Synthesis and characterization of radical cations derived from mono- And biferrocenyl-substituted 2-Aza-1,3-butadienes: A study of the influence of an asymmetric and oxidizable bridge on intramolecular electron transfer

FULL PAPER

Synthesis and Characterization of Radical Cations Derived from Mono- and

Biferrocenyl-Substituted 2-Aza-1,3-butadienes: A Study of the Influence of an

Asymmetric and Oxidizable Bridge on Intramolecular Electron Transfer

Vega Lloveras,^[a]Antonio Caballero,^[b]Alberto Tárraga,^[b]M. Desamparados Velasco,^[b]

Arturo Espinosa,^[b]Klaus Wurst,^[c]David J. Evans,^[d]José Vidal-Gancedo,^[a]

Concepció Rovira,^[a]Pedro Molina,*^[b]and Jaume Veciana*^[a]

Keywords: Charge transfer / Electrochemistry / Electron transfer / Metallocenes / Optical spectroscopy

The synthesis and study of structural and electronic proper-

ties of mono-ferrocenyl π-conjugated complexes 5a–d, whose

electronic characteristics have been systematically varied by

introducing an electron-donating or electron-withdrawing

substituent either at the 1-position or at the 4-position of the

2-aza-1,3-butadiene moiety linked to the ferrocenyl unit, are

presented. The structural and electronic properties of the

homobimetallic complex 5f, with two ferrocene units linked

through the asymmetric and oxidizable 2-aza-1,3-butadiene

bridge, is also reported. The crystal structures of complexes

5b, 5d, and 5f show a large degree of conjugation in this

family of compounds. Complexes 5 show a rich electrochemi-

cal behavior due both to the oxidation of ferrocenyl units and

the 2-aza-1,3-butadiene bridge, as revealed by cyclic voltam-

ric oxidations following their generation by absorption spec-

troscopy. The electronic properties of all reported neutral and

oxidized π-conjugated complexes have been investigated by

means of UV/Vis–near-IR, EPR and ⁵⁷Fe Mössbauer spec-

troscopy. The detailed study of mono-oxidized species 5a^+·–

5f^+·has permitted the determination of the influence of an

asymmetric bridge with an electroactive character on the in-

tramolecular electron transfer (IET) phenomenon, thus dem-

onstrating that the 2-aza-1,3-butadiene bridge promotes the

IET between the two metallic units of 5f^+·through two dif-

ferent pathways. The experimental data and conclusions are

supported by DFT computations (B3LYP/3-21G*) and time-

dependent DFT methods.

metry. Radical cations 5^+·were prepared from 5 by coulomet- Germany, 2005)

areas of metallocene chemistry is the study of general stra-

Introduction

tegies for assembling two or more such electroactive metal

fragments in close proximity in order to determine the type

of electronic interactions between the metal fragments.^[3]

Among the various possibilities to bridge two cyclopen-

tadienyl-type (Cp) ligands, emphasis has been laid on the

electronic and steric properties of organic bridges because

both will influence the interaction between the metal frag-

ments bound to the Cp’s and, in many cases, these factors

will not be independent.

It is worth noting that IET can be monitored by the

study of optical transitions occurring in mixed-valence

complexes,^[4]and that intervalence charge-transfer (IV-CT)

bands may be observed in both σ- and π-bridged systems.

In the former case, a through-space mechanism is generally

Currently, there is considerable interest in the study of

the properties of compounds bearing multiple redox centers

separated by an organic bridge and, in particular, those that

are able to produce mixed-valence species.^[1]Such species

are interesting not only because of their importance in pro-

ducing new valuable materials, such as molecular wires and

switches, that show intramolecular electron transfer (IET)

phenomena, but also to understand the role of biologically

relevant mixed-valence compounds.^[2]One of the richer

[a] Institut de Ciència de Materials de Barcelona (CSIC), Campus

Universitari de Bellaterra,

08193, Cerdanyola, Spain

Fax: +34-93-580-5729

E-mail: vecianaj@icmab.es

[b] Universidad de Murcia, Departamento de Química Orgánica, believed to be responsible, such that low-energy bands at

Facultad de Química,

the near-IR are usually only observed in species that are

Campus de Espinardo, 30100 Murcia, Spain

sufficiently flexible to allow the two metal centers to come

into close proximity. In π-bridged systems, near-IR bands

are much more widely observed, even when the metals are

well separated by rigid bridges, thereby suggesting that

through-bond mechanisms are involved. IET in such π-

bridged systems may proceed either by a superexchange

and/or a hopping mechanism.^[5]In the superexchange

Fax: +34-968-364149

E-mail: pmolina@um.es

[c] Institut für Allgemeine Anorganische und Theoretische

Chemie, Universität Innsbrück,

6020, Innrain 52a, Innsbruck, Austria

[d] Department of Biological Chemistry, John Innes Centre,

Norwich Research Park, Colney, Norwich, NR4 7UH, UK

Supporting information for this article is available on the

WWW under http://ww w .eujic.org/ or from the author.

