11
This paper is published as part of a CrystEngComm themed issue on: Crystal Engineering in Molecular Magnetism Guest Editors Concepció Rovira and Jaume Veciana Institut de Ciència de Materials de Barcelona (ICMAB), Spain Published in issue 10, 2009 of CrystEngComm Images reproduced with permission of Enrique Colacio (left) and Kunio Awaga (right) Papers published in this issue include: Towards high T c octacyanometalate-based networks Barbara Sieklucka, Robert Podgajny, Dawid Pinkowicz, Beata Nowicka, Tomasz Korzeniak, Maria Bałanda, Tadeusz Wasiutyński, Robert Pełka, Magdalena Makarewicz, Mariusz Czapla, Michał Rams, Bartłomiej Gaweł and Wiesław Łasocha, CrystEngComm, 2009, DOI: 10.1039/b905912a Cooperativity from electrostatic interactions: understanding bistability in molecular crystals Gabriele D'Avino, Luca Grisanti, Anna Painelli, Judith Guasch, Imma Ratera and Jaume Veciana, CrystEngComm, 2009, DOI: 10.1039/b907184a Anion encapsulation promoted by anion π interactions in rationally designed hexanuclear antiferromagnetic wheels: synthesis, structure and magnetic properties Enrique Colacio, Hakima Aouryaghal, Antonio J. Mota, Joan Cano, Reijo Sillanpää and A. Rodríguez-Diéguez, CrystEngComm, 2009, DOI: 10.1039/b906382j Fe(II) spincrossover complex of [1,2,5]thiadiazolo[3,4-f][1,10]phenanthroline Yoshiaki Shuku, Rie Suizu, Kunio Awaga and Osamu Sato, CrystEngComm, 2009, DOI: 10.1039/b906845g Visit the CrystEngComm website for more cutting-edge crystal engineering research www.rsc.org/crystengcomm

Crystal Engineering in Molecular Magnetism · Crystal Engineering in Molecular Magnetism ... (ICMAB), Spain Published in issue 10, 2009 of CrystEngComm Images reproduced with permission

  • Upload
    vannhi

  • View
    216

  • Download
    3

Embed Size (px)

Citation preview

This paper is published as part of a CrystEngComm themed issue on:

Crystal Engineering in Molecular Magnetism

Guest Editors Concepció Rovira and Jaume Veciana Institut de Ciència de Materials de Barcelona (ICMAB), Spain

Published in issue 10, 2009 of CrystEngComm

Images reproduced with permission of Enrique Colacio (left) and Kunio Awaga (right) Papers published in this issue include: Towards high Tc octacyanometalate-based networks Barbara Sieklucka, Robert Podgajny, Dawid Pinkowicz, Beata Nowicka, Tomasz Korzeniak, Maria Bałanda, Tadeusz Wasiutyński, Robert Pełka, Magdalena Makarewicz, Mariusz Czapla, Michał Rams, Bartłomiej Gaweł and Wiesław Łasocha, CrystEngComm, 2009, DOI: 10.1039/b905912a Cooperativity from electrostatic interactions: understanding bistability in molecular crystals Gabriele D'Avino, Luca Grisanti, Anna Painelli, Judith Guasch, Imma Ratera and Jaume Veciana, CrystEngComm, 2009, DOI: 10.1039/b907184a Anion encapsulation promoted by anion…π interactions in rationally designed hexanuclear antiferromagnetic wheels: synthesis, structure and magnetic properties Enrique Colacio, Hakima Aouryaghal, Antonio J. Mota, Joan Cano, Reijo Sillanpää and A. Rodríguez-Diéguez, CrystEngComm, 2009, DOI: 10.1039/b906382j Fe(II) spincrossover complex of [1,2,5]thiadiazolo[3,4-f][1,10]phenanthroline Yoshiaki Shuku, Rie Suizu, Kunio Awaga and Osamu Sato, CrystEngComm, 2009, DOI: 10.1039/b906845g

Visit the CrystEngComm website for more cutting-edge crystal engineering research www.rsc.org/crystengcomm

PAPER www.rsc.org/crystengcomm | CrystEngComm

Charge-transfer two-dimensional layers constructed from a 2 : 1 assemblyof paddlewheel diruthenium(II,II) complexes andbis[1,2,5]thiadizolotetracyanoquinodimethane: bulk magnetic behavior asa function of inter-layer interactions†

N. Motokawa,a T. Oyama,a S. Matsunaga,‡a H. Miyasaka,*a M. Yamashitaa and K. R. Dunbar*b

Received 18th March 2009, Accepted 29th June 2009

First published as an Advance Article on the web 3rd August 2009

DOI: 10.1039/b905486c

Two new charge-transfer compounds, [{Ru2(O2CPh-x-F)4}2(BTDA-TCNQ)]$n(solv) (x-F-PhCO2� ¼

x-fluorobenzoate; x ¼ o-, 4; p-, 5; BTDA-TCNQ ¼ bis[1,2,5]thiadizolotetracyanoquinodimethane),

have been prepared from reactions of paddlewheel diruthenium(II, II) complexes, [Ru2II,II(O2CPh-x-

F)4] (x ¼ o-, p-) and BTDA-TCNQ. Compounds 4 and 5 crystallize as two-dimensional (2-D) layered

structures composed of [Ru2] units and BTDA-TCNQ in a 2 : 1 ratio. The isomer with the m-

fluorobenzoate-bridged [Ru2] unit (3) exhibits a three-dimensional (3-D) structure. The oxidation states

of the constituent units were evaluated from their crystal structures, and assigned to be close to the

charge-polarized state represented formally as [Ru25+]–[BTDA-TCNQc�]–[Ru2

4+] including two types

of [Ru2] units that are electronically and structurally different and a one-electron-transferred BTDA-

TCNQ radical anion. The data indicate that a 1-e� transfer from one site of [Ru2] units to BTDA-

TCNQ has likely occurred. The units are paramagnetic with heterospin states for [Ru2II,II] (S ¼ 1),

[Ru2II,III] (S ¼ 3/2), and BTDA-TCNQc� (S ¼ 1/2), which lead to magnetic ordering in the layer. In the

case of 4, antiferromagnetic ordering occurs at TN ¼ 93 K due to inter-layer antiferromagnetic

interactions which involve two steps, namely canting of the ordered spins at Tc1 ¼ 87 K followed by

rearrangement of the canted spin states at Tc2 ¼ 13 K. Compound 5 exhibits 3-D ferromagnetic

ordering at Tc ¼ 83 K. Consequently, 4 and 5 behave as magnets although the origin of their

spontaneous magnetization is different. The bulk magnetic properties of 4 and 5 are in contrast to what

was observed in a similar layered compound [{Ru2(O2CCF3)4}2TCNQF4] (2; TCNQF4 ¼ tetrafluoro-

7,7,8,8-tetracyanoquino-dimethane). This compound is an antiferromagnet (at H ¼ 0) with TN ¼ 95 K

with the behavior being strongly dependent on the environment between the layers: in particular the

stacking motif can either be ‘‘in-registry’’ with an overlap between analogous groups (for 2) or ‘‘out-of-

registry’’ which involves a misalignment of similar groups between layers (for 4 and 5).

