Copper(II) cubane complexes built from electro- and photosensitive β-aminovinyl trifluoromethyl ketone ligands

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

FULL PAPER

DOI: 10.1002/ejic.201200737

Copper(II) Cubane Complexes Built from Electro- and Photosensitive

β-Aminovinyl Trifluoromethyl Ketone Ligands

Nicolas Chopin,*^[a]Guillaume Chastanet,^[b]Boris Le Guennic,^[c]Maurice Médebielle,*^[a]and

Guillaume Pilet*^[d]

Keywords: Copper / Cubane complexes / Magnetic properties / Density functional calculations

A new family of [Cu₄L₄] cubane complexes have been syn-

thesized by the self-assembly of Cu^IIsalts and β-aminovinyl

trifluoromethyl ketone ligands. The resulting polynuclear

architectures were especially designed to exhibit magnetic

properties arising from the metallic cubane architecture as

well as electro- or photosensitive properties arising from the

ligands. The optical and electrochemical behaviours of the

compounds were measured in solution to confirm that the

properties of the free ligands were retained within the elabo-

rated complexes. DFT investigations of the organic moieties

were conducted with the aim of assigning the observed elec-

tronic processes to the different redox steps and absorption

bands. The crystal structures of the polynuclear complexes

were solved and refined by single-crystal X-ray diffraction

analysis and correlated to variable-temperature magnetic

susceptibility data. The overall expected ferromagnetic be-

haviour of the [Cu₄L₄] complexes is weakly influenced by

the ligand substitutions. Ab initio configuration interaction

calculations helped us to detail the variations of the various

magnetic exchange pathways.

ligands derived from the condensation of amines with β-

diketones have been used extensively. Our research is cur-

rently devoted to the elaboration of ligands able to bring

additional properties to the magnetic core with the expecta-

tion that both properties could interact cooperatively. To do

this we have focused our attention on β-aminovinyl trifluor-

omethyl ketones, which are frequently exploited as valuable

starting compounds in the preparation of various trifluoro-

methylated aromatic and heterocycle molecules.^[8]On the

other hand, their ability to coordinate a large panel of tran-

sition-metal ions have led to the preparation of a number

of metallomesogens with interesting liquid-crystal proper-

ties^[9–12]or highly-active catalysts for olefin polymerization,

including fluorinated enaminones.^[13–15]β-Aminovinyl

ketones with a fluoroalkyl moiety are thus considered to be

interesting building blocks that can be decorated in a

number of synthetic sequences to produce diverse molecu-

lar entities for specifically designed applications.^[8,16–26]

Their potential for diverse coordination chemistry^[27–36]

combined with appropriate pendant organic functionalities

(a photosensitive unit, a stable radical, an electron donor, a

luminescent unit or a biomolecule) make them attractive for

the design of new generations of original materials. More-

over, these enaminone building blocks offer great potential

for rapidly generating molecular diversity (see the general

structure in Scheme 1) because these organic substrates can

be structurally modified at different sites (R¹, R², R³and

R⁴) by variation of 1) the amine (R¹), 2) the substitution on

the C=C double bond (R²and R³) and 3) the fluoroalkyl

moiety (R⁴).^[21–23]Organic functionalities with specific

properties that we are interested in can in principle be at-

Introduction

Polynuclear architectures have attracted a great deal of

attention, especially for their magnetic potentialities

brought on by the metallic ions.^[1–3]Among them, [M₄X₄]

cubane complexes (M = metal ions, X = S, O, for example)

have been widely examined because the particular organiza-

tion of the paramagnetic ions within these architectures fa-

vours ferromagnetic interactions.^[4,5]The resulting high-spin

ground-state can lead to single-molecule magnet (SMM)

behaviour,^[6]or such complexes can be used as building

blocks to elaborate clusters with higher nuclearity.^[7]Among

the numerous ligands employed in the synthesis of cubane

cores, bidentate β-diketone units and tridentate Schiff base

[a] Université Claude Bernard Lyon 1, Institut de Chimie et

Biochimie Moléculaire et Supramoléculaire (ICBMS), UMR

CNRS-UCBL-INSA 5246,

Bât. Curien, 43 boulevard du 11 novembre 1918,

69622 Villeurbanne Cedex, France

E-mail: nicolas.chopin@gmail.com

maurice.medebielle@univ-lyon1.fr

Homepage: http://ww w .icbms.fr/

[b] Institut de Chimie de la Matière Condensée de Bordeaux

(ICMCB), CNRS-UPR 9048,

87 avenue du Docteur Schweitzer, 33608 Pessac Cedex, France

[c] Institut des Sciences Chimiques de Rennes, UMR 6226 CNRS-

Université de Rennes 1,

6226, 35042 Rennes Cedex, France

[d] Université Claude Bernard Lyon 1, Laboratoire des multi

Matériaux et Interfaces (LMI), UMR CNRS-UCBL 5615,

Bât. Chevreul, 43 boulevard du 11 novembre 1918,

69622 Villeurbanne Cedex, France

E-mail: guillaume.pilet@univ-lyon1.fr

Supporting information for this article is available on the

WWW under http://dx.doi.org/10.1002/ejic.201200737.

Eur. J. Inorg. Chem. 0000, 0–0

1

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

tached at four positions of the enaminone using readily theory (DFT) calculations. Single-crystal X-ray diffraction

available starting materials and different synthetic strate- analysis evidenced the elaboration of cubane complexes, the

gies.

magnetic properties of which were investigated in the solid

phase. Ab initio configuration interaction (CI) calculations

allowed us to detail the magnetic exchange pathways pres-

ent in these cubanes.

Results and Discussion

Synthesis of Ligands and Complexes

The β-aminovinyl trifluoromethyl ketones (enaminones)

L2H₂–L4H₂(Scheme 2) were prepared in moderate unopti-

mized yields in three straightforward steps: acetalization of

suitable methyl ketones A–C, trifluoroacetylation of di-

Scheme 1. Fluoroalkylated enaminone: a versatile building block

for the design of task-specific ligands.

methyl acetals D,^{[36] [35b]}and F^[35b]by using a combination

E

In a previous work we used the Cu^IIcubane complex

[Cu₄(L1)₄] (1), built from the 1,1,1-trifluoro-7-hydroxy-4-

methyl-5-azahept-3-en-2-one ligand (L1H₂; R¹= C₂H₄OH,

R²= CH₃, R³= H and R⁴= F, see Scheme 1 and

Scheme 2), to elaborate heterometallic structures and, in

particular, a [Cu₆Ln₃] SMM.^[7a,7b]In this work, we focused

our attention on the design of β-aminovinyl ketones^[34,35]

and analogues of the L1H₂ligand for the synthesis of some

original cubane-based complexes bearing two different

functionalities. This study may also provide a more compre-

hensive understanding of the different parameters involved

in synthesizing such variations of Cu^IIcubane complexes

because it was not obvious at the outset whether more elab-

orated ligands than L1H₂would produce similar architec-

tures. In addition, this study will provide a better overview

of their structural, electrochemical and magnetic properties.

We present herein the synthesis of three original tridentate

(N,O,O) enaminone ligands (Scheme 2) especially designed

with the R²group bearing photo- and electrosensitive enti-

ties, that is, R²= phenyl (L2H₂), phenyldiazenyl (L3H₂) and

anthracenyl (L4H₂). Their coordination chemistry was in-

vestigated to elaborate Cu^IIcubane complexes. The optical

and electrochemical properties of the complexes thus ob-

tained were compared with the properties of the ligand pre-

cursors and interpreted with the help of density functional

of trifluoroacetic anhydride (TFAA, 2 equiv.) and pyridine

(4 equiv.) in anhydrous dichloromethane and O/N substitu-

tion of the resulting methoxyvinyl (trifluoromethyl) ketones

G,^[36]H^[35b]and I^[35b]with ethanolamine in anhydrous ace-

tonitrile (Scheme 3).

Scheme 3. Synthesis of the trifluoromethyl β-aminovinyl ketone li-

gands L2H₂–L4H₂. Reagents and conditions: a) (EtO)₃CH

(1.5 equiv.), p-CH₃C₆H₄SO₃H (0.7 mol%), MeOH, 60 °C, 24 h; D–

FϾ 95%; b) trifluoroacetic anhydride (2.0 equiv.), pyridine

(4.0 equiv.), dichloromethane, 45 °C, 8–72 h; G 91%, H 84%; I

60%; c) ethanolamine (1.0 equiv.), G, H or I (1.0 equiv.), acetoni-

trile, room temp., 18 h; L2H₂98%, L3H₂90%; L4H₂95%.