2436

DOI: 10.1002/ejic.200400992

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

mechanism the bridge serves solely to mediate the wave- spacer with two conjugated double bonds (e.g. butadiene,^[14]

functions of the two electroactive centers that act as elec- 1,4-diaza-1,3-butadiene,^[15]2,3-diaza-1,3-butadiene,^[16]and

tron-donor and -acceptor groups; the electron (or hole) is others^[17]). However, no examples have been reported of me-

never vibronically localized on the bridge. In contrast, in tal–metal interactions between metallocene units linked by

the hopping mechanism the electron (or hole) is located at an asymmetric spacer like the 2-aza-1,3-butadiene bridge.

the bridge for a short time (vibronically localized) during

Here we wish to present the synthesis and study of struc-

its journey from one redox center to the other. The former tural and electronic properties of ferrocenyl (Fc) π-conju-

mechanism is generally the origin of many of the observed gated complexes 5a–d, whose electronic properties have

IV-CT bands in mixed-valence species, while the hopping been systematically varied by introducing an electron-do-

one is much less common, probably because of the strict nating or electron-withdrawing substituent, either at the 1-

structural and electronic requirements that must be fulfilled position or at the 4-position of the 2-aza-1,3-butadiene

by the bridge. Nowadays there is a great interest in the de- moiety linked to the ferrocenyl unit. We also wish to present

velopment of organic bridges that favor the hopping mecha- here a study of the homobimetallic complex 5f, which con-

nism between the two electroactive centers since they could tains two ferrocene units linked by the asymmetric 2-aza-

provide new molecular devices that act as electronic relays 1,3-butadiene bridge. The mono-oxidized species of this

to promote IETs over large distances.

complex, 5f^+·, has permitted the study of the influence of

A large number of bis-metallocenes and metallocenes an asymmetric bridge with an electroactive character on the

linked by saturated carbon bridges have been synthesized intramolecular electron transfer (IET) phenomenon, dem-

and studied with respect to metal–metal interactions. It has onstrating that the 2-aza-1,3-butadiene bridge favors the

been demonstrated that, compared to bis-metallocenes, spe- IET between the two metallic centers.

cies linked by a single carbon bridge show substantially

weaker metal–metal interactions.^[6]In general, stronger

Results and Discussion

electrochemical metal–metal interactions occur when two

linkages are made between two metallocenes and the metals

are brought into closer proximity. On the other hand, IV-

CT absorptions are usually not observed in the mono-oxid-

Synthesis and Characterization

Compounds 5 were all prepared from the readily avail-

ized derivatives of ferrocenes with insulating hydrocarbon able diethyl aminomethylphosphonate (1),^[18]which was

bridges.^[7]Most bis-metallocenes linked by saturated carbon condensed with the appropriate aromatic or organometallic

bridges belong to class I in the Robin–Day classification,^[8]aldehyde 2 to give the corresponding N-substituted diethyl

the only interaction detected being at the electrochemical aminomethylphosphonate 3 in excellent yields (85–95%).

level and which can be attributed only to electrostatic and Generation of the metalloenamine by treatment with nBuLi

inductive effects. In contrast, ferrocenyl groups linked by at –78 °C and subsequently with one equivalent of the alde-

unsaturated bridges show slightly larger metal–metal inter- hyde 4 provided the 1,4-disubstituted 2-aza-1,3-butadienes

actions than those with analogous saturated bridges,^[2c,9–13]5 in yields ranging from 40% to 90% (Scheme 1). While two

and the corresponding mono-oxidized olefin-bridged spe- individual operations are required in the latter synthesis, it

cies show IV-CT absorptions in the near-IR region. Thus, is possible to execute the entire sequence of reactions in a

many such species belong to class II, in which there is a single flask without isolation of the intermediates 3 without

moderate coupling between the two electroactive centers affecting the yield. However, in this case, isolation of com-

mediated by the organic bridge through a superexchange pounds 5 proved to be more difficult. Compounds 5, recrys-

mechanism. Intermetallic interactions of different magni- tallized from dichloromethane/diethyl ether (1:10), were

1

tudes have been observed in diferrocene complexes in which characterized by mass spectroscopy and H and ¹³C NMR

the two metallocene units are linked through a symmetric spectroscopy, as well as by elemental analyses.