Introduction

The design of charge-transfer molecular networks using a care-

fully chosen set of electron-donor and electron-acceptor building

blocks is an efficient route for obtaining functional materials that

may exhibit interesting magnetic and/or conducting properties.

An important family of such materials are organometallic

charge-transfer salts of the type [M(C5Me5)2]+[TCNQ]�

(M ¼ Fe,1 Mn,2 Cr3) and [M(C5Me5)2]+[TCNE]� (M ¼ Fe,4

aDepartment of Chemistry, Graduate School of Science, TohokuUniversity, 6-3 Aramaki-Aza-Aoba, Aoba-ku, Sendai, Miyagi, 980-8578,Japan. E-mail: [email protected] of Chemistry, Texas A&M University, PO Box 30012,College Station, TX, 77842-3012, USA

† Electronic supplementary information (ESI) available: XRPD patterns(Fig. S1); temperature dependence of ac susceptibilities of dried samplesof 4 and 5 (Fig. S2 and S3). CCDC reference numbers 724296 and 724297.For ESI and crystallographic data in CIF or other electronic format seeDOI: 10.1039/b905486c

‡ Present address: Department of Chemistry, Faculty of Science,Kanagawa University, 2946 Tsuchiya, Hiratsuka, Kanagawa 259-1293,Japan.

This journal is ª The Royal Society of Chemistry 2009

Mn,5 Cr6), where C5Me5 is pentamethylcyclopentadienide (Cp*),

TCNQ is 7,7,8,8-tetracyano-p-quinodimethane, and TCNE is

tetracyanoethylene. These salts are known to exhibit bulk ferro-

or meta-magnetic behavior governed by inter-molecular inter-

actions between fully charge-transferred [M(C5Me5)2]+ units with

S ¼ 1/2 and TCNQ or TCNE mono-radicals. In these materials

which are molecular magnets, there are two requisite conditions

for the observation of magnetic behavior, namely: (1) the

generation of unpaired spins driven by one-electron (1-e�)

transfer from the donor (D; [M(C5Me5)2]) to the acceptor

(A; TCNQ0 or TCNE0) and (2) p–p stacking and/or van der

Waals interactions between D+ and A� molecules. The rationale

of the observed properties is that the 1 : 1 DA alternating array

leads to a charge-polarized state (i.e., ionic state) with localized

spins available for magnetic exchange coupling (Scheme 1a).

In a different scenario, a 2 : 1 stoichiometry ‘‘with the units

being essentially the same electronically’’ can lead to a charge-

delocalized state owing to a resonance of the type [D+–A��D 4

D–A–D 4 D–A��D+] in D2A systems (Scheme 1b) and vice

versa in A2D systems (vide infra). This concept is based on the

theory advanced for the well-known Creutz–Taube ion,

CrystEngComm, 2009, 11, 2121–2130 | 2121

Scheme 1

Scheme 2

[(H3N)5Ru(m-pyz)Ru(NH3)5]5+ (pyz ¼ pyrazine),7 which

experiences a charge-transfer resonance [RuIII–pyz–RuII 4 RuII–

pyz–RuIII] referred to as Class II in the Robin-Day system of mixed-

valence nomenclature.8 In this vein of research, early work by

Crutchley et al., involved the preparation of diruthenium(III)

complexes bridged by 1,4-dicyanamidebenzene dianion (Dicyd2�)

and its derivatives in which RuIII ion and Dicyd2� act as a 1-e�

acceptor and a 1-e� donor, respectively, to form an A2D system

that leads to the charge-transfer resonance of

[A�–D+–A 4 A–D–A 4 A–D+–A�].9 It is notable that these

compounds exhibit very strong intra-molecular magnetic coupling

between the RuIII S ¼ 1/2 spins with values of

J > 400 cm�1. Obviously, this behavior is a result of highly conju-

gated electron-transfer (in this case, hole-transfer) between Ru

ions via a pp HOMO of the Dicyd bridge based on the A2D charge-

transfer resonance. These compounds are discrete, however, and

it remains an open question as to what will be the resulting

bulk magnetic behavior if one expands the structures to include

multidimensional networks of D2A or A2D combinations. In this

vein, we are exploring chemistry that is aimed at the elaboration

of D2A or A2D materials that involve a 1-e� transfer of D / A.

The judicious choice of building blocks is very important as

a first principles step for a successful design strategy. To

accomplish the aforementioned goal of obtaining D2A networks,

we have chosen to use paddlewheel diruthenium(II, II)

complexes [Ru2] (1-e� donor) and TCNQ derivatives (as a 1-e�

acceptor, although they are capable of accepting a second elec-

tron if the reducing agent is sufficiently strong). The paddlewheel

diruthenium complexes were selected for this chemistry as they

are known to undergo a [Ru2II,II] / [Ru2

II,III] redox without

significant structural rearrangement.10 In addition, both redox

states of [Ru2II,II] and [Ru2

II,III] possess unpaired spins, viz.,

S ¼ 111 and S ¼ 3/2,12 respectively. Moreover, a 1-e� reduced

TCNQ with S¼ 1/2 will be present as well, thereby satisfying the

aforementioned design conditions for accessing magnetic mate-

rials. If one considers the situation from the point of view of

crystal engineering, the [Ru2] unit acts as a linear coordination-

acceptor building block with two axial coordination sites,13 and

the TCNQ derivatives act as a m4-coordination-donor building

block.14 The combination of these building blocks in the ratio

D/A ¼ 2 : 1 leads to neutral assemblies of various types,

including ladder-type chains,15 2-D networks,16,17 and infinite

3-D networks18 as depicted in Scheme 2.

2122 | CrystEngComm, 2009, 11, 2121–2130

In accordance with the stated goals, we have synthesized

the charge-transfer assemblies [{Ru2(O2CCF3)4}2TCNQ] (1)16

and [{Ru2(O2CCF3)4}2TCNQF4] (2; TCNQF4 ¼ tetrafluoro-

7,7,8,8-tetracyanoquinodimethane)17 which exhibit a 2-D

fishnet-like sheet structure with the D2A composition. In the

case of 2, a full 1-e� transfer from [Ru2] units to TCNQF4

has occurred and the system can be regarded overall as

[{Ru24.5+}–(TCNQF4

c�)–{Ru24.5+}]. This material exhibits

long-range antiferromagnetic ordering with TN ¼ 95 K

resulting from a combination of an intra-layer ferromagnetic

ordering and an inter-layer antiferromagnetic ordering

(so, antiferromagnet at H ¼ 0).17

To prepare new materials that exhibit a complete 1-e�

transfer as found for 2, we performed reactions with the

D/A set [Ru2II,II(O2CPh-x-F)4(THF)2] (x- ¼ o-, m-, p-;

x-F-PhCO2� ¼ x-fluorobenzoate) and bis[1,2,5]thiadizolote-

tracyanoquinodimethane (BTDA-TCNQ),19 in which the energy

levels of the HOMO in the former and the LUMO in the latter

are reversed from the corresponding levels for [Ru2(O2CCF3)4]

and general TCNQ derivatives.20 As a result of these attempts,

three new D2A charge-transfer compounds, [{Ru2(O2CPh-

x-F)4}2(BTDA-TCNQ)]$n(solv) (x ¼ m-, 3 (n(solv) ¼ 1.6

(4-chlorotoluene)$3.4CH2Cl2); o-, 4 (n(solv) ¼ 4CH2Cl2); p-, 5

(n(solv) ¼ 2CH2Cl2$2(4-chlorotoluene))), were obtained,

among which only 3 exhibits an infinite 3-D structure that

ferromagnetically orders at Tc ¼ 107 K (Scheme 2)18 in spite of

the fact that the general formula is identical to 4 and 5.