The methyl ketone bearing the azobenzene unit B is men-

tioned in the literature^[37]but without full spectroscopic

data; it was prepared in 71% isolated yield from p-amino-

acetophenone and nitrosobenzene (both commercially

available) in acetic acid at 90 °C overnight. The other

ketones A and C are commercially available. All the ligands

1

were obtained as Z isomers, as determined by H NMR

Scheme 2. Designed β-aminovinyl trifluoromethyl ketone ligands

L1H₂–L4H₂.

with a deshielded peak of the amino proton at δ_H

Ͼ

2

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

Copper(II) Cubane Complexes

Figure 1. Copper cubane complexes with L2^2–(phenyl, 2), L3^2–(azobenzene, 3) and L4^2–(anthracenyl, 4) and the corresponding cubane

cores with labels. Hydrogen atoms and co-crystallized solvent molecules have been omitted for clarity.

10.0 ppm due to hydrogen bonding between NH and C=O. from one ligand and a third oxygen atom from a second

The three ligands were isolated as solids and were purified ligand whereas the apical position is occupied by an oxygen

by silica gel chromatography.

atom from a third ligand. Therefore the alkoxo group of

The corresponding Cu^IImetal complexes 2, 3 and 4 (Fig- the ligand is involved in a μ₃fashion, giving rise to more or

ure 1) bearing the L2^2–, L3^2–and L4^2–ligands, respectively, less distorted {Cu₄O₄} cubic cores (Figure 1). The average

were prepared according to a previously reported pro- values of the short Cu–O and Cu–N bond lengths in the

cedure^[7b]. In all cases, single-crystals suitable for X-ray dif- O₃N₁square planes are the same for all three complexes

fraction characterization were obtained after slow evapora- and equal to 1.95 Å (Table 1), in good agreement with those

tion of an appropriate solvent (see the Exp. Section).

previously reported for 1.^[7b]As is usually observed, the

oxygen atom coordinated in the apical position presents

longer Cu–O bond lengths that are equal to 2.36, 2.40 and

2.42 Å in 2, 4 and 3, respectively. Thus, the distorted cubane

cores exhibit short (3.09, 3.15 and 3.15 Å for 2, 3 and 4,

respectively) and long (3.31, 3.35 and 3.22 Å for 2, 3 and 4,

respectively) Cu···Cu distances. For all complexes, the short

Cu···Cu distance implies two types of Cu–O–Cu angles: a

small one (from 90.1 to 92.5°, average 91.3°) and a large

one (from 104.0 to 108.0°, average 106.5°). Regarding the

long Cu···Cu distance, only one kind of Cu–O–Cu angle is

present (from 94.2 to 100.3°, average 97.5°). Table 1 details

the important bond lengths, Cu···Cu distances and bond

angles for the three complexes.

Structural Descriptions of the Cubane Complexes 2–4

The three crystal structures 2, 3 and 4 were solved in the

space groups tetragonal I4₁/a, monoclinic C2/c and ortho-

rhombic Fdd2, respectively (see the Exp. Section). The re-

fined formulae for 2, 3 and 4, including the co-crystallized

solvent molecules, are [Cu₄(L2)₄]·2.4(acetone), [Cu₄(L3)₄]

and [Cu₄(L4)₄]·6(toluene). According to these observations,

the asymmetric units are {CuL} for 2 and {Cu₂L₂} for 3

and 4. Compound 2, being the most symmetric, generates

[Cu₄(L)₄] tetranuclear complexes with one copper atom.

Moreover, the three original complexes are all neutral

[Cu₄(L)₄] as the four L^2–ligands (L^2–= L2^2–, L3^2–or L4^2–)

Ruiz and co-workers suggested a classification of cubane

are fully deprotonated and no additional ions are evidenced complexes based on the Cu···Cu distances within the cub-

within the structure.

ane core.^[5,38]Three classes were proposed: 1) when two

The Cu^IIion lies in a pentacoordinate [CuO₄N] square- short and four long Cu···Cu distances are observed within

pyramidal environment in all three compounds. Its basal the cubane structure, the class is called [2+4], 2) the second

plane is formed by one nitrogen and two oxygen atoms class presents four short and two long Cu···Cu distances

Eur. J. Inorg. Chem. 0000, 0–0

www.eurjic.org

3

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

Table 1. Important bond lengths [Å], bonds angles [°] and average the molecular peak [M + H]⁺was observed for each (see

short and long Cu···Cu distances (Å) within the structures of 2–4.

the Exp. Section). Therefore it is reasonable to think that

our electrochemical and UV data (see the following sec-

tions) should be a signature of their cubane architectures.

Complex 2

Cu1–O1

Cu1–N4

2.362(4)

1.939(6)

1.959(4)

3.09

90.8(2)

104.0(2)

Cu1–O1

Cu 1–O18

1.967(4)

1.908(4)

Cu1–O1

short Cu^…Cu

Cu1–O1–Cu1

long Cu^…Cu

Cu1–O1–Cu1

3.31

99.6(2)

Electrochemistry

To determine the stabilities of the ligands and complexes,

Complex 3

the electrochemical behaviour of the three ligands L2H₂–

Cu1–O1

Cu1–N4

Cu1–O31

Cu2–O31

2.430(3)

1.952(3)

1.969(2)

2.419(3)

1.952(3)

3.12

Cu1–O1

Cu1–O12

Cu2–O1

Cu2–O31

1.949(2)

1.912(3)

1.958(2)

1.966(2)

1.922(3)

3.15

[7a,7b]

L4H₂as well as the previously reported L1H₂

was ex-

amined in anhydrous DMF containing nBu₄PF₆as sup-

porting electrolyte. This solvent was preferred because the

complexes at the concentration used in our experiments (C

= 10^–3m) are sparingly soluble in dichloromethane or ace-

tonitrile. The data are collected in Table 2 and typical cyclic

voltammograms for L3H₂and 3 are presented in Figure 2

(for L1H₂, L2H₂and L4H₂and the corresponding cubane

complexes 1, 2 and 4, see Figures SI.2–4 in the Supporting

Information). Note that electrochemical investigations of

copper cubanes remain scarce due to the presence of one,

two or four independent copper atoms within the com-

plexes.^[39]

Cu2–N34

Cu2–O42

short Cu^…Cu

long Cu^…Cu

Cu1–O1–Cu1

short Cu^…Cu

long Cu^…Cu

Cu1–O1–Cu2

Cu2–O31–Cu2

3.31

97.60(9)

3.38

90.10(9)

107.6(1)

100.3(1)

Cu1–O31–Cu2 105.1(1)

Cu1–O31–Cu2 91.27(9)

Complex 4

Cu1–O1

Cu2–O31

Cu1–O52

1.949(3)

1.957(3)

1.909(4)

2.406(4)

3.14

3.19

92.5(1)

107.8(2)

Cu1–O31

Cu1–N34

Cu2–N4

2.388(3)

1.942(4)

1.945(4)

Cu2–O1

short Cu^…Cu

long Cu^…Cu

Cu1–O1–Cu2

short Cu^…Cu

long Cu^…Cu

Cu2–O1–Cu2

Cu1–O31–Cu1

Cu1–O31–Cu2

3.16

3.24

95.9(1)

94.2(1)

108.0(2)

Table 2. Electrochemical data for the reduction of ligands L1H₂–

L4H₂and the corresponding complexes 1–4.

Cu1–O31–Cu2 92.1(1)

[a]

[b]

[c]

E_pc[V]

E_pa[V]

E_1/2[V]

EA^[d][eV]

L1H₂–2.45 (E_pc1

)

–2.95 (E_pa3

)

–3.06 [E_1/2(3)

]

0.61

–2.73 (E_pc2

and is designated as [4+2] and 3) for the third and last class,

all six Cu···Cu distances are similar and this class is denoted

as [6+0]. The main conclusion of this study evidences the

link between the class of the cubane complex and the nature

and intensity of the magnetic interactions. According to this

classification, 2 and 3 belong to the [4+2] class. The differ-

ence between the average short (3.15 Å) and long (3.22 Å)

Cu···Cu distances within 4 is small compared with the two

other complexes and therefore its architecture can also be

considered as a [6+0] cubane.

–3.17 (E_pc3

1

–2.14 (E_pc1

–2.44 (E_pc2

–2.78 (E_pc3

–0.51 (E_pc5

–0.96 (E_pc6

+0.26 (E_pa4

+1.22 (E_pa5

–0.27 (E_pa6

)

[e]

L2H₂–1.96 (E_pc1

–1.75 (E_pa1

)

–1.86 [E_1/2(1)

]

1.21

2.02

–2.40 (E_pc2

2

–1.84 (E_pc1

–2.40 (E_pc2

–0.50 (E_pc4

–0.83 (E_pc5

+0.34 (E_pa3

+1.18 (E_pa4

–0.27 (E_pa5

)

Within each structure described herein, the cubane enti-

ties are well isolated from each other as the shortest cub-

ane–cubane distance (distance between the centres of each

cubane) is 11.6, 10.7 and 12.8 Å for 2, 3 and 4, respectively.