Scheme 1.

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2437

P. Molina, J. Veciana et al.

FULL PAPER

Assignment of the configuration of the double bonds effect is observed on the aldiminic H₁; no NOE effect is

present in the 1,4-disubstituted 2-aza-1,3-butadienes 5 was observed to H₁when irradiation takes place at δ = 6.63 ppm

1

achieved by inspection of the H NMR spectroscopic data. (H₃).

It is well established that the condensation reaction between

a primary amine and an aldehyde is not sterereoselective,

hence both (E)- and (Z)-aldimine isomers are generally

present in the reaction product. However, it must be em-

X-ray Crystal Structures

phasized that the condensation reaction between 1 and the

Single crystals of compounds 5b, 5d, and 5f, suitable for

appropriate aldehyde 2 takes place stereoselectively to give X-ray structure determination, were grown from diffusion

exclusively the (E)-isomer, as ascertained by ¹H NMR spec- of n-hexane into a solution of the compound in CH₂Cl₂.

troscopy. Indeed, an NOE effect is observed to the methyl- The crystal data and details of data collection and structure

ene group on irradiation of the aldiminic proton of 3a. On refinement are given as Supporting Information (Table S1).

the other hand, N-substituted diethyl aminophosphonates Attempts to grow good-quality single crystals of 5a and 5c

3, after deprotonation with nBuLi, react smoothly in THF for X-ray determination were unsuccessful.

¯

with the corresponding aldehyde 4 to give only the trans

Compound 5b crystallizes in the triclinic space group P1

configured carbon-carbon double bond, as is expected in a with four molecules in the unit cell. Two independent mole-

Horner–Wadsworth–Emmons olefination process. This cules were found per asymmetric unit. Figure 1 (see a)

configuration was confirmed by the characteristic vicinal shows the ORTEP drawing of the molecular structure of 5b

coupling constants (J = 14.0 Hz) in the ¹H NMR spectrum. with the atomic numbering scheme. The molecular struc-

In addition, NOE and two-dimensional NOESY experi- ture reveals that 5b is present as the (E,E)-isomer, as is ob-

1

ments on a solution of compound 5f in CDCl₃confirmed served in solution by H NMR spectroscopy, and has an s-

not only the (E,E)-configuration of the double bonds pres- trans configuration in the solid state. The interatomic dis-

ent in the bridge, but also the s-cis-conformation in this tances and bond angles of 5b are close to those previously

solvent. Thus, on irradiation at δ = 7.04 ppm (H₄) an NOE observed for ferrocene^[19]and other substituted ferrocene

Figure 1. ORTEP views of the asymmetric unit of 5b showing the atom numbering scheme (a) and of 5d showing the atom numbering

scheme (b). Thermal ellipsoids are drawn at the 50% probability level.

2438

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

compounds.^[20]The most interesting structural feature of 5b Theoretical Calculations

