging mode from

the 3-methoxy oxygens helps to form the binuclear structure,

together with the acetato-bridge. The planarity of the bridging

CuO₂Gd moiety is estimated by the dihedral angle between

CuO₂and GdO₂planes, in which the angle is 24.4°. The pack-

ing of binuclear molecules in the crystal lattice was examined

(see Figure S1). The distances between the adjacent binuclear

molecules are Gd···Gd = 10.1500(4) Å, Cu···Cu = 5.9522(8) Å,

and Gd···Cu = 7.5770(6) Å, demonstrating that the neighboring

binuclear species are well separated and thus the intermolecular

magnetic interactions should be negligible.

Molecular Structures of Co^II−Ln^IIIand Ni^II−Ln^III(Ln^III=

Gd^III, Tb^III, and Dy^III). Due to the crystal habit of the rapid de-

composition and the disorder at hfac ligands, the X-ray diﬀrac-

tion analyses were determined at 100 K. The crystallographic

data of Co^II−Ln^IIIand Ni^II−Ln^III(Ln^III= Gd^III, Tb^III, and Dy^III)

complexes at 100 K are listed in Table 3. These six Co^II−Ln^III

and Ni^II−Ln^IIIcomplexes crystallized in same monoclinic space

group P2₁/c (No. 14) with similar cell parameters, and they are

isomorphous to each other. The coordination bond distances

with their estimated standard deviations in parentheses are listed

in Table 4, where the same atom numbering is taken for six com-

plexes. As the six Co^II−Ln^IIIand Ni^II−Ln^IIIcomplexes are iso-

morphous, only the structure of Co^II−Gd^IIIis described in detail.

Figure 2 shows the molecular structure of the binuclear

Co^II−Gd^IIIcomplex with the selected atom numbering scheme.

The complex has an acetato- and diphenolato-bridged heterometal

binuclear structure with the distances of Co···Gd = 3.4059(5) Å,

Co−O(6) = 2.082(3) Å, and Gd−O(5) = 2.349(3) Å, where the

acetato ion bridges Co^IIand Gd^IIIions and two oxygen atoms

O(5) and O(6) atoms of the acetato coordinate axially to Co^II

and Gd^IIIions. The Co^IIion assumes an octahedral coordination

Molecular Structures of Cu^II−Ln^IIIand Zn^II−Ln^III(Ln^III=

Gd^III, Tb^III, and Dy^III). The crystal structures were determined by

single-crystal X-ray diﬀraction analyses at 150 K. Due to the

crystal habit of the rapid decomposition and the disorder at hfac

ligands, the X-ray diﬀraction analyses at 296 K have problems in

the accuracy and the explanation of the disorder at hfac ligand.

Though the basic structural information, such as an isomorphous

structure for a series of Cu^II−Ln^IIIand Zn^II−Ln^IIIcomplexes, an

acetate-bridged binuclear structure, and the coordination

numbers of d- and f-elements, is undoubtedly obtained even at

296 K, the X-ray analyses at 150 K give the more reliable data.

The crystallographic data of the Cu^II−Ln^IIIand Zn^II−Ln^III

complexes (Ln^III= Gd^III, Tb^III, and Dy^III) at 150 K are listed in

Table 1. All six Cu^II−Ln^IIIand Zn^II−Ln^IIIcomplexes crystallized

in the triclinic space group P1 (No. 2) with similar cell param-

̅

eters, and therefore, they are isomorphous to each other. The

C

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Table 1. Crystallographic Data for the Cu^II−Ln^IIIand Zn^II−Ln^IIIComplexes [M(3-MeOsaltn)(ac)Ln(hfac)₂] (Ln^III= Gd^III, Tb^III,

and Dy^III)

Cu^II−Gd^III

Cu^II−Tb^III

Cu^II−Dy^III

formula

fw

C₃₁H₂₅N₂O₁₀F₁₂CuGd

1034.32

C₃₁H₂₅N₂O₁₀F₁₂CuTb

1036.00

C₃₁H₂₅N₂O₁₀F₁₂CuDy

1039.57

cryst syst

space group

a/Å

triclinic

P1 (No. 2)

̅

P1 (No. 2)

̅

P1 (No. 2)

̅

10.2246(3)

11.4639(3)

17.5688(5)

86.3787(9)

75.3343(8)

66.2417(7)

1821.59(8)

150

10.2184(4)

11.4642(4)

17.5832(4)

86.3967(8)

75.2519(9)

66.2258(10)

1821.06(10)

150

10.2311(3)

11.4571(4)

17.5663(4)

86.4514(11)

75.1403(7)

66.1499(11)

1818.23(9)

150

b/Å

c/Å

α/deg

β/deg

γ/deg

V/Å³

T/K

Z

2

D

_calcd/g cm⁻³

1.886

1.889

1.899

μ/cm⁻¹

25.12

26.24

27.46

R1 [I > 2σ(I)]

wR2 [all data]

reﬂns collected

unique reﬂns, R(int)

0.0407

0.0385

0.0417

0.1126

0.1057

0.1219

17 782

17 971

17 857

8242, 0.0299

8253, 0.0197

8170, 0.0270

Zn^II−Gd^III

Zn^II−Tb^III

Zn^II−Dy^III

formula

fw

C₃₁H₂₅N₂O₁₀F₁₂ZnGd

1036.16

C₃₁H₂₅N₂O₁₀F₁₂ZnTb

1037.83

C₃₁H₂₅N₂O₁₀F₁₂ZnDy

1041.41

cryst syst

space group

a/Å

triclinic

P1 (No. 2)

̅

P1 (No. 2)

̅

P1 (No. 2)

̅

10.1860(5)

11.5306(5)

17.5576(8)

86.1213(13)

75.3054(13)

66.2677(11)

1824.63(14)

150

10.1697(7)

11.5450(9)

17.5614(14)

86.204(3)

75.275(3)

66.167(2)

1822.6(3)

150

10.1959(3)

11.5344(3)

17.5376(5)

86.2477(8)

75.1651(8)

66.0818(7)

1820.81(8)

150

b/Å

c/Å

α/deg

β/deg

γ/deg

V/Å³

T/K

Z

2

D

_calcd/g cm⁻³

1.886

1.891

1.899

μ/cm⁻¹

25.82

26.97

27.47

R1 [I > 2σ(I)]

wR2 [all data]

reﬂns collected

unique reﬂns, R(int)

0.0387

0.0532

0.0370

0.1203

0.1474

0.1080

18 003

17 277

17 892

8283, 0.0362

8037, 0.0354

8228, 0.0245

geometry, where the equatorial coordination sites consist of N₂O₂

donor atoms of the tetradentate Schiﬀ-base ligand 3-MeOsaltn

with the distances of Co−O(2) = 2.058(3) Å, Co−O(3) =

2.056(2) Å, Co−N(1) = 2.063(3) Å, and Co−N(2) = 2.070(4) Å,

and the two axial positions are occupied by one of two oxygen

atoms of an acetato ion with the distance of Co−O(6) =

2.082(3) Å and an oxygen atom of methanol with 2.233(3) Å.

The Co−N and Co−O bond distances are consistent with a high-

estimated by the dihedral angle between CoO₂and GdO₂planes,

in which the angle is 14.2°. The dihedral angle (14.2°) of the

Co^II−Gd^IIIcomplex is smaller than that (24.4°) of the Cu^II−Gd^III

complex, in which the Co^IIand Cu^IIions have a six-coordinated

octahedral and a ﬁve-coordinated square pyramidal coordination

geometries, respectively. The packing manner of binuclear mole-

cules in the crystal was examined (see Figure S2). The distances be-

tween the adjacent binuclear molecules are Gd···Gd = 9.8838(3) Å,

Co···Co = 7.0791(6) Å, and Gd···Co = 7.4170(5) Å, demonstrating

that the magnetic ions are well separated.

General Procedures for Magnetic Measurements of

M^II−Ln^IIIComplexes. As the microcrystalline samples consist-

ing of Tb^IIIand Dy^IIIions showed reorientation in the applied

magnetic ﬁeld, all samples were dispersed in paraﬃn grease to

avoid this problem, see Experimental Section. The temperature

dependent magnetic susceptibility measurements on powdered

samples of M^II−Ln^IIIwere carried out in an applied magnetic

ﬁeld of 0.1 T in the temperature range 1.9−300 K. The ﬁeld de-

pendence of the magnetization up to 5 T was measured at 1.9 K.

3

spin state of the Co^IIion (S = /₂). On the basis of a similar

examination on the coordination bond distances, also the Ni^IIion

of the Ni^II−Ln^IIIcomplexes assumes high-spin state (S = 1). The

coordination number of Gd^IIIion is nine, consisting of four

oxygen atoms of two hfac⁻, two phenolato and two methoxy

oxygen atoms of 3-MeOsaltn, and one oxygen atom of acetato

ion. The two methoxy and two phenolato oxygen atoms of the

tetradentate ligand of Co^IIcomplex coordinate to a Gd^IIIion as

the bridging atoms, with the distances of Gd−O(1) = 2.706(3) Å,

Gd−O(2) = 2.337(3) Å, Gd−O(3) = 2.331(3) Å, and Gd−O(4) =

2.647(3) Å. The planarity of the bridging CoO₂Gd moiety is

D

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

3

and χ_4f= (Ng_J²β²/3kT)[J(J + 1)], with g_J= /₂+ [S(S + 1)

Table 2. Coordination Bond Distances (Å) and Ln^III···M^II

Distances (Å) for the Cu^II−Ln^IIIand Zn^II−Ln^IIIComplexes

[M(3-MeOsaltn)(ac)Ln(hfac)₂] (Ln = Gd^III, Tb^III, and Dy^III)

− L(L + 1)]/2J(J + 1).²¹

Magnetic Properties of Zn^II−Ln^IIIComplexes (Ln^III

=

Gd^III, Tb^III, and Dy^III). The temperature dependences of the

magnetic susceptibilities for the binuclear Zn^II−Ln^IIIcomplexes

are shown in Figure 3a, as plots of χ_MT versus T, where χ_Mis the

molar magnetic susceptibility per binuclear Zn^II−Ln^IIImolecule

and T is the absolute temperature. Since Zn^IIion with d¹⁰elec-

tronic conﬁguration is diamagnetic, the magnetic properties of

the binuclear Zn^II−Ln^IIIcomplexes depends only on the Ln^IIIion

and are helpful to evaluate the magnetic details for the present

series of the M^II−Ln^IIIcomplexes involving paramagnetic M^II

ions. The constant χ_MT value of ca. 8.1 cm³K mol⁻¹over 1.9−

300 K of the Zn^II−Gd^IIIcomplex is reproduced by the theoretical

value expected for one diamagnetic Zn^IIion (3d¹⁰, S = 0) and a

Gd^III(4f⁷, J = ⁷/₂, L = 0, S = ⁷/₂, ⁸S_7/2) ion with g_Gd= 2.028. The

χ_MT value of 12.73 cm³K mol⁻¹for the Zn^II−Tb^IIIcomplex at

300 K is compatible with the value of 11.82 cm³K mol⁻¹

expected for one Tb^III(4f⁸, J = 6, S = 3, L = 3, ⁷F₆) ion in the free-

ion approximation. On lowering the temperature, the χ_MT value

decreases gradually to reach a value of 10.41 cm³K mol⁻¹at

1.9 K. The decrease in the lower temperature region is due to the

crystal ﬁeld eﬀect on the Tb^IIIion that removes the 13-fold

degeneracy of the ⁷F₆ground state.²¹The χ_MT value of 15.02 cm³

K mol⁻¹for the Zn^II−Dy^IIIcomplex at 300 K is compatible with

the value of 14.17 cm³K mol⁻¹expected for one Dy^III(4f⁹,

J = 15/2, S = 5/2, L = 5, ⁶H_15/2) ion in the free-ion approxima-

tion. On lowering the temperature, the χ_MT value decreases

gradually to 11.34 cm³K mol⁻¹at 1.9 K. The decrease in the

lower temperature region is also due to the crystal ﬁeld eﬀect on

the Dy^IIIion that removes the 16-fold degeneracy of the ⁶H_15/2

ground state.

Cu^II−Gd^III

Cu^II−Tb^III

Cu^II−Dy^III

Ln−O(1)

Ln−O(2)

Ln−O(3)

Ln−O(4)

Ln−O(5)

Ln−O(7)

Ln−O(8)

Ln−O(9)

Ln−O(10)

Cu−O(2)

Cu−O(3)

Cu−O(6)

Cu−N(1)

Cu−N(2)

Ln···Cu

2.521(5)

2.388(4)

2.342(4)

2.599(4)

2.327(3)

2.443(4)

2.387(4)

2.375(4)

2.527(3)

1.969(4)

1.974(3)

2.231(3)

1.980(4)

1.982(5)

3.4373(6)

Zn^II−Gd^III

2.515(4)

2.371(3)

2.327(4)

2.597(4)

2.310(3)

2.432(4)

2.368(4)

2.361(3)

2.507(3)

1.969(4)

1.980(3)

2.235(3)

1.986(4)

1.983(5)

3.4268(6)

Zn^II−Tb^III

2.503(5)

2.366(4)

2.317(4)

2.590(4)

2.295(3)

2.424(4)

2.360(4)

2.344(4)

2.497(4)

1.961(4)

1.973(4)

2.235(3)

1.978(4)

1.980(6)

3.4175(6)

Zn^II−Dy^III

Ln−O(1)

Ln−O(2)

Ln−O(3)

Ln−O(4)

Ln−O(5)

Ln−O(7)

Ln−O(8)

Ln−O(9)

Ln−O(10)

Zn−O(2)

Zn−O(3)

Zn−O(6)

Zn−N(1)

Zn−N(2)

Ln···Zn

2.548(4)

2.357(3)

2.323(4)

2.609(4)

2.376(3)

2.443(4)

2.391(4)

2.363(3)

2.515(3)

2.069(4)

2.062(3)

1.998(3)

2.064(4)

2.063(5)

3.4499(6)

2.546(7)

2.351(5)

2.312(7)

2.605(7)

2.355(5)

2.430(6)

2.374(6)

2.352(5)

2.503(5)

2.068(6)

2.062(5)

1.999(5)

2.067(7)

2.062(8)

3.439(1)

2.530(5)

2.338(3)

2.298(4)

2.606(4)

2.343(3)

2.423(4)

2.369(4)

2.339(4)

2.483(3)

2.061(4)

2.060(3)

2.003(3)

2.063(4)

2.062(5)

3.4292(6)

The ﬁeld dependences of the magnetization up to 5 T for the

Zn^II−Ln^IIIcomplexes were measured at 1.9 K, and the M/Nβ

versus H plots are shown in Figure 3b. The M/Nβ versus H plots

of the Zn^II−Gd^IIIcomplex were well reproduced by the Brillouin

function for S = ⁷/₂and g = 2.06 (the solid line) whose g-value is

compatible with the value (g = 2.028) obtained from the mag-

netic susceptibility measurement. As shown in Figure 3b, upon

increasing the applied external magnetic ﬁeld, the magnetization

of the Zn^II−Tb^IIIcomplex (green triangle) increases to 5.24 Nβ

The reference theoretical magnetic susceptibilities expected for

the noninteracting d−f ions, using the d-ion with spin-only value

and the 4f-ion with free ion approximation, have been calculated

with the equation χ_M= χ_3d+ χ_4f, whereχ_3d= (Ng²β²/3kT)[S(S + 1)]

Figure 1. Molecular structure of acetato- and diphenolato-bridged binuclear Cu^II−Gd^IIIcomplex with the selected atom numbering scheme. The

hydrogen and ﬂuorine atoms are omitted for clarity. (a) Perspective view projected on equatorial N₂O₂plane. (b) Side view showing acetato-bridge.