Compounds 4 and 5 are 2-D layered structures as is the case

for 1 and 2, but they exhibit different magnetic behavior which

is strongly dependent on inter-layer interactions. Compound 4

exhibits canted spins with a spontaneous magnetization at

Tc ¼ 87 K, i.e., it is a canted antiferromagnet triggered by inter-

layer antiferromagnetic ordering at TN ¼ 93 K. In contrast,

compound 5 is a 3-D ferromagnetically ordered material

(Tc ¼ 83 K). In this paper, the syntheses, structures, and

magnetic properties of 2-D layered compounds 4 and 5 are

discussed and compared to 2.

This journal is ª The Royal Society of Chemistry 2009

Table 1 Crystallographic data for 4 and 5

4 5

Formula C72H40Cl8N8O16

F8Ru4S2

C84H50Cl6N8O16

F8Ru4S2

Formula weight 2177.16 2260.47Crystal system Triclinic TriclinicSpace group P-1 P-1a/A 10.1901(17) 10.52920(10)b/A 14.441(2) 13.1690(5)c/A 14.560(2) 15.8886(2)a/� 77.697(6) 80.575(9)b/� 76.965(6) 79.368(7)g/� 73.843(5) 74.815(9)V/A3 1978.9(6) 2074.14(9)Z 1 1T/K 103 � 1 93 � 1Dcalc/g cm�3 1.827 1.810F000 1072.00 1120.00l/A 0.7107 0.7107m(Mo Ka)/cm�1 11.615 10.498Data measured 13460 14073Data unique 6803 7154Rint 0.038 0.048No. of variables 533 578GOF 1.067 1.063R1 (I > 2.00s(I))a 0.0539 0.0645R (all reflections)a 0.0806 0.0785wR2 (all reflections)b 0.1299 0.2097

a R1 ¼ R ¼ S||Fo| � |Fc||/S|Fo|. b wR2 ¼ [Sw(Fo2 � Fc

2)2/Sw(Fo2)2]1/2.

Experimental

General procedures

All synthetic procedures were performed anaerobically using

standard Schlenk-line techniques and a commercial glove box.

All chemicals were purchased from commercial sources and were

of reagent grade quality. Solvents were dried using common

drying agents and distilled under an ultrapure nitrogen gas

before use. The starting materials [Ru2(O2CPh-x-F)4(THF)2]

(x ¼ o-, p-) were prepared according to previously reported

methods,21 and BTDA-TCNQ was synthesized according to the

literature procedure.19

Preparation of [{Ru2(O2CPh-o-F)4}2(BTDA-TCNQ)]$4CH2Cl2 (4)

A CH2Cl2 solution (25 ml) of [Ru2(O2CPh-o-F)4(THF)2]

(180 mg, 0.20 mmol) was separated into 1 mL portions and

placed in narrow-diameter glass tubes (f: 8 mm). A 4-chlor-

otoluene solution (75 mL) of BTDA-TCNQ (37.5 mg,

0.10 mmol) was carefully placed in 3 mL portions onto each

CH2Cl2 layer, sealed and slow diffusion was allowed to occur.

The glass tubes were left to stand undisturbed for a week or more

to yield block-shaped dark-green crystals of 4. Yield: 43 mg,

20%. Elemental analysis (%) calcd for 4$3.5CH2Cl2C71.5H39Cl7F8N8O16Ru4S2: C 40.23, H 1.84, N 5.25. Found: C

40.20, H 1.92, N 5.48. IR (KBr): n(ChN), 2194, 2134 cm�1.

Preparation of [{Ru2(O2CPh-p-F)4}2(BTDA-TCNQ)]$2CH2Cl2$

2(4-chlorotoluene) (5)

Compound 5 was synthesized in a similar way as described for 4

with [Ru2(O2CPh-p-F)4(THF)2] (180 mg, 0.20 mmol) and

BTDA-TCNQ (37.5 mg, 0.10 mmol). Yield: 55 mg, 24%.

Elemental analysis (%) calcd for 5$0.7(4-chlorotoluene)

C72.9H36.9Cl0.7F8N8O16Ru4S2: C 45.46, H 1.93, N 5.82. Found: C

45.23, H 2.27, N 5.79. IR (KBr): n(ChN), 2181, 2129 cm�1.

Physical measurements

Infrared spectra were measured on KBr disks with a Jasco FT-IR

620 spectrophotometer. Magnetic susceptibility measurements

were conducted with a Quantum Design SQUID magnetometer

(MPMS-XL) in the temperature and dc field ranges of 1.8–300 K

and �7 to 7 T, respectively. Magnetic measurements of the ac

type were performed at various frequencies ranging from 1 to

1488 Hz with an ac field amplitude of 3 Oe. Polycrystalline

samples embedded in liquid paraffin were used for the

measurements to prevent torquing of crystals. Experimental data

were corrected for the sample holder and liquid paraffin and for

the diamagnetic contribution calculated from the Pascal

constants.22

Crystallographic analyses

Crystal data were collected on a Rigaku CCD diffractometer

(Saturn70) with graphite-monochromated Mo Ka radiation. A

single crystal of size of 0.09 � 0.05 � 0.03 mm for 4 and

0.20 � 0.10 � 0.10 mm for 5 was mounted on a thin-glass loop.

The structures were solved using direct methods (SIR97).23 The

non-hydrogen atoms were refined anisotropically, except for

This journal is ª The Royal Society of Chemistry 2009

some disordered solvent atoms which were refined isotropically;

hydrogen atoms were introduced as fixed contributors. Full-

matrix least-squares refinements on F2 converged with

unweighted and weighted agreement factors of R1 ¼ S||Fo|�|Fc||/

S|Fo| (I > 2.00s(I) and all data), and wR2 ¼ [Sw(Fo2-Fc

2)2/

Sw(Fo2)2]1/2 (all data). A Sheldrick weighting scheme was used.

All calculations were performed using the CrystalStructure

crystallographic software package.24 The crystal data and details

of the structure determination of 4 and 5 are summarized in

Table 1.