The packing cohesion is assumed to arise from weak inter-

actions such as F···H. In the case of 4, the chiral packing

(Flack parameter close to 0) is such that the cubane com-

plexes are organized into columns running along the c axis

of the unit cell. External ligand anthracene ramifications

form rectangular cavities between cubane cores constructed

by T-shape π–π interactions. In compound 3, the azobenz-

ene function is found to exist only in the E (trans) confor-

mation but does not induce the same organization as in

compound 4 (see Figure SI.1 in the Supporting Information

for the structural packing of 2–4).

L3H₂–1.51 (E_pc1

–1.66 (E_pa2

–2.24 (E_pa3

)

–1.73 [E_1/2(2)

–2.32 [E_1/2(3)

]

–1.81 (E_pc2

–2.41 (E_pc3

–1.61 (E_pc1

–1.83 (E_pc2

–2.48 (E_pc3

–0.48 (E_pc6

–0.82 (E_pc7

3

–1.64 (E_pa2

–0.58 (E_pa4

+0.35 (E_pa5

+0.99 (E_pa6

)

–1.73 [E_1/2(2)

]

–0.26 (E_pa7

)

[e]

L4H₂–1.92 (E_pc1

–1.74 (E_pa1

)

–1.83 [E_1/2(1)

]

1.42

–2.16 (E_pc2

–2.44 (E_pc3

–1.89 (E_pc1

–2.28 (E_pc2

–0.49 (E_pc4

–0.85 (E_pc5

–0.73 (E_pa4

)

4

–0.73 (E_pa3

+1.03 (E_pa4

–0.37 (E_pa5

)

[a] Cathodic peak potential in DMF at 293 K with a glassy carbon

electrode and 0.1 m nBu₄NPF₆as supporting electrolyte. All poten-

tials were determined versus Ag/Ag⁺in acetonitrile, scan rate

0.2 Vs^–1. [b] Anodic peak potential. [c] E_1/2= (E_pc+ E_pa)/2. [d]

Electron affinity obtained from DFT calculations. [e] By scanning

Mass spectra were recorded for all the cubane complexes

(electrospray ionization mode) and it was demonstrated

that their cubic structures were retained in solution because potential from 0.0 to –2.1 V and then from –2.1 to 0.0 V.

4

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

Copper(II) Cubane Complexes

Figure 2. Cyclic voltammetry of complex 3 in anhydrous dimethylformamide and 0.1 m nBu₄NPF₆, glassy carbon electrode, scan rate:

0.2 Vs^–1, T = 293 K. a) Scanning potential from 0.0 to –3.3 V, then –3.3 to +1.5 V and then +1.5 to 0.0 V. b) Inset shows the reduction

profile of L3H₂, scanning potential from 0.0 to –3.1 V and then –3.1 to 0.0 V. c) Inset shows the oxidation profile of 3, scanning potential

from 0.0 to +1.2 V, then +1.2 to –1.0 V and then –1.0 to 0.0 V.

L1H₂–L3H₂were found to be non-oxidized up to the efficient delocalization, whereas the SOMO is only localized

limit of solvent oxidation (around +1.65 V) whereas L4H₂on the enaminone moiety in the case of L1H₂. Mulliken

is oxidized irreversibly at a potential close to +1.13 V (peak atomic spin densities of the reduced ligands (Scheme 4)

potential at 0.2 Vs^–1), which should correspond to the for- clearly show that for ligands L3H₂and L4H₂the spin den-

mation of the unstable radical cation of the anthracenyl sity is mostly delocalized over the aromatic moieties and

moiety.^[40]L2H₂(Figure SI.3 in the Supporting Infor- that for L2H₂it is slightly more localized at the C-4 posi-

mation) and L4H₂(Figure SI.4) exhibit a first one-electron tion. For the ligand L1H₂bearing a methyl moiety, the spin

quasi-reversible reduction step at –1.96 (peak potential is delocalized over the enaminone with a higher density at

0.2 Vs^–1; E_1/2= –1.86 V, peak separation close to 0.14 V) C-4. These observations are also confirmed by the spin

and –1.92 V (peak potential 0.2 Vs^–1; E_1/2= –1.83 V, peak densities at C-2, which are higher for L1H₂and L2H₂.

separation close to 0.13 V), respectively, that should corre- Thus, the introduction of aromatic groups such as azobenz-

spond to the radical anion of the enaminone moiety. For

L1H₂and L3H₂, this reduction step is irreversible and is

located at –2.45 (peak potential at 0.2 Vs^–1) and –1.51 V,

respectively.^[41]In addition, L3H₂exhibits two quasi-revers-

ible steps at –1.81 (peak potential at 0.2 Vs^–1; E_1/2

–1.73 V) and –2.41 V (peak potential at 0.2 Vs^–1; E_1/2

=

–2.32 V; Figure 2, b) that should correspond to the first and

second one-electron reductions of the –N=N– moiety.

These steps should correspond to the corresponding radical

anion based on a comparison with literature data and fur-

ther reduction of this species to the corresponding dianion,

respectively.^[42–45]An irreversible reduction step was ob-

served at –2.44 V for L4H₂(Figure SI.4) that may corre-

spond to the formation of the unstable radical anion of the

anthracenyl moiety.^[45,46]

DFT calculations were performed on all ligands and

their reduced forms (radical anions) to gain more insights

into their electrochemical behaviour. A good correlation be-

tween the calculated electron affinities and the measured

reduction potentials was obtained (see Table 2). The

SOMOs of the reduced forms of the ligands L2H₂–L4H₂

(see Figure SI.5 in the Supporting Information) are spread

Scheme 4. Schematic representation of the calculated Mulliken

atomic spin densities of the reduced form of the ligands L1H₂–

L4H₂from DFT calculations. The azobenzene and anthracene

moieties of L3H₂and L4H₂, respectively, are replaced by a phenyl-

over the entire molecule in each case, which demonstrates type ring. H and F atoms have been omitted for clarity.

Eur. J. Inorg. Chem. 0000, 0–0

www.eurjic.org

5

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

ene or anthracene generates better spin delocalization and these new Cu^IIcomplexes is not yet known and will be the

therefore stabilizes the reduced form of the ligand and facil- subject of forthcoming studies. For cubanes 1, 2 and 4, sim-

itates its reduction, as confirmed by experimental cyclic vol- ilar redox behaviour was observed, as can be seen from the

tammetry measurements and electron affinity calculations.

The redox behaviour of the four corresponding com-

plexes was also examined in anhydrous DMF containing

nBu₄PF₆as the supporting electrolyte. The data are col-

lected in Table 2 and only the cyclic voltammogram of 3 is

presented in Figure 2. For the three other compounds, quite

data collected in Table 2.

Spectroscopic Properties

The UV/Vis absorption spectra of ligands L1H₂–L4H₂

similar electrochemical profiles are observed between the and their corresponding copper cubane complexes 1–4 are

ligands and corresponding complexes (Figures SI.2–4 in the depicted in Figure 3. To confirm that the absorption prop-

Supporting Information). This does not allow us to distin- erties of the ligands were retained within the complexes,

guish with certainty a reduction potential for the Cu^IIin both free ligands and complexes were studied.

the complexes. However, it is possible that such a reduction

The spectra were compared with the results of TD-DFT

occurs at around –2 V that is overlapped by a reduction of calculations (for specific transitions and the orbitals in-

the ligand (as reported for some examples in the litera- volved, see the Supporting Information) to assign the ob-

ture).^[39]All the cubanes were found to be irreversibly oxid- served bands to electronic transitions of the ligands (Table 3

ized at potentials (peak potential at 0.2 Vs^–1) higher than and Tables SI.1–4 and Figures SI.6–9 for L1H₂–L4H₂and

+1.0 V and this oxidation process should correspond to the 1–4).

one-electron irreversible oxidation of Cu^IIinto a Cu^IIIspe-

For L1H₂, a single band at 319 nm is observed that cor-

cies. For example, in the case of 3, the Cu^IIIspecies can be responds to a π–π* transition of the enaminone moiety. For

reduced into a new Cu^IIcomplex upon scanning back to the corresponding cubane complex 1, one can see two

cathodic potentials. This new Cu^IIspecies can be further bands at 313 and 267 nm, which may represent the spectral

reduced into Cu⁰at higher reduction potentials (E_pc7

=

signature of two different enaminone tautomeric forms (en-

–0.82 V), which is also characterized by a redissolution amine and imine) induced by electron delocalization.

peak at –0.26 V (E_pa7) upon scanning back to anodic poten- For L2H₂and cubane 2, the spectra reveal the presence

tials. This new anodic peak does not correspond to a dif- of an absorption band centred at around 333 nm for the

fusion type Cu^II/Cu^Iprocess because most of the time these ligand and two bands at 323 and 264 nm for the corre-

are characterized by a sharp anodic peak typical of an ad- sponding complex. We assign the bands to different π–π*

sorption process. The newly formed Cu^IIcomplexes are transitions of the phenyl ring and the enaminone moiety

probably not cubane-type complexes because none of the (see Table SI.2 and Figure SI.7 in the Supporting Infor-

Cu^IImetallic centres of these cubanes can be electrochemi- mation). It seems that within the cubane, one observes a

cally reduced between 0.0 and –1.0 V. The exact nature of similar pattern as in complex 1 with two different enami-

Figure 3. UV/Vis absorption spectra of ligands (black curves) and the corresponding cubane complexes (gray curves) in dichloromethane

(C = 1.10^–5m). a) L1H₂and 1, b) L2H₂and 2, c) L3H₂and 3, and d) L4H₂and 4.