E

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Table 3. Crystallographic Data for the Ni^II−Ln^IIIand Co^II−Ln^IIIComplexes [M(3-MeOsaltn)(MeOH)(ac)Ln(hfac)₂] (Ln = Gd^III,

Tb^III, and Dy^III)

Ni^II−Gd^III

Ni^II−Tb^III

Ni^II−Dy^III

formula

fw

C₃₃H₃₃N₂O₁₂F₁₂NiGd

1093.56

C₃₃H₃₃N₂O₁₂F₁₂NiTb

1095.24

C₃₃H₃₃N₂O₁₂F₁₂NiDy

1098.81

cryst syst

space group

a/Å

monoclinic

P2₁/c (No. 14)

17.0843(7)

12.7041(7)

19.5631(8)

110.2960(11)

3982.4(3)

100

monoclinic

P2₁/c (No. 14)

17.1879(6)

12.7311(3)

19.5571(6)

109.4290(10)

4035.8(2)

100

monoclinic

P2₁/c (No. 14)

17.1346(4)

12.6797(3)

19.4864(4)

109.7349(7)

3984.99(14)

100

b/Å

c/Å

β/deg

3

́

V/ Å

T/K

Z

4

D

_caldc/g cm⁻³

1.824

1.802

1.831

μ/cm⁻¹

22.46

23.17

24.54

R1 [I > 2σ(I)]

wR2 [all data]

reﬂns collected

unique reﬂns, R(int)

0.0498

0.0395

0.0407

0.1490

0.1040

0.1053

36 268

37 035

37 211

9013, 0.0578

Co^II−Gd^III

9126, 0.0287

Co^II−Tb^III

9035, 0.0489

Co^II−Dy^III

formula

fw

C₃₃H₃₃N₂O₁₂F₁₂CoGd

1093.79

C₃₃H₃₃N₂O₁₂F₁₂CoTb

1095.47

C₃₃H₃₃N₂O₁₂F₁₂CoDy

1099.04

cryst syst

space group

a/Å

monoclinic

P2₁/c (No. 14)

17.0988(4)

12.7846(3)

19.6171(4)

110.0599(7)

4028.17(14)

100

monoclinic

P2₁/c (No. 14)

17.1053(6)

12.7743(4)

19.6165(5)

110.0478(9)

4026.64(19)

100

monoclinic

P2₁/c (No. 14)

17.0871(4)

12.7293(4)

19.5855(5)

110.0456(7)

4001.92(16)

100

b/Å

c/Å

β/deg

V/Å³

T/K

Z

4

D

_calcd/g cm⁻³

1.803

1.807

1.824

μ/cm⁻¹

21.65

22.67

23.87

R1 [I > 2σ(I)]

wR2 [all data]

reﬂns collected

uniq reﬂns, R(int)

0.0371

0.0379

0.0377

0.0937

0.0949

0.0908

37 618

36 152

37 542

9175, 0.0352

9154, 0.0429

9033, 0.0313

at 5 T but did not reach the expected saturation value (9 Nβ for

the Tb^IIIion). This is due to the crystal ﬁeld eﬀect on the Tb^III

ion. The highest of the levels into which the ⁷F₆state is split do

not contribute to the magnetization, which therefore does not

reach the saturation value. This eﬀect is well-known for transition

metal ions with a signiﬁcant zero-ﬁeld-splitting and also applied

to rare earth ions except for Gd^IIIion which has an isotropic

compatible with the value of 10.00 cm³K mol⁻¹, expected for an

isolated S = 4 spin state resulting from ferromagnetic coupling

between the Cu^II(S = ¹/₂) and Gd^III(S = ⁷/₂) ions assuming g_Cu

=

g

_Gd= 2.00. Fits to the experimental data were performed assum-

ing for the Gd^IIIion an isotropic ⁸S_7/2state without orbital angular

momentum and using the following spin Hamiltonian, H =

g_CuβS_CuH + g_GdβS_GdH − 2J(Cu−Gd)S_Cu·S_Gdin which g_Gdand g_Cu

are the g factors for the Gd^IIIand Cu^IIions, H is the applied ﬁeld,

and J is the Heisenberg coupling constant between the two ions,

respectively. The solid line in Figure 4a shows the theoretical curve

with the best ﬁt parameters of g_Cu= 2.17, g_Gd= 2.03, J(Cu−Gd) =

+2.6 cm⁻¹, being consistent with an intramolecular ferromagnetic

interaction between the Cu^IIand Gd^IIIions and a negligible weak

intermolecular antiferromagnetic interaction.

8

ground S_7/2state. As shown in Figure 3b, upon increasing

the applied external magnetic ﬁeld, the magnetization of the

Zn^II−Dy^IIIcomplex increases to 5.59 Nβ at 5 T but did not reach

the expected saturation value (10 Nβ for the Dy^IIIion). This is

also due to the crystal ﬁeld eﬀect on the Dy^IIIion.

Magnetic Properties of Cu^II−Ln^IIIComplexes (Ln^III

=

Gd^III, Tb^III, and Dy^III). The temperature dependences of the

magnetic susceptibilities for the binuclear Cu^II−Ln^IIIcomplexes

are shown in Figure 4a. The χ_MT value of 8.81 cm³K mol⁻¹at

300 K of the Cu^II−Gd^IIIcomplex is compatible with the theo-

retical value of 8.25 cm³K mol⁻¹expected for a Cu^II(3d⁹, S = ¹/₂)

The χ_MT value of 12.76 cm³K mol⁻¹for Cu^II−Tb^IIIcomplex at

300 K is compatible with the value of 12.19 cm³K mol⁻¹ex-

pected for one Cu^IIand one Tb^IIImagnetically isolated ions. On

lowering the temperature, the χ_MT value decreases gradually to

12.05 cm³K mol⁻¹at 10.0 K and then decreases abruptly. The

χ_MT value of 15.18 cm³K mol⁻¹for the Cu^II−Dy^IIIcomplex at

300 K is comparable with the value of 14.54 cm³K mol⁻¹

expected for one Cu^IIand one Dy^IIImagnetically isolated ions.

On lowering the temperature, the χ_MT value decreases gradually

7

and a Gd^III(4f⁷, S = /₂) noninteracting ions. On lowering

temperature, the χ_MT value increases gradually to the value of

10.59 cm³K mol⁻¹at 1.9 K. The increase of the χ_MT value indicates

an intramolecular ferromagnetic interaction between Cu^IIand

Gd^IIIions. The value of χ_MT = 10.59 cm³K mol⁻¹at 1.9 K is

F

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

to reach a minimum value of 14.07 cm³K mol⁻¹at 26.0 K, and

then increases, ﬁnally showing a small decrease at the lowest

temperatures. The magnetic behavior at low temperatures of

these two complexes comes from the interplay of two eﬀects: (1)

a magnetic interaction between the Cu^IIand Ln^IIIions and (2) a

crystal ﬁeld splitting of the Tb^IIIor Dy^IIIion. The latter eﬀect

leads to a decrease of the χ_MT value in the lower temperature

region, as seen for the Zn^II−Tb^IIIand Zn^II−Dy^IIIcomplexes

where it is the only one to operate. The former eﬀect leads to an

increase of χ_MT if the magnetic interaction is ferromagnetic. The

increase of the χ_MT value in the low temperature region for the

Cu^II−Dy^IIIcomplex suggests that a ferromagnetic interaction

prevails against the eﬀect of the crystal ﬁeld splitting, while a less

clear situation occurs for the Cu^II−Tb^IIIcomplex for which the

two eﬀects seem to compensate and the χ_MT value shows a pla-

teau in the low temperature region before decreasing below 10 K.

The ﬁeld dependences of the magnetization up to 5 T for the

Cu^II−Ln^IIIcomplexes were measured at 1.9 K, and the M/Nβ vs

H plots are shown in Figure 4b. The M/Nβ versus H plots of the

Cu^II−Gd^IIIcomplex were well reproduced by the Brillouin

function for S = 4 and g = 2.09 (the solid line). Therefore, the

curve ﬁtting of the magnetization as well as that of the magnetic

susceptibility indicates clearly that the Cu^II−Gd^IIIcomplex has an

isolated S = 4 spin ground state resulting from the ferromagnetic

Table 4. Coordination Bond Distances (Å) and Ln^III···M^II

Distances (Å) for the Ni^II−Ln^IIIand Co^II−Ln^IIIComplexes

[M(3-MeOsaltn)(MeOH)(ac)Ln(hfac)₂] (Ln = Gd^III, Tb^III,

and Dy^III)

Ni^II−Gd^III

Ni^II−Tb^III

Ni^II−Dy^III

Ln−O(1)

Ln−O(2)

Ln−O(3)

Ln−O(4)

Ln−O(5)

Ln−O(8)

Ln−O(9)

Ln−O(10)

Ln−O(11)

Ni−O(2)

Ni−O(3)

Ni−O(6)

Ni−O(7)

Ni−N(1)

Ni−N(2)

Ln···Ni

2.690(4)

2.335(4)

2.324(4)

2.639(4)

2.340(4)

2.386(5)

2.423(4)

2.368(4)

2.462(4)

2.040(4)

2.036(4)

2.066(4)

2.185(4)

2.026(4)

2.031(5)

3.4009(7)

Co^II−Gd^III

2.703(4)

2.319(3)

2.311(4)

2.642(3)

2.336(3)

2.369(4)

2.410(3)

2.362(3)

2.458(3)

2.047(4)

2.037(3)

2.063(3)

2.172(3)

2.034(4)

2.038(4)

3.3985(5)

Co^II−Tb^III

2.699(4)

2.309(3)

2.298(4)

2.632(4)

2.312(3)

2.354(4)

2.404(3)

2.338(3)

2.440(3)

2.043(4)

2.038(3)

2.056(3)

2.172(4)

2.024(4)

2.035(5)

3.3839(5)

Co^II−Dy^III

Ln−O(1)

Ln−O(2)

Ln−O(3)

Ln−O(4)

Ln−O(5)

Ln−O(8)

Ln−O(9)

Ln−O(10)

Ln−O(11)

Co−O(2)

Co−O(3)

Co−O(6)

Co−O(7)

Co−N(1)

Co−N(2)

Ln···Co

2.706(3)

2.337(3)

2.331(3)

2.647(3)

2.349(3)

2.387(4)

2.427(3)

2.376(3)

2.461(3)

2.058(3)

2.056(2)

2.082(3)

2.233(3)

2.063(3)

2.070(4)

3.4059(5)

2.714(3)

2.324(3)

2.314(3)

2.647(3)

2.335(3)

2.369(4)

2.412(3)

2.356(3)

2.454(3)

2.060(3)

2.060(2)

2.083(3)

2.232(3)

2.063(3)

2.070(4)

3.3980(5)

2.714(4)

2.310(3)

2.300(3)

2.646(3)

2.321(3)

2.355(4)

2.403(3)

2.342(3)

2.444(3)

2.063(3)

2.052(3)

2.089(3)

2.228(3)

2.059(4)

2.065(4)

3.3827(5)

1

7

interaction between Cu^II(S = /₂) and Gd^III(S = /₂) ions.

Indeed, using the value of J(Cu−Gd) = +2.6 cm⁻¹from the mag-

netic susceptibility analysis, the S = 3 excited state is 20.8 cm⁻¹

(8J) above the S = 4 ground state and is not populated at the

temperature of 1.9 K at which the magnetization data have been

measured.

As shown in Figure 4b, upon increasing the applied external

magnetic ﬁeld, the magnetization of the Cu^II−Tb^IIIcomplex

(green triangle) increased to 6.65 Nβ at 5 T but did not reach the

expected saturation value (9 Nβ for the Tb^IIIion and 1 Nβ for the

Cu^IIion). This is due to the crystal ﬁeld eﬀect on the Tb^IIIion.

Upon increasing the applied external magnetic ﬁeld, the mag-

netization of the Cu^II−Dy^IIIcomplex increased to 6.23 Nβ at 5 T

but did not reach the expected saturation value (10 Nβ for the

Figure 2. Molecular structure of acetate- and diphenolato-bridged binuclear Co^II−Gd^IIIcomplex with the selected atom numbering scheme. The Co^II

ion assumes octahedral coordination geometry with equatorial tetradentate ligand and acetato and methanol oxygens. The hydrogen and ﬂuorine atoms

and solvent molecule are omitted for clarity. (a) Perspective view projected on equatorial N₂O₂plane. (b) Side view showing acetate-bridge and

coordination of methanol to Co.

G

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 3. (a) Plots of χ_MT vs T for binuclear Zn^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and Dy^III(blue)]. (b) Field dependence of the

magnetization for the Zn^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and Dy^III(blue)] at 1.9 K, as plots of M/Nβ vs H. The solid line of the

Zn^II−Gd^IIIcomplex represents the theoretical curve of M = NgβSB_S(y) for g = 2.06 and S = ⁷/₂.

Figure 4. (a) Plots of χ_MT vs T for binuclear Cu^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and Dy^III(blue)]. The solid line of the Cu^II−Gd^III

complex represents the theoretical curve for Cu^II−Gd^IIIcomplex with the best ﬁt parameters of g_Cu= 2.17, g_Gd= 2.03, J(Cu−Gd) = +2.6 cm⁻¹. (b) Field

dependence of the magnetization for the Cu^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and Dy^III(blue)] at 1.9 K, as plots of M/Nβ vs H. The

solid line of the Cu^II−Gd^IIIcomplex represents the theoretical curve for g = 2.09 and S = 4 spin state produced by ferromagnetic coupling between Cu^II

(S = ¹/₂) and Gd^III(S = ⁷/₂) ions.