Results and discussion

Synthesis and infrared spectra

Compounds 4 and 5 were prepared by slow diffusion of solutions

of diruthenium complexes in CH2Cl2 (bottom layer) and BTDA-

TCNQ in 4-chlorotoluene (top layer) in narrow-diameter glass

tubes under an ultra-pure N2 atmosphere. Crystals of 4 and 5 are

somewhat fragile due to partial loss of interstitial solvents. For

measurements of physical properties, samples were harvested

immediately from fresh batches of the compound. The infrared

spectra of 4 exhibit two n(ChN) stretches at 2194 and 2134 cm�1,

which are shifted to lower energies than the corresponding

features for neutral BTDA-TCNQ (2220 cm�1). Although

n(ChN) modes should be affected by s-bonding of Ru )

NhC, the shift to lower energies indicates reduction of the

BTDA-TCNQ moiety resulting from charge-transfer of [Ru2] /

BTDA-TCNQ with possible p-back donation as was observed in

1 or Rh derivative.16 Such shifts in the n(ChN) stretches were

also observed for 5 (2181, 2129 cm�1), implying a similar

reduction state of BTDA-TCNQ in 4 and 5. Note that 3 with an

CrystEngComm, 2009, 11, 2121–2130 | 2123

infinite 3-D structure exhibits similar n(ChN) stretches at 2196

and 2134 cm�1.18

Description of the structures of 4 and 5

Compounds 4 and 5 crystallize in the triclinic space group P-1

(#2) with a formula unit comprising two unique [Ru2] units and

one BTDA-TCNQ molecule, each of which has an inversion

center on the midpoint of the Ru–Ru bond and the center of the

ring of BTDA-TCNQ (Z ¼ 1). Thermal ellipsoid plots of the

units in 4 and 5 are depicted in Fig. 1. In both 4 and 5, the BTDA-

TCNQ molecule coordinates to four [Ru2] units through the

cyano groups in a m4-bridging mode, the result of which is the

formation of a distorted hexagonal 2-D fishnet-like structure as

observed for 1 and 2 (Fig. 2).16,17 In this motif, the hexagonal net-

Fig. 1 Thermal ellipsoid plots of 4 (a) and 5 (b) (50% probability

ellipsoids; symmetry operations (*)�x,�y,�z; (**)�x + 1,�y + 1,�z�1;

(***)�x,�y + 1,�z; (#)�x,�y + 1,�z; (##)�x + 1,�y,�z + 1; (###)�x,

�y + 1, �z + 1). Hydrogen atoms and solvent molecules are omitted

for clarity.

2124 | CrystEngComm, 2009, 11, 2121–2130

ring is composed of [–{Ru(1)2}–(syn-BTDA-TCNQ)–{Ru(2)2}–

(cis-BTDA-TCNQ)–]2 as can be seen in the projection from the

a axis (syn- and cis-BTDA-TCNQ designations are used

for BTDA-TCNQ molecules bridging through the 7,8- and

7,7-cyano groups, respectively). The inter-layer distance was

estimated as 8.82 A for 4 and 9.02 A for 5 by defining a least-

squares plane for each layer. The nearest inter-layer [Ru2]/[Ru2]

or (BTDA-TCNQ)/(BTDA-TCNQ) distance corresponds to

the a axis distance with ca. 10.19 A for 4 and ca. 10.53 A for 5

(see Table 1) which are very similar. Conversely, for 2, in which

spin cancellation between layers is observed to occur, the inter-

layer distance is 6.6 A, and the nearest inter-layer [Ru2]/[Ru2] or

(TCNQF4)/(TCNQF4) distance is ca. 8.79 A (corresponding to

the c axis of the unit cell).17 These values are significantly shorter

than those found for 4 and 5, although it should be pointed out

that a p-xylene molecule of crystallization is packed between the

TCNQF4 moieties. Fig. 3 depicts a vertical view of the stacking

of the 2-D sheets in 2, 4 and 5. In 2, each 2-D sheet slides to the

lengthwise direction of the TCNQF4 molecule along the mirror

plane of the C2/m unit cell (i.e., the (101) direction), and conse-

quently, an ‘‘in-registry’’ stacking mode consisting of a requence

of evenly spaced molecular groups, namely [/TCNQF4/TCNQF4/]N is found (dotted circle in Fig. 3a).17 In 4 and 5,

however, each sheet slides diagonally, consequently, the BTDA-

TCNQ moieties in a sheet are closer to the [Ru2(2)] moieties in

neighboring sheets (‘‘out-of-registry’’ mode: dotted circles in

Figs. 3b and 3c): The inter-layer interatomic distance between the

midpoint of Ru(2)–Ru(2) bond and C(30) in BTDA-TCNQ is

8.87 A and 9.00 A for 4 and 5, respectively (the interlayer

distance between the least-square planes defined by an atomic set

of [Ru(1), Ru(1)***, Ru(2), Ru(2)***] for 4 or [Ru(1), Ru(1)###,

Ru(2), Ru(2)###] for 5 is 8.82 and 9.02 A, respectively; Fig. 2).

These differences in the structures of 2, 4, and 5 vis-�a-vis the inter-

layer distances and stacking modes affect their bulk magnetic

properties which are strongly dependent on these interactions

dictated by symmetry.

Local structures and the effect of charge-transfer

The bond distances in the [Ru2] units and BTDA-TCNQ are

good guides for the evaluation of the degree of charge-transfer

from [Ru2] units to BTDA-TCNQ. The Ru–Ru bond distances in

4 and 5 are Ru(1)–Ru(1)* ¼ 2.2787(6) A and Ru(2)–Ru(2)** ¼2.2807(6) A and Ru(1)–Ru(1)#¼ 2.2886(7) A and Ru(2)–Ru(2)##

¼ 2.2890(6) A, respectively (symmetry operations: *,�x,�y,�z;

**, �x + 1, �y + 1, �z � 1; #, �x, �y + 1, �z; ##, �x + 1, �y,

�z + 1). It is difficult, however, to assess the degree of charge-

transfer emanating only from the Ru–Ru bond because this bond

distance, as judged by literature reports, differs only slightly for

[Ru2II,II]0 versus [Ru2

II,III]+ and is affected strongly only by

s-donation from axial ligands.10 The Ru–Oeq (Oeq¼ carboxylate

oxygen) bond distances are more sensitive to the oxidation state,

and are generally found in the range of 2.07–2.09 A for [Ru2II,II]

and 2.01–2.03 A for [Ru2II,III].10 The average Ru–Oeq distances

for the [Ru(1)2] and [Ru(2)2] units are 2.019 and 2.056 A in 4 and

2.022 and 2.070 A in 5, respectively, which indicate that the

[Ru(1)2] and [Ru(2)2] units are likely to be [Ru2II,III] and [Ru2

II,II].

Essentially, a 1-e� charge-transfer from one of the two [Ru2] units

to BTDA-TCNQ occurs to induce a charge-polarized state in

This journal is ª The Royal Society of Chemistry 2009

Fig. 2 Packing diagrams of 4 (a) and 5 (b), in which the equatorial RCO2� ligands are located around the [Ru2] center and solvent molecules are omitted

for the sake of clarity. The noted distances with no caption represent the distances between [Ru2] units, and the interlayer distance means a distance

between the least-squares planes defined by an atomic set of [Ru(1), Ru(1)*, Ru(2), Ru(2)#].