6

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

Copper(II) Cubane Complexes

Table 3. UV/Vis absorption data for ligands L1H₂–L4H₂and com-

plexes 1–4.

Finally, regarding L4H₂and complex 4 (see Table SI.4

and Figure SI.9), the spectra show the characteristic fine

structure of the anthracene pattern^[48]between 230 and

400 nm with an absorption maximum at 257 and 258 nm,

respectively. These are mainly due to π–π* transitions

centred on the anthracene. In each complex, the forbidden

d–d transitions are too weak in comparison with the other

photosensitive moieties to be clearly observable at this con-

centration.

In summary, the UV/Vis spectra show that the absorp-

tion properties of the ligands are retained in the cubane

complexes. Because in the cubane complexes the enaminone

is considered to be in the form L^2–, UV/Vis spectra were

also recorded for the deprotonated ligands in the presence

of Et₃N and it was found that there was no significant dif-

ference to the spectra of their LH₂forms. It will be interest-

ing to study whether, in the case of the azobenzene-derived

compound, solid-state photoisomerization can be achieved

and how this could affect the magnetic properties of the

metallic core of the complex. We are currently investigating

the luminescent ability of the anthracene-derived complex

both in solution and in the solid state. Its triplet state, if

accessible in the solid state, could influence the magnetic

properties of the entire architecture and must be studied in

detail.

Magnetic Properties

The magnetic properties of polycrystalline samples of the

cubane complexes 1–4 were investigated to identify the ther-

mal molar magnetic susceptibility (χ_M) behaviour and the

isothermal magnetization data. Data for compound 1 have

already been published.^[7b,49]Figure 4 shows the results of

the thermal variation of the χ_MT product and the magne-

tization curves recorded at 2 K for complex 2. Figures illus-

trating the magnetic behaviour of complexes 1, 3 and 4 are

included in the Supporting Information (see Figures SI.10–

12).

[a] Molar extinction coefficients (ε) were calculated at λ_max. [b] Ob-

tained by TD-DFT calculations.

none forms. At this stage, we cannot discriminate the π–π*

or the LMCT origin of the additional bands located at

267 nm for 1 and 264 nm for 2.

For L3H₂and cubane 3 (see Table SI.3 and Figure SI.8),

broad intense bands at 339 and 385 nm, respectively, are

evidenced. This is characteristic of the π–π* transition of

the azobenzene function. Usually this transition is ac-

companied by isomerization of the azo moiety and a shift

from the E (trans) to the Z (cis) configuration. The Z con-

figuration is generally identified at a lower energy than the

E configuration (generally beyond 400 nm in dichlorometh-

ane).^[47]Thus, in our case the E conformation is dominant

in solution and is retained in the cubane complex associated

with a slight bathochromic shift. In addition, the band at

298 nm for cubane 3 and the shoulder in the same region

for L3H₂can be attributed to a transition from the en-

aminone moiety (by comparison with L1H₂and cubane 1

and confirmed by the TD-DFT calculations for L3H₂).

Figure 4. Thermal dependence of χ_MT for 2 (inset shows the mag-

netization recorded at 2 K). The red curves represent the fits dis-

cussed in the text.

Eur. J. Inorg. Chem. 0000, 0–0

www.eurjic.org

7

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

The thermal variations of χ_MT exhibit the same feature Table 4. Magnetic coupling constants experimentally obtained [J_exp

and zJЈ_exp(cm^–1)] and computed [J_theo(cm^–1)], and summary of all

for all the cubanes. At room temperature, the χ_MT products

the structural parameters that can influence the magnetic proper-

ties.

of 2.08, 1.83 and 1.79 cm³Kmol^–1for 2, 3 and 4, respec-

tively, are higher than the expected value for four uncorre-

lated Cu^IIions with a g value of 2 (1.5 cm³Kmol^–1), which

indicates a slight anisotropy of g. Upon cooling, χ_MT in-

creases continuously and reaches a maximum of 3.62, 3.06

and 3.15 cm³Kmol^–1for 2, 3 and 4, respectively, at 10 K.

These values are close to the expected value for an S = 2

state (3 cm³Kmol^–1with g = 2). This feature indicates the

presence of intra-complex ferromagnetic interactions. Be-

low 10 K, the χ_MT product slowly decreases as a result of

weak intermolecular anti-ferromagnetic interactions and/or

zero-field splitting of the S = 2 ground state. The magne-

tization curve recorded at 2 K for 2 is shown in the inset of

Figure 4 (see Figures SI.10–12 for complexes 1, 3 and 4). In

all cases, it shows a continuous increase up to the saturation

value of 4.62, 4.03 and 4.18 μ_Bfor 2, 3 and 4, respectively,

which corresponds well with a ground-state spin of 2, in

agreement with the Brillouin curves calculated with S = 2

and g values of 2.28 (2), 2.03 (3) and 2.06 (4).

Regarding the structural data of compounds 1–4, their

magnetic behaviour was analysed by using two different

coupling constants J₁and J₂, expressed in the spin Hamil-

tonian of Equation (1) in Scheme 5, in agreement with the

[4+2] class of cubane clusters.^[5,38]The fitting procedure

also includes intermolecular interactions through a molecu-

lar field approximation with the zJЈ term.

nide^III–copper^IIcubanes that the application of such a

Scheme 5. [Cu₄O₄] cubane core with exchange couplings as dashed

methodology to binuclear moieties may also be a powerful

lines. The thick lines represent the short Cu–O bond lengths

tool.^[7b]Thus, CAS(2,2)SCF and subsequent DDCI calcula-

whereas the thin lines represent the long Cu–O bond lengths.

tions (see Computational Details in the Exptl. Sect.) were

performed on binuclear units extracted from the crystal

First note that the zJЈ terms are all anti-ferromagnetic structures. For each cubane 1–4, the number of binuclear

and weaker than –0.05 cm^–1(Table 4). The g values of 2.2, models studied is governed by the variety of intramolecular

2.2, 2.1 and 2.1 for 1, 2, 3 and 4, respectively, are close to Cu···Cu distances, which is related to the distortion of the

those obtained from the Brillouin curves. According to the cubane core. Two binuclear models were thus constructed

exchange parameters, J₁is ferromagnetic for all compounds for complex 2 with Cu···Cu distances of 3.31 (Figure 5) and

at 21.4 (1), 21.1 (2), 22.6 (3) and 14.5 cm^–1(4) whereas J₂3.09 Å, whereas four (d = 3.12, 3.15, 3.31 and 3.38 Å), four

varies from ferromagnetic at 0.2 (1) and 4.9 cm^–1(2) to anti- (d = 3.14, 3.16, 3.19 and 3.24 Å) and six (d = 3.11, 3.12,

ferromagnetic at –0.9 (4) and –3.6 cm^–1(3; Table 4).

3.13, 3.14, 3.30, 3.31 Å) dimer models were required for

To support the assignment of the magnetic exchange pa- complexes 3, 4 and 1, respectively. First, the need to mimic

rameters in cubanes 1–4, multi-reference configuration in- the influence of the other two Cu^IIions in the cubane either

teraction (MRCI) calculations were performed. Although by including pure electrostatic point charges or by using

calculations based on the DFT formalism can be carried pseudo-potentials was tested on the binuclear models of cu-

out on a whole cubane complex, as exemplified in recent bane 2, which is the most symmetric structure (see

studies,^[5,38,49,50]such complexes reach the current size Table SI.5). An evolution of at least 20% of the calculated

limitation for a complete description by using ab initio DDCI-3 J values was observed if pseudo-potentials were

MRCI approaches. However, it was shown recently in a included to describe the missing Cu^IIions. Thus, calcula-

thorough analysis of the magnetic properties of lantha- tions were conducted following this strategy.

8

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

Copper(II) Cubane Complexes

tween the models with d = 3.11 and 3.12 Å and those with

d = 3.13 and 3.14 Å. The small bond length elongation cou-

pled with the corresponding tiny increase in the Cu–O–Cu

angles thus seem to strongly affect the magnetic behaviour

in this case. One can conclude that for the cubane com-

plexes 1 and 3, a description of the magnetic behaviour in

terms of three coupling constants may finally be more ap-

propriate than the J₁versus J₂view classically used for the

[4+2] class of cubane clusters.

Magneto-structural correlations for the hydroxo- and

alkoxo-bridged Cu^IIcomplexes have been proposed by Hat-

field and co-workers.^[51]According to these authors, the

crucial geometrical parameter is the Cu–O–Cu angle.