Dy^IIIion and 1 Nβ for the Cu^IIion). This is also due to the crystal

ﬁeld eﬀect on the Dy^IIIion. These results demonstrate that the

Cu^II−Dy^IIIcomplex has a large magnetic moment and a magnetic

anisotropy in the same manner as the Cu^II−Tb^IIIcomplex.

Magnetic Properties of Ni^II−Ln^IIIComplexes (Ln^III= La^III,

Gd^III, Tb^III, and Dy^III). The temperature dependences of mag-

netic susceptibilities are shown in Figure 5a, as plots of χ_MT

versus T. The constant χ_MT value of ca. 1.2 cm³K mol⁻¹over

20−300 K of the Ni^II−La^IIIcomplex is reproduced by the theo-

retical value expected for a Ni^II(3d⁸, S = 1) ion with g_Ni= 2.2 and

one diamagnetic La^IIIion (4f⁰). On lowering the temperature, the

χ_MT value decreases gradually to reach a value of 0.40 cm³K

mol⁻¹at 1.9 K. The decrease in the lower temperature region is

due to the zero-ﬁeld-splitting (ZFS) of Ni^IIion. Fits to the experi-

mental data were performed using the spin Hamiltonian β, H =

g_NiβS_NiH + D_Ni[S_z²_Ni− S(S + 1)/3] in which g_Niis the g factor for

the Ni^IIion, H is the applied ﬁeld, S = 1 is the Ni^II

spin number, and D_Niis the ZFS parameter for Ni^II. A good

ﬁt was obtained for g_Ni= 2.21 and D_Ni= +8.6 cm⁻¹, both

parameters being in the range of values measured for other Ni^II

complexes.^20b

The χ_MT value of the Ni^II−Gd^IIIcomplex is 9.78 cm³K mol⁻¹

at 300 K, which is compatible with the calculated value of

8.88 cm³K mol⁻¹expected for a high spin Ni^II(S = 1) and a Gd^III

7

(S = /₂) noninteracting ions. On lowering temperature, the

χ_MT value increases gradually to reach a maximum value of

12.59 cm³K mol⁻¹at 3.0 K. The increase indicates an intramolecular

ferromagnetic interaction between Ni^IIand Gd^IIIions while the

very slight decrease below 3 K can be ascribed to a zero-ﬁeld-

splitting of the Ni^II, as suggested by the analysis of the magnetic

data for the Ni^II−La^IIIcomplex, see above. The maximum value

of 12.59 cm³K mol⁻¹at 3.0 K is close to the value of 12.38 cm³K

mol⁻¹expected for an isolated S = ⁹/₂spin state resulting from

ferromagnetic coupling between the Ni^II(S = 1) and Gd^III

(S = ⁷/₂) ions of the binuclear complex. Fits to the experimental

data were performed assuming for the Gd^IIIion an isotropic ⁸S_7/2

state and using the following spin Hamiltonian, H = g_NiβS_NiH +

g_GdβS_GdH + D_Ni[S_z²_Ni− S(S + 1)/3] − 2J(Ni−Gd)S_Ni·S_Gdin

H

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 5. (a) Plots of χ_MT vs T for the Ni^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), Dy^III(blue), and La^III(black)]. The orange solid line

represents the theoretical curve for the Ni^II−La^IIIcomplex with the best ﬁt parameters of g_Ni= 2.21 and D_Ni= +8.6 cm⁻¹. The black solid line represents

the theoretical curve for the Ni^II−Gd^IIIcomplex with the best ﬁt parameters of g_Ni= 2.20, g_Gd= 2.00, J(Ni−Gd) = +1.1 cm⁻¹, D = +7.1 cm⁻¹. (b) Field

dependence of the magnetization at 1.9 K for Ni^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), Dy^III(blue), and La^III(black)].The solid line for

the Ni^II−La^IIIcomplex represents the theoretical curve with the best ﬁt parameters of g_Ni= 2.28, D_Ni= 5.1 cm⁻¹. The black solid line for the Ni^II−Gd^III

complex represents the theoretical curve with the best-ﬁt parameters of g_9/2= 2.06, D_9/2= +0.16 cm⁻¹.

which g_Niand g_Gdare the g factors for the Ni^IIand Gd^IIIions, H

is the applied ﬁeld, D_Niis the ZFS parameter for Ni^II, and J(Ni−Gd)

is the Heisenberg coupling constant between the two ions. The

inclusion of the ZFS term is required to reproduce the experimental

data and is consistent with the analysis of the magnetization data, see

below. The best-ﬁt parameters to the data were g_Gd= 2.04, g_Ni= 2.20,

J(Ni−Gd) = +1.1 cm⁻¹, D_Ni= +7.1 cm⁻¹, with the values for g_Ni

and D_Nibeing compatible with the analysis of the magnetic data

for the Ni^II−La^IIIcomplex, see above. The calculated J(Ni−Gd)

value is lower than that reported by Costes et al. for a Ni^II−Gd^III

binuclear complex with two phenolato bridges of J = +3.6 cm⁻¹,²²

but slightly larger than the value observed by Chen et al. for a

Ni^II−Gd^IIIcompound with three phenolato bridges, which had

J = +0.56 cm⁻¹.²³

state and the Hamiltonian used for the ﬁt of the magnetic

susceptibility, see above. The best ﬁt, shown as a solid line in the

lower part of the Figure 5b, was obtained for g_Ni=2.28, and D_Ni=

5.1 cm⁻¹, values compatible with those obtained from the ﬁt

of χ_MT.

The data of the Ni^II−Gd^IIIcomplex are only qualitatively re-

produced by Brillouin curves for S = ⁹/₂with g = 2.03, demon-

9

strating the presence of an isolated S = /₂spin ground state

derived from the ferromagnetic coupling between Ni^II(S = 1)

and Gd^III(S = ⁷/₂) ions. Indeed, using the value of J(Ni−Gd) =

+1.1 cm⁻¹from the magnetic susceptibility analysis, the lowest

S = ⁷/₂excited state is 9.9 cm⁻¹(9J) above the S = ⁹/₂ground state

and is thus not signiﬁcantly populated at the temperature of 1.9 K

at which the magnetization data have been measured. The

data are quantitatively simulated including a ZFS term in the

Hamiltonian β, H = g_9/2βSH + D_9/2[S_z²− S(S + 1)/3] in which

g_9/2and D_9/2are the g factor and the ZFS parameters for the

S = ⁹/₂state, and the best ﬁt to the experimental data yields the

following values: g_9/2= 2.06 and D_9/2= +0.16 cm⁻¹(see solid line

in the upper part of Figure 5b). It is worth noting that the D_9/2

value obtained from the ﬁt of the magnetization data refers to the

The χ_MT value of 14.05 cm³K mol⁻¹at 300 K for the Ni^II−

Tb^IIIcomplex is slightly higher than the value of 12.82 cm³K

mol⁻¹expected for one Ni^II(S = 1) and one Tb^III(4f⁸, J = 6, S = 3,

L = 3, ⁷F₆) magnetically isolated ions. On lowering the tempera-

ture, the χ_MT value increases gradually to reach a maximum value

of 15.89 cm³K mol⁻¹at 6.0 K and then decreases abruptly. The

χ_MT value of 16.12 cm³K mol⁻¹at 300 K for the Ni^II−Dy^IIIcom-

plex is also slightly higher than the value of 15.17 cm³K mol⁻¹

expected for one Ni^II(S = 1) and one Dy^III(4f⁹, J = 15/2, S = 5/2,

9

S = /₂state and can be compared with the single ion value of

D_Ni= +5.1 cm⁻¹for Ni^IIobtained from a ﬁt of the magnetization

for the Ni^II−La^IIIcomplex using the Wigner−Eckart theorem.²⁴

For instance, using eqs 6.4.3 and 6.4.4 of ref 24 and taking into

account that the ZFS of the Gd^IIIion is negligible, it can be shown

that D_9/2= (1/36)D_Ni. This justiﬁes, at least qualitatively, the

smaller value of D_9/2, by over 1 order of magnitude.

6

L = 5, H_15/2) magnetically isolated ions, and the χ_MT value

decreases gradually with decreasing temperature to 14.98 cm³K

mol⁻¹at 12.0 K and then barely increases to 15.07 cm³K mol⁻¹at

7.0 K, before ﬁnally decreaseing abruptly.

The ﬁeld dependences of the magnetization for the Ni^II−Ln^III

complexes (Ln^III= La^III, Gd^III, Tb^III, and Dy^III) were measured

at 1.9 K, and the M/Nβ versus H curve is shown in Figure 5b.

We see that the magnetization M/Nβ for the Ni^II−La^IIIcomplex

assumes values much lower than those expected from the

Brillouin function for an S = 1 ground state: this is due to the

strong zero ﬁeld splitting, which leads to a split the S = 1 ground

state into two M_S= 0, 1 components, preventing the mag-

netization from reaching the saturation value of g_Swith S = 1. The

correct dependence of the magnetization as a function of the ﬁeld

H can be obtained through the full-matrix diagonalization of the

Hamiltonian matrix, using the three spin functions of the S = 1

As shown in Figure 5b, upon increasing the applied external

magnetic ﬁeld, the magnetization of Ni^II−Tb^IIIcomplex increases

up to 6.58 Nβ at 5 T but does not reach the expected saturation

value of 11 Nβ (9 Nβ for Tb^IIIion and 2 Nβ for Ni^IIion). This is

due to the crystal ﬁeld eﬀect on the Tb^IIIion. The magnetization

of Ni^II−Dy^IIIcomplex increases to 7.34 Nβ at 5 T but does not

reach the expected saturation value of 12 Nβ (10 Nβ for Dy^IIIion

and 2 Nβ for Ni^IIion), also due to the crystal ﬁeld eﬀect on the

Dy^IIIion.

Magnetic Properties of the Co^II−Ln^IIIComplexes (Ln^III=

La^III, Gd^III, Tb^III, and Dy^III). The temperature dependences of

I

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 6. (a) Plots of χ_MT vs T for the Co^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), Dy^III(blue), and La^III(black)]. The orange line for the

Co^II−La^IIIcomplex represents the theoretical curve with the parameters of g_Co= 2.46, D_Co= +42 cm⁻¹. The black solid line represents the theoretical

curve for the Co^II−Gd^IIIcomplex with the best ﬁt parameters of g_Co= 2.59, g_Gd= 2.09, J(Co−Gd) = +0.60 cm⁻¹, and D = +51 cm⁻¹. (b) Field

dependence of the magnetization at 1.9 K for Co^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), Dy^III(blue), and La^III(black)]. The solid line for

the Co^II−La^IIIcomplex represents the theoretical curves with g_Co= 2.59, D_Co= +49 cm⁻¹. The solid line for the Co^II−Gd^IIIcomplex represents the

theoretical curve for the best ﬁt parameters of g₅= 2.03, D₅= 0.14 cm⁻¹for the S = 5 ground state of the Co^II−Gd^IIIcomplex.

magnetic susceptibilities for the Co^II−Ln^IIIcomplexes (Ln^III

=

The best-ﬁt parameters to the data were g_Gd= 2.09, g_Co= 2.59,

J(Co−Gd) = +0.6 cm⁻¹, and D_Co= +51 cm⁻¹with g_Coand D_Co

values close to those obtained for the Co^II−La^IIIcomplex, see above.

The calculated J(Co−Gd) value is slightly lower than that reported by

Costes et al.²⁶for a Co^II−Gd^III, binuclear complex with two phenolato

bridges of J = +0.90 cm⁻¹, the only magnetic coupling constant

between Co^IIand Gd^IIIions, to the best of our knowledge.

La^III, Gd^III, Tb^III, and Dy^III) are shown in Figure 6a. The constant

χ_MT value of ca. 2.81 cm³K mol⁻¹over 120−300 K of the

Co^II−La^IIIcomplex is reproduced by the theoretical value expected

for a high spin Co^II(3d⁷, S = ³/₂) ion with g_Co= 2.45 and one

diamagnetic La^IIIion (4f⁰). On lowering the temperature, the

χ_MT value keeps constant in the temperature range 120−300 K

and then decreases gradually to reach a value of 1.86 cm³K mol⁻¹

at 1.9 K. The decrease in the lower temperature region is due to

a large zero-ﬁeld-splitting (ZFS) of Co^IIion. An isotropic spin

Hamiltonian is not strictly applicable to octahedral Co^II

complexes because of the strong spin−orbit splitting of the

⁴T_1gground term but can be employed to distorted octahedral

The χ_MT value of 16.04 cm³K mol⁻¹at 300 K for the

Co^II−Tb^IIIcomplex is higher than the value of 13.69 cm³K mol⁻¹ex-

pected for one Co^II(S = ³/₂) and one Tb^IIImagnetically isolated

ions. On lowering the temperature, the χ_MT value increases

gradually to reach a value of 18.97 cm³K mol⁻¹at 5.0 K and then

decreases. The χ_MT value of 18.29 cm³K mol⁻¹at 300 K for the

Co^II−Dy^IIIcomplex is higher than the value of 16.05 cm³K mol⁻¹

4

geometries where the orbital degeneracy of the T_1gstate is

3

expected for one Co^II(S = /₂) and one Dy^IIImagnetically

removed. Indeed, the experimental data for the Co^II−La^IIIcom-

plex could be ﬁtted using the spin Hamiltonian, H = g_CoβS_CoH +

D_Co[S_z²_Co− S(S + 1)/3] in which g_Cois the g factor for the Co^II

ion, H is the applied ﬁeld, S = ³/₂is the Co^IIspin number, and D_Co

is the ZFS parameter for Co^II. The best-ﬁt parameters to the data

were g_Co= 2.46 and D_Co= +42 cm⁻¹. Although quite high, the

calculated ZFS parameter D_Cois in the range of values measured

for other Co^IIcomplexes with a N₂O₄coordination.²⁵

isolated ions. On lowering the temperature, the χ_MT value

decreases gradually to reach a minimum value of 16.68 cm³K

mol⁻¹at 12.0 K, and then increases to 17.38 cm³K mol⁻¹at 1.9 K.

The ﬁeld dependences of the magnetization for the Co^II−Ln^III

complexes (Ln^III= La^III, Gd^III, Tb^III, and Dy^III) were measured at

1.9 K, and the M/Nβ versus H curve is shown in Figure 6b. The

magnetization M/Nβ for the Co^II−La^IIIcomplex assumes values

much lower than those expected from the Brillouin function for

an S = ³/₂ground state, due to the strong zero ﬁeld splitting. The

correct dependence of the magnetization as a function of the ﬁeld

H can be obtained through the full-matrix diagonalization of the

Hamiltonian matrix, using the four spin functions of the S = ³/₂

state and the Hamiltonian used for the ﬁt of the magnetic sus-

ceptibility, see above. The best ﬁt, as shown in the lower part of

the Figure 6b, was obtained for g_Co= 2.59, and D_Co= +49 cm⁻¹,

whose values are compatible with those obtained from the ﬁt of

χ_MT versus T.