Fig. 3 Vertical views of stacking 2-D layers in 2 (a), 4 (b), and 5 (c), in which the RCO2� ligands around the [Ru2] center and the solvent molecules have

been omitted for the sake of clarity. The dotted circle represents the characteristic groups located in the nearest position between layers, TCNQF4/TCNQF4 for 2 and [Ru2]/BTDA-TCNQ for 4 and 5.

both compounds. The Ru–Nax (Nax ¼ cyano nitrogen of BTDA-

TCNQ) distances are Ru(1)–N(1) ¼ 2.232(5) A and Ru(2)–N(2)

¼ 2.280(4) A for 4 and Ru(1)–N(1)¼ 2.244(5) A and Ru(2)–N(2)

¼ 2.291(5) A for 5, which also reflects the respective oxidation

states of each [Ru2] unit; those for [Ru2II,II] have a tendency to be

longer than those for [Ru2II,III].11e,16,25

The oxidation state of the BTDA-TCNQ moiety can be eval-

uated by a comparison of the bond distances in 4 and 5 to those

of the neutral analogue (BTDA-TCNQ0) and the fully-1-

e��transferred derivative (BTDA-TCNQc�) based on the

Kistenmacher relationship26 commonly used for TCNQ

This journal is ª The Royal Society of Chemistry 2009

derivatives and their assemblies.16–18,27 The constituent bond

distances in the BTDA-TCNQ moiety in 4 and 5 are listed in

Table 2 together with those of BTDA-TCNQ compounds repor-

ted previously.28–30 The modification of bond distances upon

reduction from the neutral state to the anionic state has significant

effects on the bonds b, c, d, and e (Figure in Table 2); the trend is

that b and d become shorter and c and e are lengthened with the

c distance being particularly sensitive. An examination of

these relevant bond distances was undertaken and the oxidation

state of BTDA-TCNQ moiety was evaluated by three

relationships: rc ¼ A1c + B1, rc/d ¼ A2(c/d) + B2, and

CrystEngComm, 2009, 11, 2121–2130 | 2125

Table 2 Comparison of bond distances (A) in BTDA-TCNQa

Compound Charge a b c d e f g rc rc/d rc/(b+d) Ref.

I 0 1.126(5) 1.444(6) 1.351(7) 1.464(6) 1.421(5) 1.330(5) 1.626(4) 0 0 0 19

II 0 1.136(4) 1.442(3) 1.358(3) 1.462(3) 1.426(3) 1.338(3) 1.616(2) �0.18 �0.13 �0.13 28

III 0 1.135(8) 1.445(7) 1.362(7) 1.463(7) 1.423(8) 1.338(7) 1.617(5) �0.28 �0.18 �0.17 29

IV �0.5 1.127(10) 1.422(10) 1.379(8) 1.449(8) 1.422(8) 1.340(8) 1.614(4) �0.72 �0.65 �0.77 30

V �0.5 1.128(14) 1.443(13) 1.372(12) 1.454(12) 1.428(12) 1.336(12) 1.615(8) �0.54 �0.48 �0.40 30

�0.5 1.135(14) 1.430(14) 1.381(12) 1.452(12) 1.434(12) 1.338(12) 1.615(9) �0.77 �0.64 �0.71 30

VI �0.9 � 0.2 1.142(5) 1.434(5) 1.387(5) 1.458(5) 1.431(5) 1.333(5) 1.611(4) �0.92 �0.64 �0.73 29

VII -1 1.140(6) 1.429(7) 1.390(6) 1.437(6) 1.433(5) 1.346(6) 1.614(4) �1 �1 �1 30

3 1.135(8) 1.416(8) 1.407(9) 1.425(8) 1.434(9) 1.330(8) 1.601(7) �1.44 �1.31c �1.45c 18

1.139(8) 1.415(9) 1.444(8) 1.335(9) 1.622(7)1.137b 1.416b 1.435 1.333b 1.612b

4 1.134(7) 1.437(8) 1.407(8) 1.431(10) 1.441(8) 1.335(9) 1.609(6) �1.44 �1.21c �1.31c This Work1.156(8) 1.411(9) 1.450(8) 1.338(8) 1.610(5)1.145b 1.424b 1.441 1.337b 1.610b

5 1.140(8) 1.419(8) 1.403(8) 1.420(8) 1.446(9) 1.348(7) 1.598(5) �1.33 �1.25c �1.30c This Work1.152(7) 1.428(8) 1.448(8) 1.353(7) 1.616(5)1.146b 1.424b 1.434 1.351b 1.607b

a I: BTDA-TCNQ; II: (TTF)(BTDA-TCNQ) (TTF ¼ tetrathiafulvalene); III: (TSeN)(BTDA-TCNQ)(C6H5Cl) (TSeN ¼ naphthaceno[5,6-cd : 11,12-c0d0]bis[1,2]diselenole); IV: [NMe(Bt)3](BTDA-TCNQ)2; V: (NEt4)(BTDA-TCNQ)2(BTDA-TCNQ); VI: (TSeN)(BTDA-TCNQ)(C6H5Cl); VII:[NEt(Me)3](BTDA-TCNQ). b Average values. c Estimated from the average values.

Fig. 4 Temperature dependence of c and cT of 4 (a) and 5 (b) measured

at 1 kOe.

rc/(b+d) ¼ A3[c/(b + d)] + B3 (Kistenmacher relationship).26 The

constants were obtained by applying the relationships to neutral

BTDA-TCNQ (r ¼ 0)19 and [NEt(Me)3](BTDA-TCNQ)

(r ¼ �1)30 with A1 ¼ �25.64, B1 ¼ 34.64, A2 ¼ �22.73,

B2¼ 20.98, A3¼�50.00 and B3¼ 23.25. The estimated values are

rc ¼ �1.435, rc/d ¼ �1.214, and rc/(b+d) ¼ �1.305 for 4 and

rc ¼ �1.333, rc/d ¼ �1.250, and rc/(b+d) ¼ �1.295 for 5, whi-

ch are greater than exactly �1 (d ¼ 0.214–0.435 for 4 and

d ¼ 0.250–0.333 for 5). We point out that the values estimated for

other compounds (Table 2) are typically +0.1 � +0.3 larger than

the expected value. In the case of non-charge-transferred

compounds of [Ru2II,II]2TCNQ (1) and for [Rh2

II,II]2TCNQ, the

values deviate from the expected value of r ¼ 0 on the order of

�0.6.16 Hence, the BTDA-TCNQ moiety in both 4 and 5 can be

reasonably assigned as a fully reduced monoanion, but it may be

the case that it is slightly more reduced, i.e., [BTDA-TCNQ(1+d)�]

(d z 0 � 0.4), which would imply a non-integer charge-

polarized state of [Ru(1)25+]–[BTDA-TCNQ(1+d)�]–[Ru(2)2

(4+d)+]