Figure 5. Magnetic orbitals of the binuclear models of cubane 2

with d = 3.31 Å. The large balls correspond to the positions of the

two other copper atoms represented by pseudo-potentials in the

calculations.

Moreover, the impact of the CF₃and the R¹group (R¹Hence, in the case of the hydroxo-bridged complexes, an

= phenyl in 2) of the ligand on the calculated J value was angle of 98.5° marks the border between ferro- and anti-

also examined. For d = 3.09 Å, J(DDCI-3) = 38.2 cm^–1with ferromagnetic complexes. Ferromagnetism is found in com-

CF₃and 43.2 cm^–1with R¹, which is to be compared with pounds with smaller angles and anti-ferromagnetism with

39.2 cm^–1if neither moiety is present in the model complex larger angles. Other influential parameters were evidenced

(see Table SI.5). Such variations being notably less signifi- such as the out-of-plane shift of the coordinated atom and

cant than those induced by the description of the neigh- the Addison parameter.^[52,53]All these parameters are re-

bouring Cu^IIions by pseudo-potentials, it was decided not ported in Table 4 for samples 1–4. Unfortunately, a detailed

to include them in the subsequent calculations. Conse- investigation of the influence of these parameters on the

quently, only the distortion of the cubane core due to the magnetic exchange intensity fails to reveal any trend for the

presence of various ligands R¹in the series 1–4 was taken four cubane compounds.

into account. The electronic influence of these moieties,

which was expected to be small without light perturbation,

was thus neglected in this study. The calculated J values at

various levels of theory for the cubanes 1, 3 and 4 are given

Conclusions

in the Supporting Information (Tables SI.6–8). The values

Three cubane complexes have been crystallized with

obtained at the best level of calculation [CAS(2,2)+DDCI- novel ligands based on β-aminovinyl trifluoromethyl

3] are summarized in Table 4. As expected, the magnetic ketones. Single-crystal X-ray diffraction analysis evidenced

orbitals in all cases are those of d-type combinations with the {Cu₄O₄} cubane core of the complexes, which exhibit

a small delocalization on the coordinated atoms of the ferromagnetic exchange interactions and therefore a non-

square-pyramidal base. A slight ferromagnetic deviation zero spin ground state. These cubanes were decorated with

was observed between most of the calculated and experi- different aromatic functions such as phenyl, phenyldiazenyl

mentally obtained J values. This may partially be attributed and anthracenyl with the aim of introducing optical proper-

to the approximations made in the model systems. Al- ties (e.g., photoisomerization, phosphorescence) to the

though two sets of exchange coupling constants were calcu- magnetic core. To characterize the different properties in

lated for cubanes 2 and 4, that is, one for both the long and an independent way, electrochemical studies and absorption

short ranges of Cu···Cu distances, in agreement with the spectroscopy in solution were performed. Our first goal was

experimental magneto-structural description, one needs to to show that the ligands retain their electronic properties

take a closer look at the calculated J values for cubanes 1 within the complexes. In this sense, we hope to observe in

and 3. For complex 3, three sets of computed coupling con- future studies a real equilibrium between the trans and cis

stants were obtained. For the longest Cu···Cu distances d = conformations of the azobenzene under UV light excitation

3.31 and 3.38 Å, J_theowas calculated to be around +3 cm^–1, in the cubane complex. Indeed, photoisomerization is

which is, except for the small ferromagnetic shift discussed known to influence the ligand field around metallic ions

above, in line with the experimental fit (J_exp= 3.6 cm^–1). and therefore their magnetic behaviour such as in spin-

For d = 3.12 and 3.15 Å, J_theoequals 30.5 and 18.9 cm^–1, crossover materials.^[54]In addition, the luminescent proper-

respectively. This unexpectedly large difference between ties of the anthracene moiety are under investigation, both

these two values may be attributed to the elongation of the in solution and in the solid state. The four coordinated an-

Cu···Cu distance combined with a decrease in the Cu–O– thracene ligands in their phosphorescent state (triplet state,

Cu angles. A similar situation is observed for complex 1, S = 1) may interact magnetically with the four copper(II)

for which three sets of J_theowere calculated. Again, for the ions of the cubane core and thus influence the global mag-

binuclear models based on the longest distances (d = 3.30 netic behaviour of the cluster. Experiments will then be per-

and 3.31 Å), good agreement is obtained between the com- formed to study the correlation between azobenzene isom-

puted and experimentally proposed J values. For the binu- erization/anthracene luminescence and the magnetic prop-

clear models with the four shortest distances, an unexpected erties of the metallic cubane core. Indeed, our goal was to

difference of around 10 cm^–1in J value was computed be- generate synergistic effects between different magnetic and

Eur. J. Inorg. Chem. 0000, 0–0

www.eurjic.org

9

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

H), 2.69 (br. s, 1 H) ppm. ¹³C NMR (75 MHz, CDCl₃): δ = 176.2

(q, J = 33.0 Hz), 170.2, 153.5, 152.4, 135.6, 131.7, 129.2, 128.6,

123.3, 123.2, 117.5 (q, J = 286.0 Hz), 90.5, 61.2, 47.4 ppm. ¹⁹F

NMR (282 MHz, CDCl₃): δ = –76.93 ppm. HRMS (ESI): calcd.

optical properties. Finally, our experience of cubane com-

plexes as building blocks for the elaboration of SMMs with

such architectures is currently being exploited.^[7a,7b]

for C₁₈H₁₅F₃N₃O₂[M

–

H]^–362.1122; found 362.1123.

C₁₈H₁₆F₃N₃O₂(363.33): calcd. C 59.50, H 4.44, N 11.57; found C

59.68, H 4.51, N 11.39.

Experimental Section

(Z)-4-(Anthracen-2-yl)-1,1,1-trifluoro-4-[(2-hydroxyethyl)amino]but-

3-en-2-one (L4H₂): Yield 95%; yellow solid; m.p.155 °C. H NMR

General: Anhydrous solvents were used from commercial suppliers.

Reagents were obtained commercially and used without further pu-

rification. Compounds A and C are commercially available. Com-

pounds B, E, F, H and I have been recently described in the litera-

ture.^[35b]Compounds D and G were prepared as described in the

literature.^{[36] 1}H, ¹⁹F and ¹³C NMR spectra were recorded with a

Bruker Avance 300 spectrometer in CDCl₃at 300, 282 and 75 MHz

respectively. Chemical shifts are given in ppm (s = singlet, d =

doublet, dd = doublet of doublets, t = triplet, m = multiplet, br. =

broad) relative to the residual solvent peak (δ_H= 7.26 ppm for

CHCl₃, δ_C= 77.0 ppm for CDCl₃) or CFCl₃(¹⁹F). Coupling con-

stants (J) are given in Hz. TLC was performed on Merck silica Gel

60 F₂₅₄plates with detection by UV light. Silica gel chromatog-

raphy was performed on Macherey–Nagel silica gel 60M (0.04–

0.063 mm). Solvents used for chromatography, work-up and

recrystallization were acetone (AC), dichloromethane (DCM), ethyl

acetate (EA), methanol (ME), petroleum ether (EP) and toluene

(TOL). Mass spectra were recorded with a Finnigan MAT 95 (EI)

and MicroTOF-Q (ESI) spectrometers and UV/Vis spectra with a

Lambda 35 spectrophotometer. Melting points were determined in

capillary tubes on a Büchi apparatus. Electrochemical measure-

ments were performed by using an EG & G-Princeton Applied Re-

search 263A all-in-one potentiostat with a standard three-electrode

set-up with a glassy carbon electrode (working electrode), platinum

wire auxiliary electrode and a non-aqueous Ag/Ag⁺system in ace-

tonitrile as the reference electrode. DMF solutions containing the

compounds under the study contained 0.1 m of the supporting elec-

trolyte nBu₄NPF₆and a voltage scan rate of 0.2 Vs^–1was used.

Under these experimental conditions the ferrocene/ferricinium cou-

ple, used as internal reference for potential measurements, was lo-

cated at E_1/2= +0.05 V in DMF. Mass spectra of all the dimethyl

acetals were not recorded due to their instability during the ioniza-

tion process.

1

(300 MHz, CDCl₃): δ = 11.44 (br. s, 1 H), 8.49 (s, 1 H), 8.46 (s, 1

H), 8.09 (s, 1 H), 8.06–8.02 (m, 3 H), 7.55–7.52 (m, 2 H), 7.40 (dd,

J = 9.0, J = 1.5 Hz, 1 H), 5.63 (s, 1 H), 3.81 (t, J = 4.8 Hz, 2 H),

3.56 (q, J = 5.4 Hz, 2 H), 2.87 (br. s, 1 H) ppm. ¹³C NMR (75 MHz,

CDCl₃): δ = 175.9, 171.1, 132.7, 132.3, 131.2, 130.5, 130.3, 129.1,

128.4, 128.3, 128.2, 127.6, 126.5, 126.5, 126.4, 126.2, 123.5, 90.8,

61.4, 47.5 ppm. ¹⁹F NMR (282 MHz, CDCl₃): δ = –76.89 ppm.