The χ_MT value of the Co^II−Gd^IIIcomplex is 12.06 cm³K mol⁻¹

at 300 K, which is higher than the calculated value of 9.75 cm³

K mol⁻¹expected for a high spin Co^II(3d⁷, S = ³/₂) and a Gd^III

7

8

(4f⁷, J = /₂, L = 0, S = /₂, S_7/2) noninteracting ions. On

lowering temperature, the χ_MT value increases gradually to a

value of 15.54 cm³K mol⁻¹at 1.9 K. The increase indicates an

intramolecular ferromagnetic interaction between Co^IIand Gd^III

ions. Fits to the experimental data were performed assuming for

8

the Gd^IIIion an isotropic S_7/2state without orbital angular

momentum and using the following spin Hamiltonian, H =

g_CoβS_CoH + g_GdβS_GdH + D_Co[S_z²_Co− S(S + 1)/3] − 2J(Co−Gd)-

S_Co·S_Gdin which g_Coand g_Gdare the g factors for the Co^IIand Gd^III

ions, H is the applied ﬁeld, D_Cois the ZFS parameter for Co^II, and J is

the Heisenberg coupling constant between the two ions. Although

not necessarily required by a decrease of the χ_MT value in low

temperature region, a ZFS term has been added as suggested by

the data for the Co^II−La^IIIcomplex and to improve the ﬁtting.

The data for the Co^II−Gd^IIIcomplex are only qualitatively

reproduced by Brillouin curves for S = 5 and g = 1.99, demon-

strating the presence of an isolated S = 5 spin ground state derived

from the ferromagnetic coupling between Co^II(S = ³/₂) and Gd^III

(S = ⁷/₂) ions. Indeed, using the value of J(Co−Gd) = +0.6 cm⁻¹

from the magnetic susceptibility analysis, the lowest S = 4 excited

state is 6.0 cm⁻¹(10J) above the S = 5 ground state and is thus

J

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 7. (a) Plots of Δ(T) = (χ_MT)_CuLn− (χ_MT)_ZnLnvs temperature T for the Cu^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and Dy^III(blue)]

and theoretical value of one Cu^IIion (−). (b) Plots of Δ(H) = M_CuLn(H) − M_ZnLn(H) vs magnetic ﬁeld H for the Cu^II−Ln^IIIcomplexes

[Ln^III= Gd^III(red), Tb^III(green), and Dy^III(blue)]. Black solid line is the theoretical magnetization curve (Brillouin functions) of an independent Cu^II

ions with S = ¹/₂and g = 2.00.

barely populated at the temperature of 1.9 K at which the mag-

netization data have been measured. The data are quantitatively

simulated including a ZFS term in the Hamiltonian β, H = g₅βSH +

D₅[S_z²− S(S + 1)/3] in which g₅and D₅are the g factor and

the ZFS parameters for the S = 5 state, and the best ﬁt to the

experimental data yields the following values: g₅= 2.03 and D₅=

+0.14 cm⁻¹(see solid line in the upper part of Figure 6b). As

already seen for the Ni^II−Gd^IIIcomplex, the D₅value obtained

from the ﬁt of the magnetization data is related to the single ion

value of D_Co= +49 cm⁻¹for Co^IIobtained from a ﬁt of the mag-

netization for the Co^II−La^IIIcomplex using the Wigner−Eckart

theorem.²⁴It can be shown that D₅= (1/30)D_Co, thus justifying,

at least qualitatively, the smaller value of D₅, by almost 2 orders of

magnitude.

As shown in Figure 6b, upon increasing the applied external

magnetic ﬁeld, the magnetizations of Co^II−Tb^IIIand Co^II−Dy^III

complexes are increased to 7.36 and 8.47 Nβ at 5 T, respectively,

but did not reach the expected saturation values of 12 Nβ (9 Nβ

for Tb^IIIion and 3 Nβ for Co^IIion) and 13 Nβ (10 Nβ for Dy^III

ion and 3 Nβ for Co^IIion), respectively, due to the crystal ﬁeld

eﬀect on the Tb^IIIand Dy^IIIions.

When the Δ(T) value is larger than (χ_MT)_Cu= 0.375 cm³K mol⁻¹,

J_CuLn(T) is positive, and consequently the magnetic interaction

between Cu^IIand Ln^IIIions is ferromagnetic, while when the Δ(T)

value is smaller than 0.375 cm³K mol⁻¹, J_CuLn(T) is negative and

indicates antiferromagnetic Cu^II−Ln^IIIinteractions. The Δ(T) value

for the Cu^II−Gd^IIIcomplex is larger than 0.375 cm³K mol⁻¹over the

entire temperature region and increases on lowering the

temperature, indicating a ferromagnetic interaction, in agree-

ment with the quantitative ﬁtting in terms of the isotropic spin

Hamiltonian described above.

The Δ(T) values for the Cu^II−Tb^IIIcomplex are constant

and slightly below 0.375 cm³K mol⁻¹in the temperature region

35−300 K and then increase to values above 0.375 cm³K mol⁻¹

reaching a maximum around 10 K to ﬁnally decrease abruptly to a

negative value. The Δ(T) values for the Cu^II−Dy^IIIcomplex are

constant and slightly below 0.375 cm³K mol⁻¹in the tem-

perature region 80−300 K and monotonically increase above

0.375 cm³K mol⁻¹on decreasing the temperature.

While for the Cu^II−Dy^IIIcomplex the monotonical increase of

Δ(T) at low temperaures indicates a ferromagnetic Cu^II−Dy^III,

the Δ(T) versus T plot for the Cu^II−Tb^IIIcomplex does not give

a clear answer about the nature of the Cu^II−Tb^IIImagnetic

interaction. The failure of the present empirical approach for the

Cu^II−Tb^IIIcomplex is quite unexpected since the compared

Cu^II−Tb^IIIand Zn^II−Tb^IIIcomplexes have an isomorphous

structure. A possible explanation for the low temperature

decrease of Δ(T) may be the occurrence of weak intermolecular

antiferromagnetic interaction, for the Cu^II−Tb^IIIthan for the

Zn^II−Tb^IIIcomplex. Another possible explanation of this

behavior could be a higher crystal ﬁeld splitting experienced by

the Tb^IIIion in the Cu^II−Tb^IIIthan in the Zn^II−Tb^IIIcomplex so

that, at low temperatures, the (χ_MT)_Tbcontribution in Cu^II−Tb^III

is lower than that in Zn^II−Tb^IIIand their diﬀerence overcomes

the small interaction term J_CuLn(T). This eﬀect is probably

related to the larger size of the Zn^IIthan the Cu^IIion, see Table 2,

leading to a signiﬁcant variation of the coordination sphere

around the Ln^IIIion when passing from the Zn^II−Ln^IIIto the

Cu^II−Ln^IIIcomplex, and is observed only for the Cu^II−Tb^III

complex for which the interaction term J_CuLn(T) is smaller. How-

ever, the increase of Δ(T) in the temperature range 50−10 K

suggests an overall ferromagnetic Cu^II−Tb^IIIinteraction.

Magnetic Interaction between Cu^IIand Ln^IIIIons. Due to

the orbital contribution of the Ln^IIIcomponent, it is diﬃcult to

theoretically simulate the magnetic properties of the Cu^II−Tb^III

and Cu^II−Dy^IIIcomplexes. The nature of the magnetic interac-

tion between Cu^IIand Ln^IIIions were investigated by an empir-

ical approach^27,28based on a comparison of the magnetic sus-

ceptibilities of Cu^II−Ln^IIIwith those of the isostructural Zn^II−Ln^III

complexes involving the diamagnetic Zn^IIion. According to

this approach, deﬁning Δ(T) = (χ_MT)_CuLn− (χ_MT)_ZnLn, we can

write the following equation: Δ(T) = [(χ_MT)_Cu+ (χ_MT)_Ln

+

J_CuLn(T)] − [(χ_MT)_Zn+ (χ_MT)_Ln] = (χ_MT)_Cu+ J_CuLn(T), where

(χ_MT)_Znis diamagnetic and (χ_MT)_Curepresents the χ_MT value

imputable to an isolated Cu^IIion (Curie constant: (χ_MT)_Cu

=

0.375 cm³K mol⁻¹) and the temperature-dependent contribu-

tion J_CuLn(T) is related to the nature of the overall exchange

interactions between the Cu^IIand Ln^IIIions. It should be noted

that in general the M^II−Ln^IIImagnetic interaction J_MLn(T) is

weak and the pronounced Δ(T) value appears at the lower tem-

perature region. The plots of Δ(T) versus T are given in Figure 7a.

The above equation gives the relation of J_CuLn(T)=Δ(T)−(χ_MT)_Cu.

K

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 8. (a) Plots of Δχ_M(T) = (χ_MT)_NiLn− (χ_MT)_ZnLn− (χ_MT)_NiLavs T for the Ni^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and

Dy^III(blue)]. (b) Plots of Δ(H) = M_NiLn(H) − M_ZnLn(H) − M_NiLa(H) vs H for the Ni^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green), and

Dy^III(blue)].

Although not investigated by this empirical approach, ferro-

magnetic Cu^II−Ln^IIIinteractions, (Ln^III= Gd^III, Dy^III, Tb^III) have

been reported for dinuclear [Cu^IILn^III] complexes by Costes

et al.,^28a,b2D ladder-type [Ln^IIICu^II] complexes by Kahn et al.,²⁹

above M_Cu(H) in the whole range of H, thus conﬁrming the

ferromagnetic nature of the interaction between Cu^IIand Gd^III

ions. Also the Δ(H) versus H plot for the Cu^II−Dy^IIIcomplex lies

above M_Cu(H), although slightly lower at high applied ﬁeld,

indicating again a ferromagnetic Cu^II−Dy^IIIinteraction. On the

other hand, the Δ(H) versus H plot for the Cu^II−Tb^IIIcomplex

lies below M_Cu(H) at low magnetic ﬁeld and above M_Cu(H) at

magnetic ﬁeld above 1 T: the behavior at low ﬁeld is probably due

to larger intermolecular interactions or a higher crystal ﬁeld

splitting experienced by the Tb^IIIion in the Cu^II−Tb^IIIthan in the

Zn^II−Tb^IIIcomplex (see discussion above) whose eﬀect is

maximum at the lowest considered temperature of 1.9 K. How-

ever, the increase at higher ﬁeld conﬁrms that also the Cu^II−Tb^III

interaction is ferromagnetic. The result from the ﬁeld dependent

magnetization measurements is consistent with the result from

the temperature-dependent magnetic susceptibility measure-

ments, taking into account that the Δ(T) and Δ(H) values are

not always above the reference Cu curves over whole regions of T

and H. The detailed explanation for the anomalous Δ(T) and

Δ(H) behaviors is left to a further study.

2

3

and trinuclear complexes Cu^II−Ln^III−Cu^IIcomplexes

[Ln^IIIL₂(NO₃)₂{Cu^II(CH₃OH)}₂](NO₃)(CH₃OH) (L = 2,6-

bis(acetylacetonato)pyridine) by Ishida et al.³⁰and Okawa

et al.³¹An overview of the plots of Δ(T) versus T of Figure 7a

shows that the plot for the Cu^II−Gd^IIIcomplex lies on the top,

that for the Cu^II−Dy^IIIcomplex in the middle, and that for the

Cu^II−Tb^IIIcomplex on the bottom, indicating the J_CuLn(T) is in

the order J_CuGd(T) > J_CuDy(T) > J_CuTb(T). This trend slightly

diﬀers from the results by Ishida et al.,³⁰who studied the

ferromagnetic exchange coupling of Cu^II−Ln^III−Cu^IIcomplexes

[Ln^IIIL₂(NO₃)₂{Cu^II(CH₃OH)}₂](NO₃)(CH₃OH) by high-

frequency electron paramagnetic resonance and magnetization

studies and revealed that the exchange coupling J(Cu−Ln) was in

the order J(Cu−Gd) > J(Cu−Tb) > J(Cu−Dy).

The same empirical approach was applied to the magnet-

ization curves at constant temperature and gives complementary

information on the magnetic interaction between Cu^IIand Ln^III

ions at the lowest considered temperature of 1.9 K. Let us now

deﬁne Δ(H) = M_CuLn(H) − M_ZnLn(H) = [M_Cu(H) + M_Ln(H) +

Magnetic Interaction between Ni^IIand Ln^IIIIons. Since

the magnetic properties of the Ni^II−Ln^IIIcomplexes involve

important orbital contributions from the Ni^IIand Ln^IIIions, the

nature of the magnetic interaction between Ni^IIand the 4f

magnetic ions was investigated by the same empirical approach

employed above.^22,23Now the Ni^II−Ln^IIIcomplexes involve an

orbital contribution from the Ni^IIion too, and therefore, we have

included also χ_MT values of Ni^II−La^IIIcomplex in the empirical

approach. The diﬀerence between the χ_MT values for the Ni^II−

J_CuLn(H)] − [M_Zn(H) + M_Ln(H)] = M_Cu(H) + J_CuLn(H), where

M is the magnetization measured on the complex denoted by the

subscript and the magnetization of the diamagnetic Zn^IIcom-

ponent M_Zn(H) and the contribution from the Zn^II−Ln^III

interaction should be zero. This can be written as Δ(H) =

M_Cu(H) + J_CuLn(H), where M_Cu(H) is the magnetization for an

independent Cu^IIion, that can be calculated by the

corresponding Brillouin function, while the extra contribution

Ln^III, Zn^II−Ln^III, and Ni^II−La^IIIcomplexes, Δ(T) = (χ_MT)_NiLn

−

(χ_MT)_ZnLn− (χ_MT)_NiLa= [(χ_MT)_Ni+ (χ_MT)_Ln+ J_NiLn(T)] −

[(χ_MT)_Zn+ (χ_MT)_Ln] − [(χ_MT)_Ni+ (χ_MT)_La] = J_NiLn(T), where

the terms (χ_MT)_Znand (χ_MT)_Lareferring to diamagnetic species

are zero and the temperature-dependent contribution J_NiLn(T) is

related to the nature of the overall exchange interactions between

the Ni^IIand Ln^IIIions. Since the Ni^II−Ln^IIImagnetic interaction

J_CuLn(H) is related to the nature of the overall exchange magnetic

interactions between the Cu^IIand Ln^IIIions. As the quantity

(Δ(H) − M_Cu(H)) represents the deviation from the limit situa-

tion in which Cu^IIand Ln^IIIions are magnetically independent,

positive values of J_CuLn(H), i.e., Δ(H) lying above M_Cu(H), indi-

cate ferromagnetic Cu^II−Ln^IIIinteractions while negative values

of J_CuLn(H), i.e., Δ(H) lying below M_Cu(H), indicate antiferro-

magnetic Cu^II−Ln^IIIinteractions. The values of the parameter

Δ(H) are plotted in Figure 7b as a function of the ﬁeld and

compared with the Brillouin function for one independent Cu^II

ion. The Δ(H) versus H plot for the Cu^II−Gd^IIIcomplex lies well

J

_NiLn(T) is weak, a positive or a negative value of the Δ(T) in the

lower temperature region is directly related to a ferro- or anti-

ferromagnetic interaction, respectively. Figure 8a reports the

values of Δ(T) versus T for the Ni^II−Gd^III, Ni^II−Tb^III, and

Ni^II−Dy^IIIcomplexes. Since J_NiLn(T) = Δ(T), when Δ(T) and then

J

_NiLn(T) is positive, the magnetic interactions between Ni^IIand

L

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 9. (a) Plots of Δ(T) = (χ_MT)_CoLn− (χ_MT)_CoLa− (χ_MT)_ZnLnvs temperature T for the Co^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III

(green), and Dy^III(blue)]. (b) Plots of Δ(H) = M_CoLn(H) − M_CoLa(H) − M_ZnLn(H) vs H for the Co^II−Ln^IIIcomplexes [Ln^III= Gd^III(red), Tb^III(green),

and Dy^III(blue)].