(d z 0 � 0.4). This trend has also been observed for the 3-D

compound 3 with an identical formula to 4 and 5, and the

observed charge-polarized state may be a result of the structural

symmetry; the space group is P-1 with Z ¼ 1 for 4 and 5 and

C2/c with Z ¼ 4 for 3,18 whereas it is C2/m with Z ¼ 2 for 2.17

Magnetic susceptibility measurements for 4 and 5

Field-cooled dc magnetic susceptibility (FCM) data were

collected on freshly-harvested polycrystalline samples of 4 and 5

2126 | CrystEngComm, 2009, 11, 2121–2130 This journal is ª The Royal Society of Chemistry 2009

suspended in Nujol oil in the temperature range of 1.8–300 K at

1 kOe. The plots of c and cT as a function of temperature are

depicted in Fig. 4. The cT value of 3.54 cm3 K mol�1 for 4 and

2.96 cm3 K mol�1 for 5 at 300 K is much higher than that

expected from the spin-only value 2.00 cm3 K mol�1 for a set of

two S ¼ 1 spins with g ¼ 2.00 for isolated [Ru2II,II] units as

calculated according to no redox reaction, an indication that the

charge-transfer from the [Ru2II,II] unit to the BTDA-TCNQ

moiety occurs in both 4 and 5 and that the resulting magnetic

centers are significantly coupled through the BTDA-TCNQc� S

¼ 1/2 centers even at high temperatures. Upon decreasing the

temperature, the cT value gradually increases (33.5 cm3 K mol�1

at 101 K for 4 and 29.3 cm3 K mol�1 at 96 K for 5) and then

drastically increases to reach 296 cm3 K mol�1 at 82 K for 4 and

349 cm3 K mol�1 at 72 K for 5 followed by a decrease at 1.8 K to

8.57 cm3 K mol�1 for 4 and 11.0 cm3 K mol�1 for 5. The c value

(at 1 kOe) for both compounds exhibits a rapid increase

at� 90–100 K as expected from their cT behavior and a stepwise

small increase at �20 K for 4 without any decrease being

observed over the entire temperature range. The observation of

a rapid increase in c and a considerably high cT value for the

maximum indicates the onset of long-range ordering at

� 80–100 K for both 4 and 5.

Fig. 5 Temperature dependence of the ac magnetic susceptibilities c0 (in-

phase) and c00 (out-of-phase) at zero dc field and 3 Oe ac oscillating field

for 4 (a) and 5 (b).

This journal is ª The Royal Society of Chemistry 2009

In order to gain more information about the magnetic tran-

sitions, temperature dependence ac magnetic susceptibility

(c0, real part; c00, imaginary part) data were measured under zero

dc field and a 3 Oe oscillating field at several ac frequencies

(Fig. 5). In the case of 4, the c0 response shows three distinct

peaks at 93, 87, and 13 K upon cooling, whereas the c00 signal

only responds to the two c0 peaks at 87 and 13 K without

a distinct frequency dependence (Fig. 5a). Note that weak c00

signals responding at high frequencies are observed at ca. 92 K,

which are probably associated with the creation/movement of

antiferromagnetic domains that are fixed at Tc1 ¼ 87 K (vide

infra). Therefore, the c0 peak at 93 K implies a long-range anti-

ferromagnetic ordering and is assigned to a N�eel temperature

with TN ¼ 93 K. Antiferromagnetic ordering is a consequence of

antiferromagnetic interactions between ferromagnetically-

ordered layers as observed for 2.17 The other peaks (87 and 13 K)

involving c00 signals indicate the involvement of two types of

phase transitions with spontaneous magnetizations; the first

phase transition at Tc1 ¼ 87 K is due to the onset of a new

spin-canted state following the antiferromagnetic ordering at

TN¼ 93 K, and the second phase transition at Tc2¼ 13 K is most

likely a result of a rearrangement of ordered spins in the estab-

lishment of domain walls. Indeed, the c00 value observed in the

range between Tc1 and Tc2 tends to show non-zero signals, albeit

weak, as a result of the dynamic behavior of domains of various

sizes.

Conversely, the c0 value for 5 exhibits a single distinct peak at

83 K along with c00 signals without frequency dependence for the

maximum (Fig. 5b), suggesting the onset of long-range ferro-

magnetic order (Tc ¼ 83 K). At temperatures below ca. 80 K

(to 10 K), however, both c0 and c00 display multi-relaxation

processes roughly confirmed as mode-I at around 80–50 K and

mode-II at 50–10 K. These slow relaxation processes of the

magnetization can be associated with the dynamic behavior of

various domain sizes, and the freezing of the magnetization at

mode-II may be attributed to the fixing of domain walls, as was

observed for 4 (Fig. 4). Thus, although 4 and 5 are of opposite

nature in terms of their inter-layer interactions, i.e., antiferro-

magnetic and ferromagnetic, respectively, their bulk magnetic

behavior induced by spontaneous magnetizations (originated

from spin canting in 4 and ferromagnetically ordered spins in 5)

at low temperatures is similar to each other, but completely

different from that of 2, a reflection of their particular inter-layer

stacking arrangements.

Investigation of the antiferromagnetic phase for 4

Compound 4 exists in an antiferromagnetic phase at TN ¼ 93 K

under H ¼ 0, whereas 5 behaves as a ferromagnet at low

temperatures below Tc ¼ 83 K as detailed in the aforementioned

section. To define the antiferromagnetic phase in 4 as a function

of temperature and dc field, the following two types of magnetic

measurements were further performed: FCM measurements in

low applied fields and magnetization measurements as a function

of field at several temperatures. Fig. 6 depicts FCM curves for 4

measured at 10, 100, and 500 Oe. The magnetization curves

exhibit a peak near TN and then a two-step increase at Tc1 and

Tc2 when fields of 10 and 100 Oe are applied, but no longer

display a feature indicative of a transition to the

CrystEngComm, 2009, 11, 2121–2130 | 2127

Fig. 7 Field dependence of the initial magnetization of 4 measured at

several temperatures between 1.8–100 K.

Fig. 8 H–T phase diagram of 4, where P and AF stand for the para-

magnetic and antiferromagnetic phases, respectively, and TN (¼ 93 K)

was determined from the c0 vs T data at H¼ 0 (Fig. 5). The dashed line is

a guide for the eye.

Fig. 6 Field-cooled magnetization curves of 4 measured under dc fields

of 10, 100, and 500 Oe.

Fig. 9 Field dependence of the magnetization of 4 (a) and 5 (b) at several

temperatures between 1.8–100 K. Inset: temperature dependence of the

remnant magnetization at H ¼ 0 and the coercive field Hc.

antiferromagnetic phase under 500 Oe dc field. This behavior is

evidence of metamagnetic behavior, and is not observed in 5 at

any field. Fig. 7 shows the initial M–H curves for 4 measured at

several temperatures up to 100 K. Except for the data at 90 and

100 K, the magnetization rapidly increases at low fields up to

2128 | CrystEngComm, 2009, 11, 2121–2130

2 kOe, which is due to the spontaneous magnetization observed

below Tc1 (the behavior of the magnetization at 1.8 and 5 K at

low fields (H� 1 kOe) is associated with the anisotropic nature of

the ordered spins). The magnetization curve at 1.8 K reveals a

stepwise (or sigmoidal) increase with an inflection point at �1 T,

which shifts to lower fields with increasing temperature. This

inflection event is due to a spin flip of antiferromagnetically-

coupled intra-layer-ordered spins consistent with what was

observed in the M vs T data, and these inflection points and the

TN value define a boundary between the antiferromagnetic (AF)

phase and paramagnetic (P) phase defined by respective fields/

temperatures. Ultimately, the data lead to an H–T phase diagram

for 4 as shown in Fig. 8. It should be noted that the spin-flip field

(H1) tends to increase abruptly at temperatures below �20 K.

This fact suggests that the domain structure over this tempera-

ture range is different from that above 20 K. This hypothesis is

consistent with what we concluded earlier, namely that the

rearrangement of ordered spins in a domain occurs at Tc2

(i.e., even after the rearrangement, the canting state is preserved).