HRMS (ESI): calcd. for C₂₀H₁₇F₃NO₂[M + H]⁺360.1206; found

360.1218. C₂₀H₁₆F₃NO₂(359.34): calcd. C 66.85, H 4.49, N 3.90;

found C 67.02, H 4.18, N 4.12.

General Procedure for the Synthesis of Complexes: A solution of

Cu(ClO₄)₂·6H₂O (1.2 mmol, 1.2 equiv.) in methanol (10 mL) was

added dropwise to a stirred mixture of L2H₂, L3H₂or L4H₂

(1 mmol, 1 equiv.) and triethylamine (2 mmol, 2 equiv.) in dichloro-

methane (20 mL) resulting in instantaneous change in color of the

solution to deep-blue (for 2) or deep-green (for 3 and 4). After

stirring for 2 h at room temperature, solvents were removed under

reduced pressure and filtered through a very small pad of silica

(eluent: EA). The filtrate was evaporated under reduced pressure

and crystals were obtained by slow evaporation of EA/AC (1:1) for

[Cu₄(L2)₄]·2.4[OC(CH₃)₂] (2), EP/DCM (9:1) for [Cu₄(L3)₄] (3) and

TOL for [Cu₄(L4)₄]·6(MePh) (4).

2: C_55.2H_54.4Cu₄F₁₂N₄O_10.4(1422.42): calcd. (after loss of 2.4 equiv.

of acetone) C 44.93, H 3.14, N 4.37; found C 44.88, H 3.25, N

4.39.

3: C₇₂H₅₆Cu₄F₁₂N₁₂O₈(1699.47): calcd. C 50.89, H 3.32, N 9.89;

found C 51.06, H 3.67, N 9.59.

4: C₁₂₂H₁₀₄Cu₄F₁₂N₄O₈(2236.35): calcd. (after loss of 6 equiv. of

toluene) C 57.08, H 3.35, N 3.33; found C 56.70, H 3.40, N 3.65.

Single-Crystal X-ray Diffraction: Single-crystal X-ray studies of 2–

4 were carried out by using a Gemini diffractometer and the related

analysis software.^[55]An absorption correction based on the crystal

faces was applied to the data sets (analytical).^[56]All structures were

solved by direct methods using the SIR97 program^[57]combined

with Fourier difference syntheses and refined against F using reflec-

tions with [I/σ(I) Ͼ 3] by using the CRYSTALS program.^[58]All

atomic displacement parameters for non-hydrogen atoms were re-

fined with anisotropic terms. The hydrogen atoms were theoretic-

ally located on the basis of the conformation of the supporting

atom and refined by using the riding model. Selected crystallo-

graphic data are presented in Table 5.

General Procedure for the Addition of Ethanolamine: Ethanolamine

(1 mmol, 1 equiv.) was added dropwise to a stirred solution of

methoxyvinyl trifluoromethyl ketones G,

H and I (1 mmol,

1 equiv.) in anhydrous acetonitrile (2 mL) at 0 °C. The mixture was

stirred at room temperature overnight (TLC monitoring). The sol-

vent was then removed under reduced pressure and the crude solid

was purified by silica gel chromatography (eluent: EP/EA, 1:1).

(Z)-1,1,1-Trifluoro-4-[(2-hydroxyethyl)amino]-4-phenylbut-3-en-2-

one (L2H₂): Yield 98 %; yellow solid; m.p. 50 °C. ¹H NMR

(300 MHz, CDCl₃): δ = 11.40 (br. s, 1 H), 7.45–7.41 (m, 3 H), 7.39–

7.35 (m, 2 H), 5.42 (s, 1 H), 4.41 (br. s, 1 H), 3.73 (t, J = 5.1 Hz,

2 H), 3.45 (q, J = 5.4 Hz, 2 H) ppm. ¹³C NMR (75 MHz, CDCl₃):

δ = 175.0 (q, J = 32.4 Hz), 171.2, 133.4, 130.3, 128.6, 127.3, 117.6

(q, J = 286.4 Hz), 90.1, 60.6, 47.3 ppm. ¹⁹F NMR (282 MHz,

CDCl₃): δ = –76.8 (s) ppm. HRMS (ESI): calcd. for C₁₂H₁₃F₃NO₂

[M + H]⁺260.0893; found 260.0904. C₁₂H₁₂F₃NO₂(259.22): calcd.

C 55.60, H 4.67, N 5.40; found C 55.53, H 4.72, N 5.36.

Magnetic Measurements: Magnetic susceptibilities were measured

with a MPMS5 SQUID magnetometer (Quantum Device) in the

temperature range 5–300 K under an applied magnetic field of

2000 Oe. Measurements at 5000 Oe did not evidence significant dif-

ferences in the susceptibility curves. Magnetization curves were re-

corded at 2 K up to 7 T by using the field sweeping mode of the

magnetometer. The samples were precisely weighted and correc-

tions were applied to account for the compounds and the sample

holder diamagnetic contributions.

(Z)-1,1,1-Trifluoro-4-[(2-hydroxyethyl)amino]-4-{4-[(E)-phenyl-

diazenyl]phenyl}but-3-en-2-one (L3H₂): Yield 90%; orange-red so-

1

lid; m.p. 101 °C. H NMR (300 MHz, CDCl₃): δ = 11.35 (br. s, 1

H), 8.02–7.94 (m, 4 H), 7.56 (d, J = 8.4 Hz, 2 H), 7.53–7.52 (m, 3 Computational Details: DFT and TD-DFT calculations were car-

H), 5.52 (s, 1 H), 3.80 (t, J = 4.8 Hz, 2 H), 3.51 (q, J = 5.4 Hz, 2 ried out with the Gaussian 09 program^[59]at the PBE0/6-

10

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

Copper(II) Cubane Complexes

Table 5. Single-crystal X-ray diffraction data and structure refinement details for complexes 2–4.

2

3

4

Refined formula

M [gmol^–1]

Crystal system

Space group

a [Å]

b [Å]

c [Å]

C

_55.2H_54.4Cu₄F₁₂N₄O_10.4

C₇₂H₅₆Cu₄F₁₂N₁₂O₈

1699.47

monoclinic

C2/c

18.927(2)

25.682(1)

18.468(1)

120.50(1)

7735(1)

4

C₁₂₂H₁₀₄Cu₄F₁₂N₄O₈

2236.35

orthorhombic

Fdd2

37.382(2)

29.827(4)

18.2370(7)

90

1422.42

tetragonal

I 4₁/a

22.611(2)

11.621(2)

90

β [°]

V [Å³]

5942(1)

4

20334(2)

8

Z

T [K]

110

1.459

1.174

9729

110

1.461

0.911

11384

D [gcm^–3]

1.587

1.507

3697

0.104

0.0641

0.0677

1.13

μ [mm^–1]

Independent reflections

R_int

R(F) [IϾ3σ(I)]

R_w(F) [IϾ3σ(I)]

S

0.052

0.033

0.0470

0.0502

1.02

0.0563

0.0695

1.09

Reflections

Refined parameters

Δρ_max[eÅ^–3]

Δρ_min[eÅ^–3]

Flack parameter

Absorption correction

1871

208

2.10

–1.14

–

5317

487

0.80

–0.32

8569

677

1.27

–0.57

–0.02(2)

analytical

–

analytical

311G(2d,p) level of theory. Explicitly correlated ab initio (configu-

ration interactions, CI) and density functional theory (DFT)

broken-symmetry calculations have both proven to be powerful

tools in the understanding of magnetic systems. Nevertheless, we

decided to perform CI calculations that use the exact Hamiltonian

because the information extracted from a multi-configurational

wavefunction provides useful insights into the mechanism of mag-

netic interactions. These calculations were performed on binuclear

Cu^IIunits ensuing from the crystallographic data without any ge-

ometry optimization. With a view to reducing computational ef-

fort, the impact of the CF₃/R¹ligands on the calculated exchange

coupling parameter J was examined for R¹= phenyl, that is, cub-

ane 1 (see main text). From the results of this first study, for the

rest of the calculations, these moieties were replaced by H. Com-

plete active space self-consistent field (CASSCF) calculations, in-

cluding two electrons in two molecular orbitals (MOs), were per-

formed by using the MOLCAS 7.2 package^[60]to generate a refer-

ence space [CAS(2,2)] that consists of the configurations that quali-

tatively describe the problem. The dynamic correlation effects were

then incorporated on top of the triplet CASSCF wavefunction by

using the dedicated difference configuration interaction (DDCI)

method^[61]implemented in the CASDI code.^[62]With this approach,

one concentrates on the differential effects rather than on the evalu-

ation of the absolute energies. All atoms were depicted with ANO-

RCC-type basis sets. The Cu atoms were described with a

of χ_MT for 1, 3 and 4, calculated magnetic coupling constants at

various CI levels for 1–4.