Ln^IIIions are ferromagnetic, while negative values of Δ(T) and

then of J_NiLn(T) indicate antiferromagnetic Ni^II−Ln^IIIinter-

actions. As shown in Figure 8a, the Δ(T) values for Ni^II−Gd^III

and Ni^II−Tb^IIIcomplexes are positive in the whole temperature

range 1.9−300 K and show a marked increase below 100 K, while

those for Ni^II−Dy^IIIcomplex are slightly negative at room tem-

perature but become positive on lowering the temperature, also

showing a marked increase below 100 K. This empirical approach

thus shows for all three Ni^II−Gd^III, Ni^II−Tb^III, and Ni^II−Dy^III

complexes an intramolecular d−f ferromagnetic interaction.

The same empirical approach was applied to the magnet-

ization values at constant temperature and gives complementary

information on the magnetic interaction between Ni^IIand Ln^III

ions. Let us now deﬁne Δ(H)=M_NiLn(H)−M_ZnLn(H)−M_NiLa(H)=

[M_Ni(H) + M_Ln(H) + J_NiLn(H)] − [M_Zn(H) + M_La(H)] − [M_Ni(H)

+ M_La(H)], where M is the magnetization measured on the complex

denoted by the subscript. The terms of M_Zn(H) and M_La(H)

consisting of diamagnetic species should be zero. This parameter

can be written as Δ(H) = J_NiLn(H), where J_NiLn(H) is related to

the nature of the overall exchange magnetic interactions between

the Ni^IIand Ln^IIIions. As the quantity Δ(H) represents the

deviation from the limit situation in which Ni^IIand Ln^IIIions are

magnetically independent, positive values of J_NiLn(H) indicate

ferromagnetic Ni^II−Ln^IIIinteractions while negative values of

for the Ni^II−Tb^IIIand Ni^II−Dy^IIIinteractions with respect to that

of the Cu^II−Tb^IIIand Cu^II−Dy^IIIinteractions, see discussion above.

This result slightly diﬀers from the results by Pasatoiu et al.,^20b

who studied the ferromagnetic exchange coupling of closely

related binuclear Ni^II−Ln^IIIcomplexes, [Ni^II(CH₃CN)(H₂O)(3-

MeOsaltn)Ln^III(NO₃)₃(H₂O)₃], with the same tetradentate

ligand for Ni^IIion but, diﬀerent from our compounds, without

an acetate ligand bridging Ni^IIand Ln^IIIions, ﬁnding that the

exchange coupling J_Ni‑Lnwas in the order J_NiDy> J_NiTb. This

diﬀerent behavior indicates that the magnetic Ni^II−Ln^III

interactions are sensible to small changes of the number and

the nature of bridging ligands. In addition to the existence or

absence of a bridging acetato, the notable structural diﬀerence is

seen in that the bridging core of Ni^IIO₂Ln^IIIis almost planar for

the Ni^II−Ln^IIIcomplexes by Pasatoiu et al., while the dihedral

angle between Ni^IIO₂and O₂Ln^IIIis 13.1° for the present

complexes.

Magnetic Interaction between Co^IIand Ln^IIIIons. Since

the magnetic properties of the Co^II−Ln^IIIcomplexes involve

orbital contributions from both Co^IIand Ln^IIIions, the nature of

the magnetic interaction between Co^IIand the 4f magnetic ions

was investigated by the same empirical approach for Ni^II−Ln^III

complexes. The diﬀerence between the χ_MT values for the Co^II−

Ln^III, Zn^II−Ln^III, and Co^II−La^IIIcomplexes is Δ(T) = (χ_MT)_CoLn

− (χ_MT)_ZnLn− (χ_MT)_CoLa= J_CoLn(T), where the temperature-

dependent contribution J_CoLn(T) is related to the nature of the

overall exchange interactions between the Co^IIand Ln^IIIions.

Figure 9a reports the values of Δ(T) versus T for the Co^II−Gd^III,

Co^II−Tb^III, and Co^II−Dy^IIIcomplexes. Since J_CoLn(T) = Δ(T),

when Δ(T) and then J_CoLn(T) is positive, the magnetic interac-

tions between Co^IIand Ln^IIIions are ferromagnetic, while nega-

tive values of Δ(T) and then of J_CoLn(T) indicate antiferro-

magnetic Co^II−Ln^IIIinteractions. The Δ(T) values for all

the Co^II−Ln^IIIcomplexes are positive in the whole temper-

ature range 1.9−300 K and increase abruptly on lowering the

temperature, indicating a ferromagnetic interaction. This

empirical approach applied to the magnetic susceptibili-

ties thus indicates ferromagnetic interactions for all three

Co^II−Ln^IIIcomplexes.

J

_NiLn(H) indicate antiferromagnetic Ni^II−Ln^IIIinteractions.

The values of the parameter Δ(H) are plotted in Figure 8b as

the function of the magnetic ﬁeld. The Δ(H) versus H plots for

the all three Ni^II−Ln^IIIcomplexes lie well above zero in the whole

range of H, thus suggesting the ferromagnetic nature of the

interaction between Ni^IIand Ln^IIIions for all three Ni^II−Gd^III,

Ni^II−Tb^III, and Ni^II−Dy^IIIcomplexes. The ferromagnetic feature

derived from the magnetization measurements is thus consistent

with the result from the magnetic susceptibility measurements.

Regarding the peculiar Δ(H) proﬁles, showing a maximum

around 0.4−0.7 T, the detailed explanation is left to a further

study.

Overall, both Δ(T) versus T and Δ(H) versus H plots in

Figure 8 show that the plot for the Ni^II−Gd^IIIcomplex lies close

(on the top at high temperatures and ﬁelds) to that for the

Ni^II−Tb^IIIcomplex and that for the Ni^II−Dy^IIIcomplex on the

bottom, indicating that the ferromagnetic Ni^II−Ln^IIIinteractions

are in the order J_NiGd≥ J_NiTb> J_NiDy, with an inversion of the order

The same empirical approach was applied to the magnet-

ization values at the constant temperature of 1.9 K and gives com-

plementary information on the magnetic interaction between

M

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Co^IIand Ln^IIIions. Let us now deﬁne Δ(H) = M_CoLn(H) −

M_ZnLn(H) − M_CoLa(H), where M is the magnetization measured

on the complex denoted by the subscript. This parameter can

be written as Δ(H) = J_CoLn(H), where J_CoLn(H) is related to the

nature of the overall exchange magnetic interactions between

the Co^IIand Ln^IIIions. As the quantity Δ(H) represents the

deviation from the limit situation in which Co^IIand Ln^IIIions are

magnetically independent, positive values of J_CoLn(H) indicate

ferromagnetic Co^II−Ln^IIIinteractions while negative values of

J_CoLn(H) indicate antiferromagnetic Co^II−Ln^IIIinteractions. The

values of the parameter Δ(H) are plotted in Figure 9b versus the

magnetic ﬁeld H. The Δ(H) versus H plots lie well above zero, in

the whole range of H for the Co^II−Gd^IIIand Co^II−Dy^III

complexes, assuming small negative values only at high ﬁeld

(above 1.5 T) for Co^II−Tb^III, thus conﬁrming the ferromagnetic

nature of the interaction between Co^IIand Ln^IIIions. Although

the Δ(T) versus T and Δ(H) versus H plots in Figure 9 do not

give a very clear picture when considered in the whole considered

temperature and ﬁeld ranges, at low temperatures and ﬁelds

the plot for the Co^II−Gd^IIIcomplex lies close to that for the

Ni^II−Tb^IIIcomplex (above at high temperatures) and that for the

Co^II−Dy^IIIcomplex on the bottom, suggesting that the

Figure 10. Plots of the in-phase (χ_M′) and out-of-phase (χ_M′′) ac

susceptibilities as a function of temperature and frequency in the range

10−10 000 Hz, at H_ac= 5.0 Oe, and with a dc bias ﬁeld of 1000 G for (a)

the Cu^II−Tb^IIIcomplex, (b) the Cu^II−Dy^IIIcomplex.

the Tb^IIIion is in a less symmetrical ligand ﬁeld, it has an easy-axis

anisotropy and shows SMM behavior, whereas when it is in a

more symmetrical environment, it has an easy-plane anisotropy

and exhibits non-SMM behavior. The present Cu^II−Tb^IIIand

Cu^II−Dy^IIIcomplexes have a rather weaker ferromagnetic

interaction between Cu^IIand Ln^IIIions and have a symmetrical

ligand ﬁled around Ln^IIIion.

ferromagnetic Co^II−Ln^IIIinteractions are in the order J_CoGd

≥

J_CoTb> J_CoDy, roughly the same order found for the Ni^II−Ln^III

interactions.

Alternating Current Magnetic Properties of Ni^II−Ln^III

complexes (Ln^III= Gd^III, Tb^III, and Dy^III). The Ni^II−Ln^III

complexes exhibiting ferromagnetic interactions have been ex-

amined by ac magnetic susceptibility measurements. Figure S4

shows the results of the ac magnetic susceptibility measurements

in a 3.0 Oe ﬁeld oscillating at the indicated frequencies (10−

1000 Hz) and with a zero dc ﬁeld for Ni^II−Gd^III, Ni^II−Tb^III, and

Ni^II−Dy^IIIcomplexes, as plots of χ_M′ and χ_M′′, respectively. As

seen in Figure S4, three Ni^II−Ln^IIIcomplexes show no frequency-

dependent signals in the same temperature range, showing the

absence of SMM behavior and demonstrating the inﬂuence of

the 3d metal ion on the SMM behavior observed for the Cu^IILn^III

which therefore is not the sole fact of the Tb^IIIor Dy^IIIcenters. To

evaluate a possible SMM behavior for the present systems, we

further investigated χ_M′ and χ_M′′ with a dc bias ﬁeld of 1000 Oe.

Figure 11a−c shows the results of the ac magnetic susceptibility

measurements for the Ni^II−Gd^III, Ni^II−Tb^III, and Ni^II−Dy^III

complexes, respectively. The Ni^II−Gd^IIIcomplex shows practi-

cally no frequency-dependent signal, while the Ni^II−Dy^III

complex shows a small dependence at the very low temperature.

As for the Ni^II−Tb^IIIcomplex, notable frequency dependence

emerged. The Ni^II−Tb^IIIcomplex has higher frequency depen-

dence than that of the Ni^II−Dy^IIIcomplex. On cooling, an increase

of χ_M′′ was found together with a decrease of χ_M′. Cole−Cole

plots and Arrhenius plots of Ni^II−Tb^IIIcomplex are shown in

Figure 11d,e, respectively. When we plotted the χ_M′′ against χ_M′

Alternating Current Magnetic Properties of Cu^II−Ln^III

Complexes (Ln^III= Tb^IIIand Dy^III). The Cu^II−Ln^IIIcomplexes

exhibiting ferromagnetic interactions have been examined by ac

magnetic susceptibility measurements, since one of the

characteristics of an SMM is the observation of an out-of-phase

(χ_M′′) ac susceptibility signal. The temperature dependent ac

magnetic measurements were carried out in a 3.0 Oe ﬁeld

oscillating at the indicated frequencies (10−1000 Hz) and with a

zero dc ﬁeld, down to a minimum temperature of 1.8 K. The

Cu^II−Gd^IIIshows no frequency dependence under the

experimental conditions. Figure S3 shows the results of the ac

magnetic susceptibility measurements for the Cu^II−Tb^IIIand

Cu^II−Dy^IIIcomplexes, as plots of χ_M′ and χ_M′′ versus T in the

1.8−20 K temperature range. As seen in Figure S3, the

Cu^II−Tb^IIIcomplex shows no frequency-dependent signals in

the same temperature range, while the Cu^II−Dy^IIIcomplex shows

a very small dependence at the very low temperature. To evaluate

a possible SMM behavior for the present systems, we further

investigated χ_M′ and χ_M′′ with a dc bias ﬁeld of 1000 Oe. The dc

bias was applied to reduce a possible quantum tunneling of the

magnetization.³²The ac amplitude was 5.0 Oe ﬁeld and

frequencies of 10−10 000 Hz down to a lowest temperature of

1.8 K. Figure 10a,b shows the results of the ac magnetic

susceptibility measurements for the Cu^II−Tb^IIIand Cu^II−Dy^III

complexes. The Cu^II−Tb^IIIcomplex shows only slightly

frequency-dependent signals, and Cu^II−Dy^IIIcomplex shows a

small dependence at the very low temperature; an indication of

χ_M′′ upsurge was then observed for the latter complex, but no

peak appeared above 1.8 K. The Cu^II−Dy^IIIcomplex has higher

frequency dependence than that of the Cu^II−Tb^IIIcomplex.

It has been known that the appearance of ac signals for

Cu^II−Ln^IIISMMs is aﬀected not only by the intramolecular magnetic

interaction but also by the symmetry of the ligand ﬁeld. Kajiwara

et al. reported a correlation between the symmetry of the ligand

ﬁeld and the magnetic anisotropy of Tb^III−Cu^IIdinuclear

systems. He studied two related Cu^II−Tb^IIIcomplexes, which

were prepared by the reactions [Tb^IIICu^II(o-vanilate)₂(NO₃)₃]

with either methoxypropylamine or ethoxyethylamine.¹³When

at various temperatures according to the Cole−Cole analysis,³³

a

semicircle was clearly drawn at each temperature. The α value is

considerably small (α = 0.221(7) at 2 K) in the Debye model,

χ(ω) = χ_S+ (χ_T− χ_S)/(1 + (iωτ)^1−α) where χ_Tand χ_Sare the

isothermal and adiabatic susceptibilities, respectively.³⁴This

ﬁnding guarantees a single relaxation process for this complex.