This journal is ª The Royal Society of Chemistry 2009

Field-dependence of the magnetization

Field dependence data of the magnetization for 4 and 5 were

measured in the range of �7 to 7 T at various temperatures up to

100 K (Fig. 9). In the high field region of 2–7 T, magnetization

linearly increases to reach 2.21 mB for 4 and 1.89 mB for 5 (at 7 T)

at 1.8 K, but does not saturate even at 7 T. This behavior is

generally observed for magnetic materials containing [Ru2II,II] or

[Ru2II,III] units,17,18,31,32 the origin of which is strong intrinsic

anisotropy of the constituent [Ru2] units (D z 270 cm�1 for

[Ru2II,II] with S ¼ 1 and D z 70 cm�1 for [Ru2

II,III] with S ¼ 3/2)

as well as structural anisotropy as expected for 2-D layered

magnets in which spin-canting can occur in the layer. The field

sweep between 7 and �7 T reveals the magnetization hysteresis:

the temperature dependence of the coercive field and remnant

magnetization is plotted in the inset of Fig. 9. The coercivity for

both compounds quasi-exponentially decreases with increasing

temperature and finally disappears at ca 80–90 K corresponding

to Tc1 for 4 and Tc for 5. The remnant magnetization for 4,

measured after a sweep of H ¼ 0 / 7 T / 0, rapidly decreases

upon reaching T > 20 K, while the value for 5 begins to decrease

rapidly at T > 60 K. These results indicate that the domains in 4

are ‘‘softer’’ than those in 5, which would display the difference

between the canting antiferromagnet-domains and the ferro-

magnet-domains, respectively. The inflection point at �20 K in 4

is close to the value Tc2, at which temperature the softer domains

rearrange to form harder domains (see Fig. 5). Indeed, another

inflection point is observed at less than 1 kOe, which indicates the

modification of the distribution of domains (Weiss domains)

generally found in hard magnets, is detectable at temperatures

below Tc2 (Fig. 7).

Concluding remarks

Two charge-transfer compounds whose structures consist of

fishnet-like 2-D networks were prepared from a 2 : 1 ratio of

paddlewheel-type diruthenium complexes ([Ru2II,II]) with mono-

fluorine-substituted benzoate ligands (o-F-PhCO2�: 4 or p-F-

PhCO2�: 5) and BTDA-TCNQ. A 3-D network compound (3)

forms in the identical [Ru2]/(BTDA-TCNQ) formulation

ratio with [Ru2II,II] units and m-F-PhCO2

� ligands. These new

hybrid [Ru2II,II]/BTDA-TCNQ materials undergo charge-

transfer from the [Ru2II,II] units to BTDA-TCNQ to produce

a charge-polarized state of the type [Ru(1)25+]–[BTDA-

TCNQ(1+d)�]–[Ru(2)2(4+d)+] (d z 0 � 0.5). The 1-e� transferred

BTDA-TCNQc� unit is S ¼ 1/2 and serves as an excellent

pathway for magnetic communication with [Ru2II,II] (S ¼ 1) and/

or [Ru2II,III] (S ¼ 3/2) building blocks as evidenced by the

observation of magnetic ordering with spontaneous magnetiza-

tion. Compound 3 is the only member of the series to exhibit

ferromagnetic ordering which occurs at Tc ¼ 107 K, a reflection

of the fact that it is a 3-D network consisting of conjugated

[Ru2]–NhCBTDA-TCNQ interactions. The high Curie temperature

is an excellent indication that the magnetic exchange interaction

between the [Ru2] centers via BTDA-TCNQc� is quite strong.18

The bulk magnetic properties of 4 and 5 were found to be

strongly dependent on the environment between the layers which

directly influences inter-layer magnetic interactions between

ferromagnetically-ordered layers. A related 2-D compound,

This journal is ª The Royal Society of Chemistry 2009

[{Ru2(O2CCF3)4}2(TCNQF4)] (2), exhibits antiferromagnetic

ordering with TN ¼ 95 K, and is an antiferromagnet at H ¼ 0.17

Similarly, 4 exhibits an antiferromagnetic transition at TN ¼ 93

K, but, in this case, the origin is a freezing of a spin-canted state

involving a spontaneous magnetization at Tc1 ¼ 87 K to become

a canted antiferromagnet. The canted state in a domain is rear-

ranged and fixed at Tc2 ¼ 13 K to form a harder domain struc-

ture. In contrast, 5 displays only ferromagnetic ordering at

Tc ¼ 83 K, an indication that ferromagnetic interlayer interac-

tions are dominant. The characteristic structural features of these

compounds are governed by the stacking mode of the 2-D layers:

In 2, the TCNQF4 moieties show considerable overlap between

layers (i.e., in-registry stacking), whereas in 4 and 5, the BTDA-

TCNQ moieties are misaligned (i.e., out-of-registry) and, instead,

are closer to [Ru2] units rather than to each other. Moreover,

slight differences in the out-of-registry stacking mode and inter-

layer distances lead to magnetic variations between 4 and 5. It is

necessary to point out also that magnetic properties of 4 and 5

are strongly affected by the history of the samples. If dried

samples are used instead of fresh samples from the mother liquid,

the magnetic phase and the transition temperature are drastically

altered; the temperature dependence of ac susceptibility data of

dried samples (4-dry and 5-dry) with their XRPD patterns are

provided in the supporting information (Fig. S1–S3).†

The present study demonstrates that mixed [Ru2]/TCNQ

systems exhibit rich magnetic behavior with relatively high Tc or

TN values for 2-D layered compounds of this type. To date,

however, we have no evidence of magnetic properties being

strongly associated with a charge-transfer resonance. If this

situation can be induced, higher Tc values and even room

temperature values or higher may be anticipated. Perhaps, even

more exciting is the promise of synergy between magnetic

ordering and electrical conductivity in such systems. Ongoing

work in our laboratories is aimed at these important goals.

Acknowledgements

This work was financially supported by The Asahi Glass Foun-

dation and in part by the CREST project, Japan Science and

Technology Agency (JST). N.M. thanks the JSPS Research

Fellowships for Young Scientists for financial support. K.R.D.

thanks the National Science Foundation and the Department of

Energy (DE-FG02-02ER45999) for support of this project at

Texas A&M University.

References

1 (a) G. A. Candela, L. J. Swartzendruber, J. S. Miller and M. J. Rice,J. Am. Chem. Soc., 1979, 101, 2755; (b) J. S. Miller, A. H. Reis, Jr.,E. Gebert, J. J. Ritsko, W. R. Salaneck, L. Kovnat, T. W. Capeand R. P. Van Duyne, J. Am. Chem. Soc., 1979, 101, 7111.

2 W. E. Broderick, J. A. Thompson, E. P. Day and B. M. Hoffman,Science, 1990, 249, 401.

3 W. E. Broderick and B. M. Hoffman, J. Am. Chem. Soc., 1991, 113,6334.

4 J. S. Miller, J. C. Calabrese, H. Rommelmann, S. R. Chittipeddi,J. H. Zhang, W. M. Reiff and A. J. Epstein, J. Am. Chem. Soc.,1987, 109, 769.