Acknowledgments

The authors thank the Région Rhône-Alpes (CIBLE project “EN-

AMBRIMOLE”), Université Claude Bernard Lyon 1 and the Cen-

tre National de la Recherche Scientifique (CNRS) for financial sup-

port. N. C. is grateful to the Région Rhône-Alpes for a Ph.D. fel-

lowship. G. P. thanks the Centre de Diffraction Henri Longcham-

bon of the University Claude Bernard Lyon 1 for access to the

single-crystal X-ray diffraction diffractometers. B. L. G. acknowl-

edges the Pôle Scientifique de Modélisation Numérique (PSMN)

at ENS-Lyon for computing facilities. We are grateful to the Centre

Commun de Spectrométrie de Masse (CCSM) of Université Claude

Bernard Lyon 1 (F. Albrieux, C. Duchet and N. Enriques) for the

MS analyses. G. C. would like to thank Rodolphe Clérac for fruit-

ful discussions. P. Goekjian is warmly thanked for careful reading

of the manuscript and for suggestions.

[1] P. Gutlich, H. A. Goodwin (Eds.), Spin Crossover in Transition

Metal Compounds I–III, in: Topics in Current Chemistry 233–

235, Springer, Heidelberg, Berlin, 2004.

[2] a) D. Gatteschi, R. Sessoli, Angew. Chem. 2003, 115, 278; An-

gew. Chem. Int. Ed. 2003, 42, 268–297; b) G. Christou, Polyhe-

dron 2005, 24, 2065–2075; c) D. Gatteschi, R. Sessoli, J. Villain,

Molecular Nanomagnets, Oxford University Press, Oxford,

2006.

[3] a) C. Coulon, H. Miyasaka, R. Clérac, Struct. Bonding

(Berlin) 2006, 122, 163–207; b) R. Sessoli, Angew. Chem. 2008,

120, 5590; Angew. Chem. Int. Ed. 2008, 47, 5508–5510.

[4] a) T. Shiga, H. Oshio, Sci. Technol. Adv. Mater. 2005, 6, 565–

570; b) O. Kahn, Molecular Magnetism, VCH Publishers, New

York, 1993.

(21s15p10d6f4g2h)/[5s4p2d1f] contraction.^[63]

A (14s9p4d3f2g)/

[3s2p1d] contraction was used for O and N, and a (14s9p4d3f2g)/

[3s2p] contraction was used for C.^[64]A (8s4p3d1f)/[2s] contraction

was used for the H atoms.^[65]Finally, to mimic the influence of the

other two Cu^IIions in the cubane, copper cations were included

either as pure electrostatic point charges or through a Zn^IIpseudo-

potential.^[66]

Supporting Information (see footnote on the first page of this arti-

cle): Structural packing for 2–4, cyclic voltammograms for L1H₂,

L2H₂, L4H₂and 1, 2, 4, SOMO and spin density of the reduced

form for L2H₂–L4H₂, TD-DFT calculations, thermal dependence

[5] J. Tercero, E. Ruiz, S. Alvarez, A. Rodriguez-Fortea, P. Alem-

any, J. Mater. Chem. 2006, 16, 2729–2735.

Eur. J. Inorg. Chem. 0000, 0–0

www.eurjic.org

11

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

[6] For example, see: a) A. Ferguson, J. Lawrence, A. Parkin, J.

Sanchez-Benitez, K. V. Kamenev, E. K. Brechin, W.

Wernsdorfer, S. Hill, M. Murrie, Dalton Trans. 2008, 6409–

6414; b) H. Oshio, N. Hoshino, T. Ito, M. Nakano, J. Am.

Chem. Soc. 2004, 126, 8805–8812; c) C. J. Milios, A. Presci-

mone, A. Mishra, S. Parsons, W. Wernsdorfer, G. Christou,

S. P. Perlepes, E. K. Brechin, Chem. Commun. 2007, 153–155;

d) H. Oshio, M. Nakano, Chem. Eur. J. 2005, 11, 1702–1708;

e) K. W. Galloway, A. M. Whyte, W. Wernsdorfer, J. Shancez-

Benitez, K. V. Kamenev, A. Parkin, R. D. Peacock, M. Murrie,

Inorg. Chem. 2008, 47, 7438–7442; f) K. Isele, F. Gigon, A. F.

Williams, G. Bernardinelli, P. Franz, S. Decurtins, Dalton

Trans. 2007, 332–341.

[7] For example, see: a) C. Aronica, G. Pilet, G. Chastanet, J.-F.

Jacquot, D. Luneau, Angew. Chem. 2006, 118, 4775; Angew.

Chem. Int. Ed. 2006, 45, 4659–4662; b) C. Aronica, G. Chas-

tanet, G. Pilet, B. Le Guennic, V. Robert, W. Wernsdorfer, D.

Luneau, Inorg. Chem. 2007, 46, 6108–6119; c) B. Moubaraki,

K. S. Murray, T. A. Hudson, R. Robson, Eur. J. Inorg. Chem.

2008, 4525–4529.

[32] V. A. Reznikov, I. Yu. Bagryanskaya, Yu. V. Gatilov, Russ.

Chem. Bull. 2000, 49, 899–905.

[33]

G. I. Roshchupkina, T. V. Rybalova, Y. V. Gatilov, V. A. Rezni-

kov, Eur. J. Org. Chem. 2003, 3599–3602.

[34] G. Pilet, J.-B. Tommasino, F. Fenain, R. Matrak, M. Médebi-

elle, Dalton Trans. 2008, 5621–5626.

[35] a) G. Pilet, M. Médebielle, J.-B. Tommasino, G. Chastanet, B.

Le Guennic, C. Train, Eur. J. Inorg. Chem. 2009, 4718–4726; b)

N. Chopin, M. Médebielle, G. Pilet, Eur. J. Inorg. Chem. 2012,

1093–1103.

[36] N. Chopin, S. Decamps, A. Gouger, M. Médebielle, S. Picot,

A.-L. Bienvenu, G. Pilet, J. Fluorine Chem. 2011, 132, 850–857.

[37] J. H. Collins, H. H. Jaffe, J. Am. Chem. Soc. 1962, 84, 4708–

4712.

[38] E. Ruiz, A. Rodrıguez-Fortea, P. Alemany, S. Alvarez, Polyhe-

dron 2001, 20, 1323–1327.

[39] a) M. A. El-Sayed, A. H. Abdel Salam, T. A. El-Zayat, A. El-

Dissouki, Transition Met. Chem. 2005, 30, 616–622; b) J. Chak-

raborty, S. Thakurta, G. Pilet, D. Luneau, S. Mitra, Polyhedron

2009, 28, 819–825.

[8] S. V. Druzhinin, E. S. Balenkova, V. G. Nenajdenko, Tetrahe-

dron 2007, 63, 7753–7808.

[40] O. Hammerich, V. D. Parker, J. Am. Chem. Soc. 1974, 96, 4289–

4296.

[9] U. M. Kauhanka, M. M. Kauhanka, Liq. Cryst. 2006, 33, 213–

[41] We already know from ongoing synthetic work in these

laboratories that electrochemical reduction of several trifluoro-

methylated enaminones give rise to trifluoromethylated amino-

derived cyclopentanes resulting from an intramolecular dimer-

ization aldol process as well as the formation of the corre-

sponding amines; C. Agostini, A. Bianco, M. Médebielle, J.-B.

Tommasino, unpublished results.

[42] J. L. Sadler, A. J. Bard, J. Am. Chem. Soc. 1968, 90, 1979–1989.

[43] R. Lines, B. Svensmark Jensen, V. D. Parker, Acta Chem.

Scand., Ser. B 1978, 32, 510–514.

218.

[10] C. P. Roll, A. G. Martin, H. Görls, G. Leibeling, D. Guillon,

B. Donnio, W. Weigand, J. Mater. Chem. 2004, 14, 1722–1730.

[11] U. Pietrasik, J. Szydlowska, A. Krowczynski, Chem. Mater.

2004, 16, 1485–1492.

[12] A. Krowczynski, J. Szydlowska, D. Pociecha, J. Przedmojski,

E. Gorecka, Liq. Cryst. 1998, 25, 117–121.

[13] L. Zhang, M. Brookhar, P. S. White, Organometallics 2006, 25,

1868–1874.

[14] H.-Y. Wang, J. Zhang, X. Meng, G.-X. Jin, J. Organomet.

Chem. 2006, 691, 1275–1281.

[15] X.-F. Li, K. Dai, W.-P. Ye, L. Pan, Y.-S. Li, Organometallics

2004, 23, 1223–1230.

[16] U. Beckmann, E. Eichberger, M. Lindner, M. Bongartz, P. C.