The Arrhenius plot³⁵of the Ni^II−Tb^IIIcomplex shows a straight

line for the χ_M′′ peak, and the activation energy (Δ) for the

magnetization reversal was estimated as Δ/k_B= 14.9(6) K with

τ₀= 2.1(5) × 10⁻⁷s, where τ₀stands for the pre-exponential

factor in the Arrhenius equation, ln(2πν) = −ln (τ₀) − Δ/k_BT.

The linear Arrhenius behavior down to 2 K indicates that the

relaxation takes place mainly via thermal activation and that the

N

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 11. Plots of the in-phase (χ_M′) and out-of-phase (χ_M′′) ac susceptibility of (a) the Ni^II−Gd^III, (b) Ni^II−Tb^III, and (c) Ni^II−Dy^IIIcomplexes

measured at a dc bias ﬁeld of 1000 Oe, (d) Cole−Cole plot, and (e) Arrhenius plot of the Ni^II−Tb^IIIcomplex.

quantum tunneling of the magnetization is practically negligible

at the 1000 Oe dc bias ﬁeld. The present ac susceptibility

behavior is typical of SMMs.

Alternating Current Magnetic Properties of Co^II−Ln^III

Complexes (Ln^III= Gd^III, Tb^III, and Dy^III). The Co^II−Ln^III

complexes exhibiting ferromagnetic interactions have been ex-

amined by ac magnetic susceptibility measurements. Figure S5

shows the results of the ac magnetic susceptibility measurements

for Co^II−Gd^III, Co^II−Tb^III, and Co^II−Dy^IIIcomplexes, as plots of

χ_M′ and χ_M′′. As shown in Figure S5, Co^II−Gd^IIIand Co^II−Dy^III

complexes show no frequency-dependent signals in the same

temperature range. On the other hand, out-of-phase signal (χ_M′′)

of Co^II−Tb^IIIcomplex exhibits frequency-dependent behavior.

This behavior is indicative of the slow relaxation of the mag-

netization, which is similar to those found for some lanthanide

paramagnetic ions. Slow relaxation of the magnetization is

directly responsible for behavior in SMM or SCM systems. Un-

fortunately, due to the 1.8 K temperature limit of the instrument,

a maximum in the χ_M′′ signal was not observed at frequencies as

high as 1000 Hz. Slow relaxation of the magnetization is directly

responsible for behavior in SMM systems.

We moved to study the SMM behavior with a dc bias ﬁeld of

1000 Oe applied in this system. Figure 12a−c show the results of

the ac magnetic susceptibility measurements for the Co^II−Gd^III,

Co^II−Tb^III, and Co^II−Dy^IIIcomplexes, respectively. The

Co^II−Gd^IIIcomplex shows practically no frequency-dependent

signal, while the Co^II−Dy^IIIcomplex exhibits a small dependence at

the very low temperature. Figure 12b displays the results on

the Co^II−Tb^IIIcomplex, clarifying a drastic improvement of fre-

quency dependence. The Co^II−Tb^IIIcomplex has higher fre-

quency dependence than that of the Co^II−Dy^IIIcomplex.

According to the Cole−Cole analysis, a semicircle was clearly

drawn at each temperature. The α value is considerably small

(α = 0.241(6) at 2 K), suggesting the presence of a single

relaxation process for this complex. TheArrheniusplotoftheCo^II−

Tb^IIIcomplex (Figure 12e) aﬀorded Δ/k_B= 17.0(4) K with τ₀=

6.1(10) ×10⁻⁸s. The linear Arrhenius behavior down to 2 K indicates

It is worth comparing the magnetic properties of our

Ni^II−Ln^IIIcomplexes with those of similar binuclear Ni^II−Ln^III

complexes. Pasatoiu et al. reported the structure and magnetic

properties of the closely related binuclear Ni^II−Ln^IIIcomplexes,

[Ni^II(CH₃CN)(H₂O)(3-MeOsaltn)Ln^III(NO₃)₃(H₂O)₃],^20b

with the same tetradentate ligand for Ni^IIion but, diﬀerently from

our compounds, without an acetate ligand bridging Ni^IIand Ln^III

ions. Due to the acetato-bridging of Ni^IIand Ln^IIIions at their

apical sites and at the same side, the Ni−O₂−Ln bridging core

is more bent for our acetato-bridged compounds. While our

Ni^II−Tb^IIIand Ni^II−Dy^IIIcomplexes show a pronounced and a

slight frequency dependence of the ac susceptibilities under 1000 dc

bias ﬁeld, respectively, the [Ni^II−Tb^III] and [Ni^II−Dy^III]

complexes of ref 20b show a slight and a moderate frequency

dependence under the same conditions, respectively. Colacio

et al. reported the structure and magnetic properties of acetato-

or nitrato-bridged binuclear Ni^II−Dy^IIIcomplexes with N,N′,N′′-

trimethyl-N,N′′-bis(2-hydroxy-3-methoxy-5-methylbenzyl)-di-

ethylenetriamine, wherein for the nitrato-bridged complex a

stronger ferromagnetic interaction between Ni^IIand Dy^IIIions is

observed and an apparent frequency dependence of ac magnetic

susceptibility is observed.³⁶The diﬀerence in the magnetic

properties between two complexes is ascribed to the planarity of

Ni−O₂−Dy bridging core, in which the nitrato-bridged complex

with a more planar Ni−O₂−Dy core gives a stronger ferromagnetic

interaction between Ni^IIand Dy^IIIion and apparent frequency-

dependent ac susceptibility. These data suggest that the acetate-

bridging of the present Ni^II−Ln^IIIcomplexes gives a deviation from

planar Ni−O₂−Dy bridging core to lead a diﬀerent tendency of

ferromagnetic interaction between Ni^IIand Ln^IIIions.³⁶

O

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 12. Plots of the in-phase (χ_M′) and out-of-phase (χ_M′) ac susceptibility of (a) the Co^II−Gd^III, (b) Co^II−Tb^III, and (c) Co^II−Dy^IIIcomplexes

measured at a dc bias ﬁeld of 1000 Oe, (d) Cole−Cole plot, and (e) Arrhenius plot of the Co^II−Tb^IIIcomplex.

that the relaxation takes place mainly via thermal activation and that

the quantum tunneling of the magnetization is practically negligible at

the 1000 Oe dc bias ﬁeld. From these ac susceptibility analyses, the

Co^II−Tb^IIIcomplex is concluded to be an SMM.

While some of the results are not simply indicative of the

magnetic interaction, all these complexes suggested ferromag-

netic interactions between M^IIand Ln^IIIions, in agreement

with previous studies indicating that these types of Cu^II−Ln^III

(Ln^III= Gd^III, Tb^III, and Dy^III) complexes usually showed

ferromagnetic interaction. The magnitude of the ferromag-

netic interaction, J_MLn(T) and J_MLn(H), are in the order

Cu^II−Gd^III> Cu^II−Dy^III> Cu^II−Tb^III, while those are in the order

of M^II−Gd^III≈ M^II−Tb^III> M^II−Dy^IIIfor M^II= Ni^IIand Co^II.

The alternating current magnetic measurement revealed that

the Ni^II−Gd^IIIand Co^II−Gd^IIIcomplexes show practically no

frequency-dependent signal, the Ni^II−Dy^IIIand Co^II−Dy^III

complexes show a small dependence at the very low tem-

perature, and the Ni^II−Tb^IIIand Co^II−Tb^IIIcomplexes show

notable frequency dependence. The behavior of the frequency

dependence in ac measurements may be related to the result

of the magnetic interaction J_NiTb> J_NiDyand J_CoTb> J_CoDyfrom

the temperature and ﬁeld dependent magnetic measurements.

The alternating current magnetic measurement revealed that

the Ni^II−Tb^IIIand Co^II−Tb^IIIcomplexes could be attractive

candidates for SMM behavior. From the Arrhenius analysis the

energy barrier for the spin ﬂipping was characterized to

be 14.9(6) and 17.0(4) K for Ni^II−Tb^IIIand Co^II−Tb^III,

respectively, under a dc bias ﬁeld of 1000 Oe. Such a behavior

would clearly originate from the anisotropic properties of both

Ni^IIor Co^IIand Tb^IIIions and their ferromagnetic exchange

coupling. This observation suggests a possible way to pursue

new 3d−4f SMM complexes by increasing the spin state and

magnetic anisotropy through the use of anisotropic d metal

ions, like Ni^IIand Co^II. The present complexes do show ac

signals at low temperature and would give eﬀective information

for the design of d−f SMMs.

Combining the results on the Ni^II−Ln^IIIand Co^II−Ln^III

complexes, we can point out that a Tb^IIIion is a more promising

component candidate for SMMs than a Dy^IIIion (Figures 11 and

12). On the other hand, the results on Cu^II−Ln^IIIcomplexes

indicate that the Dy^IIIanalogue has better characteristics than the

Tb^IIIanalogue (Figure 10). The measurement conditions

including the magnitude of a bias ﬁeld were not optimized, but

a similar trend could be found from a closer look at the results

without any dc ﬁeld (Figure S3b), and the present diﬀerence is

substantial. The Ni^II−Ln^IIIand Co^II−Ln^IIIcomplexes are

isomorphous to each other but diﬀerent from those of the

Cu^II−Ln^IIIcomplexes, as the crystal structure analysis revealed.

A plausible reason for this diﬀerence will be attributed to the

structures around the Ln ions. Namely, the single-ion magnetic

anisotropy and the exchange coupling are very sensitive to the

Ln coordination structure, which regulates the SMM character-

istics.

CONCLUSIONS

■

We have reported the series of acetato- and diphenolato-bridged

3d−4f heterometal binuclear complexes, [M^II(3-MeOsaltn)-

(MeOH)_x(ac) Ln^III(hfac)₂] (M^II= Co^II, Ni^II, Cu^II, Zn^II; Ln^III

=

Gd^III, Tb^III, Dy^III, and La^III) (x = 0 for M^II= Cu^II, Zn^II; x = 1 for

M^II= Co^II, Ni^II), where 3-MeOsaltn and hfac denote N,N′-bis(3-

methoxysalicyldene)-1,3-propanediaminato) and hexaﬂuoroace-

tylacetonato, respectively. Magnetic interactions in these

complexes were determined by removing the ﬁrst-order

orbital contributions of the M^II(M^II= Co^IIor Ni^II) and Ln^III

(Ln^III= Tb^IIIor Dy^III) ions using an empirical approach.^27,28

P

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

[Zn(3-MeOsaltn)(ac)Tb(hfac)₂]. The complex was prepared by a

similar method for [Zn(3-MeOsaltn)(ac)Gd(hfac)₂]. Yellow plate crys-

tals. Yield: 0.129 g (62%). Anal. Calcd for C₃₁H₂₅N₂O₁₀F₁₂ZnTb =

[Zn(3-MeOsaltn)(ac)Tb(hfac)₂]: C; 35.88, H; 2.43, N; 2.70%. Found:

C; 36.05, H; 2.54, N; 2.84%. IR (KBr disk): ν(CO) 1658; ν(CN)

1625;ν_asm(CO) 1554; ν_sym(CO) 1442; ν(C−F) 1257, 1218, 1147 cm⁻¹.

[Zn(3-MeOsaltn)(ac)Dy(hfac)₂]. The complex was prepared by a

similar method for [Zn(3-MeOsaltn)(ac)Gd(hfac)₂]. Yellow plate crys-

tals. Yield: 0.135 g (65%). Anal. Calcd for C₃₁H₂₅N₂O₁₀F₁₂ZnDy =

[Zn(3-MeOsaltn)(ac)Dy(hfac)₂]: C; 35.75, H; 2.42, N; 2.69%. Found:

C; 35.85, H; 2.78, N; 2.71%. IR (KBr disk): ν(CO) 1658; ν(CN)

1639;ν_asm(CO) 1562; ν_sym(CO) 1440; ν(C−F) 1257, 1218, 1147 cm⁻¹.

[Ni(3-MeOsaltn)(MeOH)(ac)Gd(hfac)₂]. A mixed solution of ace-

tone and methanol (1:1 by volume) of [Gd(ac)(Hhfac)(hfac)₂(H₂O)₂]

(0.175 g, 0.2 mmol) was added to a mixed solution of acetone and

methanol (1:1 by volume) of [Ni(3-MeOsaltn)] (0.080 g, 0.2 mmol).

The solution was left at room temperature for several days to precipitate

blue plate crystals. They were collected by ﬁltration and dried in vacuo.

The crystal solvent methanol was lost on drying. Yield: 0.117 g (54%).

Anal. Calcd for C₃₂H₃₁N₂O₁₂F₁₂NiGd = [Ni(3-MeOsaltn)(MeOH)-

(ac)Gd(hfac)₂]·H₂O: C; 35.60, H; 2.89, N; 2.59%. Found: C;

35.69, H; 2.94, N; 2.76%. IR (KBr disk): ν(CO) 1654; ν(CN)

1633;ν_asm(CO) 1560; ν_sym(CO) 1440; ν(C−F) 1257, 1220, 1151 cm⁻¹.

[Ni(3-MeOsaltn)(MeOH)(ac)Tb(hfac)₂]. The complex was pre-

pared by a almost similar method for [Ni(3-MeOsaltn)(MeOH)(ac)-

Gd(hfac)₂]. Blue plate crystals. Yield: 0.123 g (57%). Anal. Calcd for

C₃₂H₃₁N₂O₁₂F₁₂NiTb = [Ni(3-MeOsaltn)(MeOH)(ac)Tb-

(hfac)₂]·H₂O: C; 35.55, H; 2.89, N; 2.59%. Found: C; 35.61, H; 2.70,

N; 2.85%. IR (KBr disk): ν(CO) 1654; ν(CN) 1633;ν_asm(CO)

1560; ν_sym(CO) 1440; ν(C−F) 1257, 1220, 1149 cm⁻¹.

EXPERIMENTAL SECTION

■

Materials. All reagents and solvents, obtained from Tokyo Kasei Co.

and Wako Pure Chemical Industries, in the syntheses were of reagent

grade, and they were used without further puriﬁcation. The tetradentae

Schiﬀ-base ligand N,N′-bis(3-methoxy-2-oxybenzylidene)-1,3-propane-

diamine), abbreviated as H₂3-MeOsaltn, was obtained as yellow crystals

by mixing 3-methoxysalicylaldehyde and 1,3-diaminopropane in a

1

2:1 mol ratio in methanol. The ligand was identiﬁed by the H NMR

spectrum and mp = 93 °C. The d-component complexes [M(3-

MeOsaltn)(H₂O)_x] (M = Co^II, Ni^II, Cu^II, and Zn^II) were synthesized by

mixing metal(II) acetate hydrate and H₂(3-MeOsaltn) in a 1:1 mol ratio

in methanol, according to the literature applied for the M(salen)

complexes,³⁷in which the synthesis of [Co(3-MeOsaltn)(H₂O)₂] was

carried out under a nitrogen atmosphere in order to avoid the air

oxidation. Recrystallization was performed from chloroform.