5 G. T. Yee, J. M. Manriquez, D. A. Dixon, R. S. McLean,D. M. Groski, R. B. Flippen, K. S. Narayan, A. J. Epstein andJ. S. Miller, Adv. Mater., 1991, 3, 309.

6 (a) J. S. Miller, R. S. McLean, C. Vazquez, J. C. Calabrese, F. Zuoand A. J. Epstein, J. Mater. Chem., 1993, 3, 215; (b)

CrystEngComm, 2009, 11, 2121–2130 | 2129

D. M. Eichhorn, D. C. Skee, W. E. Broderick and B. M. Hoffman,Inorg. Chem., 1993, 32, 491.

7 C. Creutz and H. Taube, J. Am. Chem. Soc., 1969, 91, 3988.8 M. Robin and P. Day, Adv. Inorg. Chem., 1968, 10, 247.9 M. A. S. Aquino, F. L. Lee, E. J. Gabe, C. Bensimon, J. E. Greedan

and R. J. Crutchley, J. Am. Chem. Soc., 1992, 114, 5130.10 (a) F. A. Cotton, R. A. Walton, Multiple Bonds Between Metal Atoms,

Oxford University Press, Oxford, 2nd edn, 1993; (b) F. A. Cotton,C. A. Murillo, R. A. Walton, ed. Multiple Bonds Between MetalAtoms, Springer Science and Business Media, Inc., New York, 3rdedn, 2005.

11 (a) F. A. Cotton, V. M. Miskowski and B. Zhang, J. Am. Chem. Soc.,1989, 111, 6177; (b) P. Maldivi, A.-M. Giroud-Godquin, J.-C. Marchon, D. Guillon and A. Skoulios, Chem. Phys. Lett., 1989,157, 552; (c) L. Bonnet, F. D. Cukiernik, P. Maldivi, A.-M. Giroud-Godquin, J.-C. Marchon, M. Ibn-Elhaj, D. Guillon andA. Skoulios, Chem. Mater., 1994, 6, 31; (d) A. Cogne, E. Belorizky,J. Laugier and P. Rey, Inorg. Chem., 1994, 33, 3364; (e)H. Miyasaka, R. Cl�erac, C. S. Campos-Fern�andez andK. R. Dunbar, J. Chem. Soc., Dalton Trans., 2001, 858; (f)S. Furukawa, M. Ohba and S. Kitagawa, Chem. Commun., 2005,865; (g) E. V. Dikarev, A. S. Filatov, R. Cl�erac andM. A. Petrukhina, Inorg. Chem., 2006, 45, 744.

12 M. A. S. Aquino, Coord. Chem. Rev., 1998, 170, 141.13 (a) M. A. S. Aquino, Coord. Chem. Rev., 2004, 248, 1025; (b)

M. Mikuriya, D. Yoshioka and M. Handa, Coord. Chem. Rev.,2006, 250, 2194.

14 (a) W. Kaim and M. Moscherosch, Coord. Chem. Rev., 1994, 129, 157;(b) K. R. Dunbar, Angew. Chem., Int. Ed. Engl., 1996, 35, 1659.

15 N. Motokawa, T. Oyama, S. Matsunaga, H. Miyasaka, K. Sugimoto,M. Yamashita, N. Lopez and K. R. Dunbar, Dalton Trans., 2008,4099.

16 H. Miyasaka, C. S. Campos-Fern�andez, R. Cl�erac and K. R. Dunbar,Angew. Chem., Int. Ed., 2000, 39, 3831.

17 H. Miyasaka, T. Izawa, N. Takahashi, M. Yamashita andK. R. Dunbar, J. Am. Chem. Soc., 2006, 128, 11358.

18 N. Motokawa, H. Miyasaka, M. Yamashita and K. R. Dunbar,Angew. Chem., Int. Ed., 2008, 47, 7760.

19 T. Suzuki, H. Fujii, Y. Yamashita, C. Kabuto, S. Tanaka,M. Harasawa, T. Mukai and T. Miyashi, J. Am. Chem. Soc., 1992,114, 3034.

2130 | CrystEngComm, 2009, 11, 2121–2130

20 In order to know preliminarily the possibility of occurring charge-transfer, we have confirmed the energy levels of HOMO (SOMO)on [Ru2

II,II(O2CR)4(THF)2] and LUMO on TCNQ derivativesincluding BTDA-TCNQ by calculations using Gaussian03 witha UB3LYP/LANL2DZ basis function (unpublished results), andobtained a gap between d*-HOMO and LUMO levels, +0.37 eV fora set of [Ru2

II,II(O2CCF3)4(THF)2]/TCNQF4, and �0.52 eV and�0.11 eV for sets of [Ru2

II,II(O2CPh-x-F)4] (x ¼ o-, p-)/BTDA-TCNQ, respectively, which suggested the easiness of charge-transferin the latter cases.

21 S. Furukawa and S. Kitagawa, Inorg. Chem., 2004, 43, 6464.22 E. A. Boudreaux, L. N. Mulay, Theory and Applications of

Molecular Paramagnetism, John Wiley and Sons, New York(1976) 491.

23 A. Altomare, M. Burla, M. Camballi, G. Cascarano, C. Giacovazzo,A. Guagliardi, A. Moliterni, G. Polidori and R. Spagna, J. Appl.Crystallogr., 1999, 32, 115.

24 CrystalStructure 3.8: Crystal Structure Analysis Package, Rigaku andRigaku Americas (2000–2007). 9009 New Trails Dr. The WoodlandsTX 77381 USA.

25 (a) H. Miyasaka, R. Cl�erac, C. S. C. Fern�andez and K. R. Dunbar,Inorg. Chem., 2001, 40, 1663; (b) M. Handa, D. Yoshida,Y. Sayama, K. Shiomi, M. Mikuriya, I. Hiromitsu and K. Kasuga,Chem. Lett., 1999, 1033.

26 T. J. Kintenmacher, T. J. Emge, A. N. Bloch and D. O. Cowan, ActaCrystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1982, 38,1193.

27 N. Lopez, H. Zhao, A. V. Prosvirin, A. Chouai, M. Shatruk andK. R. Dunbar, Chem. Commun., 2007, 4611.

28 T. Suzuki, C. Kabuto, T. Yamashita and T. Mukai, Bull. Chem. Soc.Jpn., 1987, 60, 2111.

29 K. Iwasaki, A. Ugawa, A. Kawamoto, Y. Yamashita,K. Yakushi, T. Suzuki and T. Miyashi, Bull. Chem. Soc. Jpn.,1992, 65, 3350.

30 T. Suzuki, C. Kabuto, Y. Yamashita, T. Mukai, T. Miyashi andG. Saito, Bull. Chem. Soc. Jpn., 1988, 61, 483.

31 (a) Y. Liao, W. W. Shum and J. S. Miller, J. Am. Chem. Soc., 2002,124, 9336; (b) T. E. Vos, Y. Liao, W. W. Shum, J.-H. Her,P. W. Stephens, W. M. Reiff and J. S. Miller, J. Am. Chem. Soc.,2004, 126, 11630.

32 T. E. Vos and J. S. Miller, Angew. Chem., Int. Ed., 2005, 44, 2416.

This journal is ª The Royal Society of Chemistry 2009