is that all atoms of the 2-aza-1,3-butadiene bridge and the

Geometry optimizations at the DFT level were carried

out for neutral 2-aza-1,3-butadienes 5 as well as for the oxi-

dized species 5^·+; the resulting geometrical parameters are

gathered in Table 1. For those compounds whose X-ray

structures are reported here, a high degree of coincidence is

observed with the calculated values, as shown by the follow-

ing critical geometrical parameters: the calculated

N(bridge)···Fe distances for 5b, 5d, and 5f are 5.309, 3.868,

and 3.885 (Fe₁) and 5.380 Å (Fe₄), while the corresponding

experimental distances are 5.242, 3.893, and 3.967 and

5.169 Å, respectively; the calculated Fe₁···Fe₄distance in 5f

is 9.225 Å, while the experimental one is 9.107 Å. The other

two calculated structures 5a and 5c display similar values

(5.350 and 3.860 Å, respectively) for the N(bridge)···Fe dis-

tance. On the other hand, all calculated structures have a

pure s-trans conformation (C1–N2–C3–C4 dihedral angles

of 179.7°, 178.3°, 177.5°, 178.6°, and 179.2° for 5a–d and

5f, respectively), with the substituted phenyl rings almost

parallel to the 2-aza-1,3-butadiene plane (angles between

mean planes of 0.3°, 3.5°, 4.7°, and 1.4° for 5a–d, respec-

tively) and the bridge-attached Cp rings with small twisting

angles (13.6°, 5.6°, 7.8°, 3.5°, 1.4°, and 12.1° for 5a–d and

Fc₁and Fc₄in 5f, respectively). Where available, all experi-

mental ring (Cp or substituted phenyl) twisting angles are

systematically higher than the corresponding calculated val-

ues. From these calculations it is possible to conclude that

both neutral 2-aza-1,3-butadienes 5 and oxidized species 5^·+

show planar structures exhibiting a large conjugation and

a high structural rigidity (Table 1).

two attached rings are disposed in a planar configuration,

thus indicating a large degree of π-conjugation and rigidity

of this organic ligand.

Compound 5d crystallizes in the monoclinic space group

P21/n with four molecules in the unit cell. Figure 1 (see b)

shows the ORTEP drawing with the atomic numbering

scheme. Molecules of this compound are also present as the

(E,E)-isomer, as observed in solution by ¹H NMR spec-

troscopy, and have an s-trans configuration in the solid

state. The bond and dihedral angles reveal that all the

atoms of the bridge and the two linked rings are disposed

in a planar configuration, as in compound 5b, thus indicat-

ing a large degree of conjugation in the π-system.

Compound 5f crystallizes in the monoclinic space group

P21/c with two molecules in the unit cell. Figure 2 shows

the ORTEP drawing of 5f with the atomic numbering

scheme. The molecular structure of 5f reveals a 1:1 disorder

of the two atoms C12 and N1 at the same position. The

crystallographic asymmetric unit is only one half of the

molecule, since the other half is generated by a symmetry

operation with a center of symmetry, as shown in Figure 2.

Since the molecule does not really possess a symmetry cen-

ter it must be disordered, therefore C12 and N1 were re-

fined each with an occupancy of 0.5 with the same coordi-

nates and temperature factors. Molecules of 5f are also

1

present as the (E,E)-isomer, as observed in solution by H

NMR spectroscopy, but in the solid state they adopt an s-

trans conformation in contrast to the major conformation

adopted in solution (vide supra). One of the two equivalent

possibilities of the molecule is shown in Figure 2.

Electronic Spectroscopy

The UV/Visible data obtained in CH₂Cl₂for 5a–d and 5f

are collected in Table 2. These spectra are characterized by

a very strong absorption with a maximum between 336 and

375 nm, which is assigned to a localized π-π* excitation

mainly within the 2-aza-1,3-butadiene bridge. In addition

to this band, another weaker absorption is visible between

469 and 535 nm, which is assigned to another localized exci-

tation with a lower energy produced either by two nearly

degenerate transitions, an Fe^IId-d transition^[21](e.g.

HOMO–1 Ǟ LUMO+3 in 5a), or by a metal-to-ligand

charge transfer (MLCT) process (d_π-π*)^[22](e.g. HOMO–3

Ǟ LUMO+1 in 5a). By comparison of the compounds with

the ferrocenyl unit in the same position of the bridge it is

easy to conclude that the presence of an electron-donating

methoxy group (5a and 5c) results in a significant decrease

of the wavelength of the absorption maximum relative to

the compounds with a nitro substituent (5b and 5d). On the

other hand, the compounds with the ferrocenyl group at the

4-position of the bridge show a shift of the absorption max-

ima to longer wavelengths relative to the compounds with

the ferrocenyl group at the 1-position. Therefore, the calcu-

lated energy gap between HOMO and LUMO in the first

position (3.6558 and 2.8194 eV for 5a and 5b, respectively)

is slightly smaller than in the second position (3.6889 and

Figure 2. Top: ORTEP view of the asymmetric unit of 5f showing

the atom numbering scheme. Thermal ellipsoids are drawn at the

50% probability level. Bottom: drawing of a complete molecule of

5f.

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2439

P. Molina, J. Veciana et al.

Table 1. Selected calculated^[a]and geometrical parameters^[b]for neutral 2-aza-1,3-butadienes 5 and oxidized species 5^+·.