[Gd(ac)(Hhfac)(hfac)₂(H₂O)₂]. To a solution of gadolinium acetate

tetrahydrate Gd^III(ac)₃·4H₂O (0.406 g, 1 mmol) in 50 mL of water was

added Hhfac (hexaﬂuoroacetylacetone) (0.618 g, 3 mmol) in small

amount of methanol under stirring at room temperature in a hood. The

mixture was stirred for 1 h, and the white precipitate was collected by

ﬁltration. The ﬁltrate was recrystallized from diethyl ether. Colorless

needle crystals. Anal. Calcd for C₁₇H₁₂O₁₀F₁₈Gd = [Gd(ac)(Hhfac)-

(hfac)₂(H₂O)₂]: C; 23.32, H; 1.38%. Found: C; 23.35, H; 1.27%. IR

(KBr disk):1650, 1560, ν(C−F) 1255, 1209, 1143 cm⁻¹.

[Tb(ac)(Hhfac)(hfac)₂(H₂O)(MeOH)]. The complex was prepared

by a similar method for [Gd^III(ac)(Hhfac)(hfac)₂(H₂O)₂]. Colorless

needle crystals. Anal. Calcd for C₁₇H₇O₈F₁₈Tb·H₂O·MeOH = [Tb(ac)-

(Hhfac)(hfac)₂(H₂O)(MeOH)]: C; 24.28, H; 1.47%. Found: C; 24.38, H;

1.45%. IR (KBr disk):1650, 1564, ν(C−F) 1255, 1209, 1143 cm⁻¹.

[Dy(ac)(Hhfac)(hfac)₂(H₂O)(MeOH)]. The complex was prepared

by a similar method for [Gd^III(ac)(Hhfac)(hfac)₂(H₂O)₂]. Colorless

needle crystals. Anal. Calcd for C₁₇H₇O₈F₁₈Dy·H₂O·MeOH = [Dy(ac)-

(Hhfac)(hfac)₂(H₂O)(MeOH)]: C; 24.19, H; 1.47%. Found: C; 24.53, H;

1.34%. IR (KBr disk): 1650, 1560, ν(C−F) 1255, 1209, 1141 cm⁻¹.

[La(ac)(Hhfac)(hfac)₂(H₂O)₂]. Colorless needle crystals. Anal. Calcd

for C₁₇H₁₂O₁₀F₁₈La = [La(ac)(Hhfac)(hfac)₂(H₂O)₂]: C; 23.85, H; 1.30%.

Found: C; 23.88, H; 1.45%.

[Ni(3-MeOsaltn)(MeOH)(ac)Dy(hfac)₂]. The complex was pre-

pared by a almost similar method for [Ni(3-MeOsaltn)(MeOH)(ac)-

Gd(hfac)₂]. Blue plate crystals. Yield: 0.121 g (56%). Anal. Calcd for

C₃₂H₃₃N₂O₁₃F₁₂NiDy = [Ni(3-MeOsaltn)(MeOH)(ac)Dy-

(hfac)₂]·2H₂O: C; 34.85, H; 3.02, N; 2.54%. Found: C; 34.89, H; 2.83,

N; 2.78%. IR (KBr disk): ν(CO) 1654; ν(CN) 1633;ν_asm(CO) 1560;

ν_sym(CO) 1440; ν(C−F) 1257, 1220, 1151 cm⁻¹.

[Ni(3-MeOsaltn)(MeOH)(ac)La(hfac)₂]. The complex was pre-

pared by a almost similar method for [Ni(3-MeOsaltn)(MeOH)(ac)-

Gd(hfac)₂]. Blue plate crystals. Yield: 0.115 g (54%). Anal. Calcd for

C₃₂H₃₁N₂O₁₂F₁₂NiLa = [Ni(3-MeOsaltn)(MeOH)(ac)La-

(hfac)₂]·H₂O: C; 36.22, H; 2.94, N; 2.64%. Found: C; 35.88, H; 3.00,

N; 2.67%.

[Co(3-MeOsaltn)(MeOH)(ac)Gd(hfac)₂]. A mixed solution of ace-

tone and methanol (1:1 by volume) of [Gd(ac)(Hhfac)(hfac)₂(H₂O)₂]

(0.175 g, 0.2 mmol) was added to a mixed solution of acetone and

methanol (1:1 by volume) of [Co^II(3-MeOsaltn)] (0.080 g, 0.2 mmol).

The resulting solution was ﬁltered, and the ﬁltrate was left at room

temperature for several days to precipitate orange plate crystals. They

were collected by ﬁltration and dried in vacuo. The crystal solvent

methanol was lost on drying. Yield: 0.108 g (50%). Anal. Calcd for

C₃₂H₃₁N₂O₁₂F₁₂CoGd = [Co(3-MeOsaltn)(MeOH)(ac)Gd-

(hfac)₂]·H₂O: C; 35.60, H; 2.89, N; 2.59%. Found: C; 35.54, H; 2.91,

N; 2.82%. IR (KBr disk): ν(CO) 1656; ν(CN) 1629; ν_asm(CO)

1556; ν_sym(CO) 1440; ν(C−F) 1257, 1218, 1147 cm⁻¹.

[Co(3-MeOsaltn)(MeOH)(ac)Tb(hfac)₂]. The complex was pre-

pared by a similar method for [Co(3-MeOsaltn)(MeOH)(ac)Gd-

(hfac)₂]. Orange plate crystals. Yield: 0.110 g (51%). Anal. Calcd

for C₃₂H₃₁N₂O₁₂F₁₂CoTb = [Co^II(3-MeOsaltn)(MeOH)(ac)-

Tb^III(hfac)₂]·H₂O: C; 35.54, H; 2.89, N; 2.59%. Found: C; 35.92,

H; 2.71, N; 2.85%. IR (KBr disk): ν(CO) 1656; ν(CN)

1629;ν_asm(CO) 1556; ν_sym(CO) 1440; ν(C−F) 1255, 1211, 1147 cm⁻¹.

[Co(3-MeOsaltn)(MeOH)(ac)Dy(hfac)₂]. The complex was pre-

pared by a similar method for [Co(3-MeOsaltn)(MeOH)(ac)Gd(hfac)₂].

Orange plate crystals. Yield: 0.111 g (51%). Anal. Calcd for

C₃₂H₃₁N₂O₁₂F₁₂CoDy = [Co(3-MeOsaltn)(MeOH)(ac)Dy-

(hfac)₂]·H₂O: C; 35.42, H; 2.88, N; 2.58%. Found: C; 35.79, H; 2.88,

N; 2.72%. IR (KBr disk): ν(CO) 1656; ν(CN) 1629; ν_asm(CO)

1556; ν_sym(CO) 1440; ν(C−F) 1255, 1216, 1147 cm⁻¹.

[Cu(3-MeOsaltn)(ac)Gd(hfac)₂]. To a solution of [Gd(ac)(Hhfac)-

(hfac)₂(H₂O)₂] (0.175 g, 0.2 mmol) in 20 mL of acetone was added a

solution of [Cu(3-MeOsaltn)] (0.081 g, 0.2 mmol) in 20 mL of acetone

at room temperature. The mixture was ﬁltered, and the ﬁltrate, clear

green solution, was left at room temperature for several days to

precipitate green plate crystals. They were collected by ﬁltration and

dried in vacuo. Yield: 0.139 g (67%). [Cu(3-MeOsaltn)(ac)Gd(hfac)₂],

Anal. Calcd for C₃₁H₂₅N₂O₁₀F₁₂CuGd: C; 36.00, H; 2.44, N; 2.71%.

Found: C; 36.00, H; 2.67, N; 2.84%. IR (KBr disk): ν(CO) 1658;

ν(CN) 1623; ν_asm(CO) 1554; ν_sym(CO) 1442; ν(C−F) 1255, 1199,

1147 cm⁻¹.

[Cu(3-MeOsaltn)(ac)Tb(hfac)₂]. The complex was prepared by a

similar method for [Cu(3-MeOsaltn)(ac)Gd(hfac)₂]. Green plate crys-

tals. Yield: 0.131 g (63%). Anal. Calcd for C₃₁H₂₅N₂O₁₀F₁₂CuTb =

[Cu(3-MeOsaltn)(ac)Tb(hfac)₂]: C; 35.94, H; 2.43, N; 2.70%. Found:

C; 35.83, H; 2.56, N; 2.88%. IR (KBr disk): ν(CO) 1658; ν(CN)

1623;ν_asm(CO) 1554; ν_sym(CO) 1442; ν(C−F) 1255, 1216, 1147 cm⁻¹.

[Cu(3-MeOsaltn)(ac)Dy(hfac)₂]. The complex was prepared by a

similar method for [Cu(3-MeOsaltn)(ac)Gd(hfac)₂]. Green plate

crystals. Yield: 0.141g (68%). Anal. Calcd for C₃₁H₂₅N₂O₁₀F₁₂CuDy =

[Cu(3-MeOsaltn)(ac)Dy(hfac)₂]: C; 35.82, H; 2.42, N; 2.69%. Found:

C; 35.96, H; 2.52, N; 2.71%. IR (KBr disk): ν(CO) 1658; ν(CN)

1623;ν_asm(CO) 1554; ν_sym(CO) 1442; ν(C−F) 1255, 1216, 1147 cm⁻¹.

[Zn(3-MeOsaltn)(ac)Gd(hfac)₂]. A acetone solution of [Gd(ac)-

(Hhfac)(hfac)₂(H₂O)₂] (0.175 g, 0.2 mmol) was added to a acetone

solution of [Zn(3-MeOsaltn)] (0.081 g, 0.2 mmol). The solution was

left at room temperature for several days to precipitate yellow plate crys-

tals. They were collected by ﬁltration and dried in vacuo. Yield: 0.120 g

(58%). Anal. Calcd for C₃₁H₂₅N₂O₁₀F₁₂ZnGd = [Zn(3-MeOsaltn)-

(ac)Gd(hfac)₂]: C; 35.93, H; 2.43, N; 2.70%. Found: C; 36.25, H; 2.56,

N; 2.85%. IR (KBr disk): ν(CO) 1656; ν(CN) 1625;ν_asm(CO)

1554; ν_sym(CO) 1442; ν(C−F) 1255, 1218, 1147 cm⁻¹.

Q

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

[Co(3-MeOsaltn)(MeOH)(ac)La(hfac)₂]. The complex was pre-

pared by a similar method for [Co(3-MeOsaltn)(MeOH)(ac)Gd-

(hfac)₂]. Orange plate crystals. Yield: 0.101 g (48%). Anal. Calcd for

C₃₂H₃₁N₂O₁₂F₁₂CoLa = [Co(3-MeOsaltn)(MeOH)(ac)La-

(hfac)₂]·H₂O: C; 36.21, H; 2.94, N; 2.64%. Found: C; 36.08, H; 2.93,

N; 2.77%.

Hendrickson, D. N. J. Am. Chem. Soc. 1993, 115, 1804−1816.

(d) Thomas, L.; Lionti, F.; Ballou, R.; Gatteschi, D.; Sessoli, R.;

Barbara, B. Nature 1996, 383, 145−147. (e) Cadiou, C.; Murrie, M.;

Paulsen, C.; Villar, V.; Wernsdorfer, W.; Winpenny, W. E. P. Chem.

Commun. 2001, 2666−2667. (f) Shores, M. P.; Sokol, J. J.; Long, J. R. J.

Am. Chem. Soc. 2002, 124, 2279−2292. (g) Boskovic, C.; Brechin, E. K.;

Streib, W. E.; Folting, K.; Bollinger, J. C.; Hendrickson, D. N.; Christou,

G. J. Am. Chem. Soc. 2002, 124, 3725−3736. (h) Ritter, S. Chem. Eng.

News 2004, 82, 29−32. (i) Gatteschi, D.; Sessoli, R. Angew. Chem., Int.

Ed. 2003, 42, 268−297. (j) Wernsdorfer, W.; Sessoli, R. Science 1999,

284, 133−135. (k) Aubin, S. M. J.; Sun, Z.; Pardi, L.; Krzystek, J.;

Folting, K.; Brunel, L. C.; Rheingold, A. L.; Christou, G.; Hendrickson,

D. N. Inorg. Chem. 1999, 38, 5329−5340.

(2) (a) Ishikawa, N.; Sugita, M.; Ishikawa, T.; Koshihara, S.; Kaizu, Y. J.

Am. Chem. Soc. 2003, 125, 8694−8695. (b) Ishikawa, N.; Sugita, M.;

Wernsdorfer, W. J. Am. Chem. Soc. 2005, 127, 3650−3651.

(3) (a) Bencini, C.; Benelli, A.; Caneschi, A.; Carlin, R. L.; Dei, A.;

Gatteschi, D. J. Am. Chem. Soc. 1985, 107, 8128−8136. (b) Benelli, A.;

Bencini, C.; Caneschi, A.; Dei, A.; Gatteschi, D. Inorg. Chem. 1986, 25,

572−575. (c) Benelli, A.; Caneschi, A.; Gatteschi, D.; Guillou, O.; Pardi,

L. Inorg. Chem. 1990, 29, 1750−1755. (d) Matsumoto, N.; Sakamoto,

M.; Tamaki, H.; Okawa, H.; Kida, S. Chem. Lett. 1990, 853−854.

(4) (a) Osa, S.; Kido, T.; Matsumoto, N.; Re, N.; Pochaba, A.;

Mronzinski, J. J. Am. Chem. Soc. 2004, 126, 420−423. (b) Kido, T.;

Nagasato, S.; Sunatsuki, Y.; Matsumoto, N. Chem. Commun. 2000,

2113−2114.

Physical Measurements. Elemental analyses (C, H, N) were

carried out at the Center for Instrumental Analysis of Kumamoto

University. Infrared spectra were recorded at room temperature using a

JEOL JIR-6500W spectrometer with samples in KBr disks. Thermog-

ravimetric analyses were carried out on a TG/DTA6200 (SII Nano

Technology Inc.) instrument at the 10 K min⁻¹heating rate using ca.