Kunz, Eur. J. Org. Chem. 2008, 4139–4147.

[44] J. C. Dabrowiak, D. P. Fisher, F. C. McElroy, D. J. Macero, In-

org. Chem. 1979, 18, 2304–2307.

[45]

J. Simonet, J.-F. Pilard, Electrogenerated Reagents, in: Organic

Electrochemistry (Eds.: H. Lund, O. Hammerich), 4th ed., Mar-

cel Dekker, New York, 2001, pp. 1163–1225.

[46]

[47]

K. Meerholz, J. Heinze, J. Am. Chem. Soc. 1989, 111, 2325–

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

K. Vlasov, N. Kalinovich, E. Lork, G.-V. Röschenthaler, J. Flu-

orine Chem. 2010, 131, 996–999.

J. Przyborowski, E. Lork, G.-V. Röschenthaler, J. Fluorine

Chem. 2000, 104, 207–213.

J. A. Barten, A. A. Kadyrov, G.-V. Röschenthaler, J. Fluorine

Chem. 2002, 113, 101–103.

J. A. Barten, A. A. Kadyrov, E. Lork, G.-V. Röschenthaler, J.

Fluorine Chem. 2002, 116, 87–92.

M. Médebielle, S. Hohn, E. Okada, H. Myoken, D. Shibata,

Tetrahedron Lett. 2005, 46, 7817–7821.

M. Médebielle, O. Onomura, R. Keirouz, E. Okada, H. Yano,

T. Terauchi, Synthesis 2002, 2601–2608.

F. Fenain, M. Médebielle, M. Rocher, O. Onomura, E. Okada,

D. Shibata, J. Fluorine Chem. 2007, 128, 1286–1299.

E. Okada, T. Kinomura, Y. Higashiyama, H. Takeuchi, M.

Hojo, Heterocycles 1997, 46, 129–132.

2326.

For example, see: a) J. Arias, M. Bardaji, P. Espinet, C. L. Fol-

cia, J. Ortega, J. Etxebarria, Inorg. Chem. 2009, 48, 6205–6210;

b) M. Han, D. Ishikawa, E. Muto, M. Hara, J. Lumin. 2009,

129, 1163–1168.

[48]

[49]

N. Iza, M. Gil, J. L. Montero, J. Morcillo, J. Mol. Struct. 1988,

175, 323–328.

C. Aronica, Y. Chumakov, E. Jeanneau, D. Luneau, P. Neuge-

bauer, A.-L. Barra, B. Gillon, A. Goujon, A. Cousson, J.

Tercero, E. Ruiz, Chem. Eur. J. 2008, 14, 9540–9548.

S. Thakurta, P. Roy, R. J. Butcher, M. Salah El Fallah, J.

Tercero, E. Garribba, S. Mitra, Eur. J. Inorg. Chem. 2009,

4385–4395.

[50]

[51]

[52]

V. H. Crawford, H. W. Richardson, J. R. Wasson, D. J. Hodg-

son, W. E. Hatfield, Inorg. Chem. 1976, 15, 2107–2110.

a) E. Ruiz, P. Alemany, S. Alvarez, J. Cano, J. Am. Chem. Soc.

1997, 119, 1297–1303; b) E. Ruiz, P. Alemany, S. Alvarez, J.

Cano, Inorg. Chem. 1997, 36, 3683–3688.

[25]

[26]

M. Hojo, R. Masuda, E. Okada, Synthesis 1986, 1013–1014.

M. Hojo, R. Masuda, E. Okada, H. Yamamoto, K. Morimoto,

K. Okada, Synthesis 1990, 195–198.

[53]

[54]

[55]

[56]

[57]

A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn, G. C. Ver-

schoor, J. Chem. Soc., Dalton Trans. 1984, 7, 1349–1356.

M.-L. Boillot, J. Zarembowich, A. Sour, Top. Curr. Chem.

2004, 234, 261–276.

CrysAlisPro, v. 1.171.33.46 (rel. 27-08-2009 CrysAlis171.NET),

Oxford Diffraction Ltd., 2009.

[27]

[28]

N. Ota, T. Tomoda, N. Terai, Y. Kamitori, D. Shibata, M.

Médebielle, E. Okada, Heterocycles 2008, 76, 1205–1217.

N. S. Boltachova, O. V. Fedorova, I. G. Ovchinnikova, O. N.

Kazheva, A. N. Chekhlov, O. A. Dyachenko, G. L. Rusinov,

V. I. Filyakova, V. N. Charushin, J. Fluorine Chem. 2007, 128,

762–768.

J. de Meulenaar, H. Tompa, Acta Crystallogr. 1965, 19, 1014–

1018.

[29]

E. Yu. Fursova, V. I. Ovcharenko, G. V. Romanenko, V. N.

Ikorskii, Yu. G. Shvedenkov, I. P. Asanov, L. N. Mazalov, J.

Struct. Chem. 2001, 42, 251–260.

G. Cascarano, A. Altomare, C. Giacovazzo, A. Guagliardi,

A. G. G. Moliterni, D. Siliqi, M. C. Burla, G. Polidori, M. Ca-

malli, Acta Crystallogr., Sect. A 1996, 52, C-79.

D. J. Watkin, C. K. Prout, J. R. Carruthers, P. W. Betteridge,

CRYSTALS Issue 11, Chemical Crystallography Laboratory,

Oxford, 1999.

[30]

[31]

G. V. Romanenko, I. V. Korobkov, V. I. Ovcharenko, J. Struct.

Chem. 2000, 41, 1035–1040.

H. C. Budnikov, O. Yu. Kargina, J. Electroanal. Chem. 1984,

169, 195–205.

[58]

12

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

Copper(II) Cubane Complexes

[59] M. J. Frisch, W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.

Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci,

G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.

Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-

nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-

gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,

T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M.

Bearpak, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov,

R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C.

Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Mil-

lam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo,

J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J.

Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,

K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.

Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Fore-

sman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,

rev. A.02, Gaussian, Inc., Wallingford, CT, 2009.

Roos, L. Serrano-Andrés, M. Urban, V. Veryazov, R. Lindh, J.

Comput. Chem. 2010, 31, 224–247.

a) J. Miralles, J.-P. Daudey, R. Caballol, Chem. Phys. Lett.

1992, 198, 555–562; b) J. Miralles, O. Castell, R. Caballol, J.-P.

Malrieu, Chem. Phys. 1993, 172, 33–43.

N. Ben Amor, D. Maynau, Chem. Phys. Lett. 1998, 286, 211–

220.

B. O. Roos, R. Lindh, P. A. Malmqvist, V. Veryazov, P. O. Wid-

mark, J. Phys. Chem. A 2005, 109, 6575–6579.

[61]

[62]

[63]

[64] B. O. Roos, R. Lindh, P. A. Malmqvist, V. Veryazov, P. O. Wid-

mark, J. Phys. Chem. A 2004, 108, 2851–2858.

[65] P. Widmark, P. A. Malmqvist, B. O. Roos, Theor. Chim. Acta

1990, 77, 291–306.

[66] S. Lopez-Moraza, J. L. Pascual, Z. Barandiaran, J. Chem.

Phys. 1995, 103, 2117–2125.

Received: July 3, 2012

Published Online: ᭿

[60] F. Aquilante, L. De Vico, N. Ferré, G. Ghigo, P.-A. Malmqvist,

P. Neogrady, T. Bondo Pedersen, M. Pitonak, M. Reiher, B. O.

Eur. J. Inorg. Chem. 0000, 0–0

www.eurjic.org

13

Job/Unit: I20737

/KAP1

Date: 11-09-12 16:21:21

Pages: 14

N. Chopin, G. Chastanet, B. Le Guennic, M. Médebielle, G. Pilet

FULL PAPER

Copper Cubane Complexes

N. Chopin,* G. Chastanet, B. Le Guennic,

M. Médebielle,* G. Pilet* ............... 1–14

Copper(II) Cubane Complexes Built from

Electro- and Photosensitive β-Aminovinyl

Trifluoromethyl Ketone Ligands

Keywords: Copper / Cubane complexes /

Magnetic properties / Density functional

calculations

A new family of [Cu₄L₄] copper(II) cubane

complexes with a magnetic core and photo-

sensitive ligands have been synthesized by

the self-assembly of Cu^IIsalts and β-

aminovinyl trifluoromethyl ketone ligands

and characterized by single-crystal X-ray

diffraction, UV/Vis spectrophotometry,

electrochemistry, magnetism and theoreti-

cal calculations.

14

www.eurjic.org

Eur. J. Inorg. Chem. 0000, 0–0

Article Doi

Copper(II) cubane complexes built from electro- and photosensitive β-aminovinyl trifluoromethyl ketone ligands

DOI: 10.1002/ejic.201200737

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejic.201200737

Full text of DOI:10.1002/ejic.201200737

Products guided by the article

R&D Labs maybe for 1421609-50-8

Relevant to this article

Hot Product