2 mg sample. Temperature-dependent magnetic susceptibilities in the

temperature range 1.9−300 K under an external magnetic ﬁeld of 0.1 T

and ﬁeld-dependent magnetization measurements in an applied mag-

netic ﬁeld from 0 to 5 T at 1.9 K were measured with an MPMS XL5

SQUID susceptometer (Quantum Design, Inc.). Some microcrystalline

samples consisting of Tb^IIIand Dy^IIIions showed apparent reorientation

in the applied magnetic ﬁeld of 0.5 T. All samples dispersed in liquid

paraﬃn to avoid orientation in the ﬁeld, and the results are given in the

text. The calibrations were performed with palladium. Alternating cur-

rent magnetic measurements were carried out at University of Wroclaw

in a 3.0 G ac ﬁeld oscillating over the range 10−1000 Hz and on a PPMS

equipped with an ac/dc magnetic probe (Quantum Design, Inc.) at The

University of Electro-Communications. Corrections for diamagnetism

were applied using Pascal’s constants.

X-ray Crystallography. X-ray data were collected on a Rigaku

Rapid imaging plate diﬀractometer with graphite monochromated Mo

Kα radiation (λ = 0.710 69 Å) at 150 or 100 K. All the crystals have a

tendency of eﬄorescense, to decompose gradually. Each crystal was

coated by epoxy resin quickly and mounted on a glass rod and measured

at 150 or 100 K. An empirical absorption correction was applied. The

data were also corrected for Lorentz and polarization eﬀects. The crystal

structures were determined and reﬁned by direct method using

the CrystalStructure crystallographic software package.³⁸The X-ray

diﬀraction analyses have problems in the explanation of the disorder at

some of hfac ligands even at the low temperature. Hydrogen atoms were

reﬁned using the riding model. CCDC 829232−829243 contain the

supplementary crystallographic data for 12 complexes. These data can

be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/

retrieving.html, or from the Cambridge Crystallographic Data Centre,

12 Union Road, Cambridge CB2 1EZ, U.K.; fax (+44) 1223−336−033;

or e-mail deposit@ccdc.cam.ac.uk.

(5) Mishra, A.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. J. Am.

Chem. Soc. 2004, 126, 15648−15649.

(6) Zaleski, C. M.; Depperman, E. C.; Kampf, J. W.; Kirk, M. L.;

Pecoraro, V. Angew. Chem., Int. Ed. 2004, 43, 3912−3914.

(7) (a) Costes, J. P.; Dahan, F.; Wernsdorfer, W. Inorg. Chem. 2006, 45,

5−7. (b) Hamamatsu, T.; Yabe, K.; Towatari, M.; Matsumoto, N.; Re,

N.; Pochaba, A.; Mrozinski, J. Bull. Chem. Soc. Jpn. 2007, 80, 523−529.

(c) Hamamatsu, T.; Yabe, K.; Towatari, M.; Osa, S.; Matsumoto, N.; Re,

N.; Pochaba, A.; Mrozinski, J.; Gallani, J. L.; Barla, A.; Imperia, P.;

Paulsen, C.; Kappler, J. P. Inorg. Chem. 2007, 46, 4458−4468.

(8) Sun, Y. Q.; Liang, M.; Dong, W.; Yang, G. M.; Liao, D. Z.; Jiang, Z.

H.; Yan, S. P.; Cheng, P. Eur. J. Inorg. Chem. 2004, 7, 1514−1521.

(9) Aronica, C.; Pilet, G.; Chastanet, G.; Wernsdorfer, W.; Jacquot, J.

F.; Luneau, D. Angew. Chem., Int. Ed. 2006, 45, 4659−4662.

(10) Ferbinteanu, M.; Kajiwara, T.; Choi, K. Y.; Nojiri, H.; Nakamoto,

A.; Kojima, N.; Cimpoesu, F.; Fujimura, Y.; Takaishi, S.; Yamashita, M. J.

Am. Chem. Soc. 2006, 128, 9008−9009.

(11) Yamaguchi, T.; Sunatsuki, Y.; Ishida, H.; Kojima, M.; Akashi, H.;

Re, N.; Matsumoto, N.; Pochaba, A.; Mrozinski, J. Inorg. Chem. 2008, 47,

5736−5745.

(12) Liu, J.-L.; Guo, F.-S.; Meng, Z.-S.; Zheng, Y.-Z.; Leng, J.-D.; Tong,

M.-L.; Ungur, L.; Chibotaru, L. F.; Heroux, K. J.; Hendrickson, D. N.

Chem. Sci. 2011, 2, 1268−1272.

ASSOCIATED CONTENT

* Supporting Information

■

S

Additional ﬁgures and crystallographic data in CIF format. This

material is available free of charge via the Internet at http://pubs.

acs.org.

(13) (a) Kajiwara, T.; Nakano, M.; Takaishi, S.; Yamashita, M. Inorg.

Chem. 2008, 47, 8604−8606. (b) Kajiwara, T.; Nakano, M.; Takahashi,

K.; Takaishi, S.; Yamashita, M. Chem.Eur. J. 2011, 17, 196−205.

(14) (a) Chandrasekhar, V.; Pandian, B. M.; Azhakar, R.; Vittal, J. J.;

AUTHOR INFORMATION

Corresponding Author

*E-mail: naohide@aster.sci.kumamoto-u.ac.jp. Fax: +81-96-342-

3390.

■

́

Clerac, R. J. Inorg. Chem. 2007, 46, 5140−5142. (b) Yamaguchi, T.;

Costes, J.-P.; Kishima, Y.; Kojima, M.; Sunatsuki, Y.; Bruefuel, N.;

Tuchagues, J.-P.; Vendier, L.; Wernsdorfer, W. Inorg. Chem. 2010, 49,

9125−9135. (c) Costes, J.-P.; Vendier, L.; Wernsdorfer, W. Dalton

Trans. 2011, 40, 1700−1706.

Notes

The authors declare no competing ﬁnancial interest.

(15) (a) Sutter, J.-P.; Dhers, S.; Rajamani, R.; Ramasesha, S.; Costes, J.-

P.; Duhayon, C.; Vendier, L. Inorg. Chem. 2009, 48, 5820−5828.

(b) Long, J.; Chamoreau, L.-M.; Marvaud, V. Dalton Trans. 2010, 39,

2188−2190. (c) Mori, F.; Nyui, T.; Ishida, T.; Nogami, T.; Choi, K.-Y.;

Nojiri, H. J. Am. Chem. Soc. 2006, 128, 1440−1441. (d) Ueki, S.; Ishida,

T.; Nogami, T.; Choi, K.-Y.; Nojiri, H. Chem. Phys. Lett. 2007, 440, 263−

267.

(16) (a) Towatari, M.; Hamamatsu, T.; Yabe, K.; Matsumoto, N.;

Mrozinski, J. Proceedings of XVth Winter School on Coordination

Chemistry, Karpacz, Poland, Dec. 4−8, 2006. They reported complexes

[Cu(3-MeOsaltn)(ac)Ln(hfac)₂] (Ln =Gd^III, Dy^III, Tb^III). (b)Towatari,

ACKNOWLEDGMENTS

■

K.N. and T.F. were supported by the Research Fellowship of

the Japan Society for the Promotion of Science, KAKENHI

00248556 and 00248498.

REFERENCES

■

(1) (a) Sessoli, R.; Gatteschi, D.; Caneschi, A.; Novak, M. A. Nature

1993, 365, 141−143. (b) Gatteschi, D.; Caneschi, A.; Pardi, L.; Sessoli,

R. Science 1994, 265, 1054−1058. (c) Sessoli, R.; Tsai, H. L.; Schake, A.

R.; Wang, S.; Vincent, J. B.; Folting, K.; Gatteschi, D.; Christou, G.;

R

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

M.; Matsumoto, N.; Kojima, M.; Re, N.; Mrozinski, J. Proceedings of

VIX th Annual Symmposium on Coordination Chemistry of Japan, 2PB-

013 2009. Nagasaki, Japan. They reported the systematic syntheses and

magnetic properties of binuclear M(II)−Ln(III) complexes.

(17) Ramade, I.; Kahn, O.; Jeannin, Y.; Robert, F. Inorg. Chem. 1997,

36, 930−936.

(38) Crystalstructure 3.7.0, Crystal Structure Analysis Package; Rigaku

and Rigaku/MSC: The Woodlands, TX, 2000−2005.

(18) (a) Fukuhara, C.; Tsuneyoshi, K.; Katsura, K.; Matsumoto, N.;

Kida, S.; Mori, M. Bull. Chem. Soc. Jpn. 1989, 62, 3939−3943.

(b) Fukuhara, C.; Tsuneyoshi, K.; Matsumoto, N.; Kida, S.; Mikuriya,

M.; Mori, M. J. Chem. Soc., Dalton Trans. 1990, 3473−3479.

(19) (a) Sakamoto, M.; Hashimura, M.; Matsuki, K.; Matsumoto, N.;

Inoue, K.; Okawa, H. Bull. Chem. Soc. Jpn. 1991, 64, 3639−3641.

(b) Costes, J. P.; Dahan, F.; Dupuis, A.; Laurent, J.-P. Inorg. Chem. 1996,

35, 2400−2402. (c) Miyasaka, H.; Matsumoto, N.; Okawa, H.; Re, N.;

Gallo, E.; Floriani, C. Angew. Chem., Int. Ed. Engl. 1995, 34, 1446−1448.

(20) (a) Wang et al. reported Ni^II−Gd^IIIcomplex of [Ni(3-MeOsaltn)

(MeOH)₂Ln(NO₃)₃]. Wang, J.-H.; Yan, P. F.; Li, G. M.; Zhang, J. W.;

Chen, P.; Suda, M.; Einaga, Y. Inorg. Chim. Acta 2010, 363, 3706−3713.

(b) Andruh et al. reported a series of [Ni(3-MeOsaltn)Ln(NO₃)₃] and

[Zn(3-MeOsaltn)Ln(NO₃)₃]. Pasatoiu, T. D.; Tiseanu, C.; Madalan, A.

M.; Jurca, B.; Duhayon, C.; Sutter, J. P.; Andruh, M. Inorg. Chem. 2011,

50, 5890−5898. (c) Pasatoiu, T. D.; Tiseanu, C.; Madalan, A. M.; Jurca,

B.; Duhayon, C.; Sutter, J. P.; Andruh, M. Inorg. Chem. 2011, 50, 5879−

5889.

(21) Casey, A. T.; Mitra, S. In Theory and Applications of Molecular

Paramagnetism; Boudreaux, E.A., Mulay, L.N., Eds.; Interscience: New

York, 1976; pp 135 and 271.

(22) Costes, J. P.; Dahan, F.; Dupuis, A.; Laurent, J. P. Inorg. Chem.

1997, 36, 4284−4286.

(23) Chen, Q. Y.; Luo, Q.; Zheng, L. M.; Wang, Z. L.; Chen, J. T. Inorg.

Chem. 2002, 41, 605−609.

(24) Kahn, O. Molecular Magnetism; VCH: Weinhein, 1993; Part 6.4

and references therein.

(25) (a) Hudak, J.; Boca, R.; Dlhan, L.; Kozisek, J.; Moncol, J.

Polyhedron 2011, 30, 1367−1373. (b) Titis, J.; Boca, R. Inorg. Chem.

2011, 50, 11838−11845.

(26) Costes, J. P.; Dahan, F.; Garcìa-Tojal, J. Chem.Eur. J. 2002, 8,

5430−5434.

(27) Kahn, M. L.; Mathoniere, C.; Kahn, O. Inorg. Chem. 1999, 38,

3692−3697.

(28) (a) Costes, J. P.; Dahan, F.; Dupuis, A.; Laurent, J.-P. Chem.Eur.

J. 1998, 4, 1616−1620. (b) Costes, J. P.; Donnadieu, B.; Gheorghe, R.;

Novitchi, G.; Tuchagues, J. P.; Vendier, L. Eur. J. Inorg. Chem. 2008,

5235−5244. (c) Kido, T.; Ikuta, Y.; Sunatsuki, Y.; Ogawa, Y.;

Matsumoto, N.; Re, N. Inorg. Chem. 2003, 42, 398−408.

(29) Daiguebonne, C.; Guillou, O.; Kahn, M.; Kahn, O.; Oushoorn, R.

L.; Boubekeur, K. Inorg. Chem. 2001, 40, 176−178.

(30) Shimada, T.; Okazawa, A.; Kojima, N.; Yoshii, S.; Nojiri, H.;

Ishida, T. Inorg. Chem. 2011, 50, 10555−10557. (b) Ishida, T.;

Watanabe, R.; Fujiwara, K.; Okazawa, A.; Kojima, N.; Tanaka, G.; Yoshii,

S.; Noriji, H. Dalton Trans. 2012, 41, 13609−13619.

(31) Shiga, T.; Ohba, M.; Okawa, H. Inorg. Chem. 2004, 43, 4435−

4446.

(32) (a) Castro, S. L.; Sun, Z.; Grant, C. M.; Bollinger, J. C.;

Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 1998, 120, 2365−

2375. (b) Ako, A.; Mereacre, M.; Hewitt, I. J.; Clerac, R.; Lecren, L.;

Anson, C. E.; Powell, A. K. J. Mater. Chem. 2006, 16, 2579−2586.

(c) Okazawa, A.; Nojiri, H.; Ishida, T.; Kojima, N. Polyhedron 2011, 30,

3140−3144.

(33) Cole, K. S.; Cole, H. R. J. Chem. Phys. 1941, 9, 341−351.

(34) Sessoli, R.; Powell, A. K. Coord. Chem. Rev. 2009, 253, 2328−

2341.

(35) (a) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular Nanomagnets;

Oxford University Press, New York, 2006. (b) Gatteschi, D.; Sessoli, R.

Angew. Chem., Int. Ed. 2003, 42, 268−297.

(36) Colacio, E.; Ruiz-Sanchez, J.; White, F. J.; Brechin, E. K. Inorg.

Chem. 2011, 50, 7268−7273.

(37) Gruber, S. J.; Harris, C. M.; Sinn, E. J. Inorg. Nucl. Chem. 1968, 30,

1805−1830.

S

dx.doi.org/10.1021/ic400594u | Inorg. Chem. XXXX, XXX, XXX−XXX

Article Doi

Syntheses, structures, and magnetic properties of acetato- and diphenolato-bridged 3d-4f binuclear complexes [M(3-MeOsaltn)(MeOH) x(ac)Ln(hfac)2](M = ZnII, CuII,NiII,CoII;Ln = LaIII,GdIII, TbIII, DyIII; 3-MeOsaltn = N,N-Bis(3-methoxy-2-oxybenzylidene)-1,3

DOI: 10.1021/ic400594u

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ic400594u

Full text of DOI:10.1021/ic400594u

Products guided by the article

R&D Labs maybe for 5767-53-3

Relevant to this article

Hot Product