FULL PAPER

R¹–C¹

C¹–N¹

N²–N³

C³–C⁴

C⁴–R⁴

C¹N²C³C¹N²C³C⁴R^1#Azd^#Azd^#R^4#Fe¹···N²Fe⁴···N²Fe¹···Fe⁴

5a

5b

1.457

1.462

(1.454)

1.444

1.439

(1.442)

1.439

1.294

1.293

(1.280)

1.294

1.297

(1.282)

1.297

1.399

1.395

(1.387)

1.401

1.398

(1.394)

1.398

1.348

1.350

(1.335)

1.348

1.349

(1.320)

1.349

1.447

1.443

(1.447)

1.459

1.458

(1.460)

1.458

119.8

120.6

(117.7)

119.5

119.0

(117.1)

119.0

(117.1)

119.5

(–0.3)

121.3

(+1.8)

121.0

(+2.0)

118.8

(–0.8)

119.3

(+0.5)

179.7

178.3

(176.7)

177.5

178.6

(179.2)

178.6

(179.2)

177.5

(–2.2)

178.3

(+0.8)

179.9

(+1.3)

176.1

(–3.1)

178.1

(–0.2)

0.3

3.5

(5.2)

7.8

3.5

(3.9)

1.4

(8.1)

1.0

(+0.7)

4.2

(–3.6)

7.6

(+4.1)

3.4

13.6

5.6

(9.6)

4.7

5.350

5.309

(5.242)

5c

5d

3.860

3.868

(3.893)

3.885

1.4

(4.2)

12.1

(8.1)

7.4

(–6.2)

2.1

(–2.6)

1.0

(–0.4)

9.0

5f

5.380

(5.169)

5.445

9.225

(9.107)

(1.442)

1.431

(1.282)

1.313

(1.394)

1.370

(1.320)

1.366

(1.460)

1.424

(3.967)

5a^+·

5c^+·

5d^+·

5f^+·

5f²⁺

(–0.026) (+0.019) (–0.029) (+0.018) (–0.023)

1.433 1.308 1.374 1.367 1.436

(–0.011) (+0.014) (–0.027) (+0.019) (–0.023)

1.446 1.296 1.392 1.354 1.455

(+0.007) (–0.001) (–0.006) (+0.005) (–0.003)

1.420 1.314 1.373 1.364 1.425

(–0.024) (+0.020) (–0.028) (+0.017) (–0.022)

1.412 1.409 1.295 1.458 1.415

(–0.008) (+0.094) (–0.078) (+0.094) (–0.010)

(+0.095)

3.984

(+0.124)

3.900

(+0.032)

3.842

5.456

9.261

(+2.0)

24.5^[c]

(+21.1)

(–1.6)

(–0.043) (+0.076) (+0.036)

3.057 4.124 7.172

35.7^[c]

(+26.7) (–0.785) (–1.332) (–2.089)

[a] The experimental values for neutral compounds, where available, and, for oxidized species, the calculated absolute difference in relation

to the precedent structure before one-electron oxidation, are given in parentheses. [b] Distances [Å] and bond and dihedral angles [°]; #

stands for the azadiene skeleton (Azd), Cp or substituted-phenyl mean planes. [c] Unusually large angles between mean planes do not

reflect a twisting rotation around a C_Cp–C_Azdbond, but a bending distortion of the exocyclic substituent towards Fe atom in order to

alleviate the electron deficiency. See for instance ref.^[49]

Table 2. Calculated and experimental electronic absorption bands of neutral compounds 5a–d and 5f.

Compound

R¹

R²

Calculated

λ_max[nm] (oscillator strength)

Experimental^[a]

λ_max[nm] (10^–3ε, m^–1cm^–1)

5a

5b

5c

5d

5f

4-CH₃OC₆H₄

4-NO₂C₆H₄

Fc

343.6 (0.7961), 375.8 (0.1944), 469.7 (0.0052) 345 (24.5), 360 (sh), 470 (2.1)

378.7 (0.4862), 526.9 (0.0965) 375 (16.9), 535 (4.4)

341.3 (1.2479), 369.2 (0.0132), 470.3 (0.0033) 342 (20.7), 356 (sh), 469 (1.9)

366.1 (0.5466), 522.6 (0.0690) 371 (18.7), 504 (4.5)

336.2 (0.0415), 339.9 (0.1006), 476.2 (0.0033) 336 (25.5), 346 (sh), 478 (3.6)

4-CH₃OC₆H₄

4-NO₂C₆H₄

Fc

[a] In CH₂Cl_2.

2.9889 eV for 5c and 5d, respectively) and this gap is greatly 2-aza-1,3-butadiene bridge. Indeed, monoferrocenyl deriva-

increased by the presence of an electron-donating substitu- tives 5a–d show two oxidation waves in the range 0–1.5 V

ent at the aromatic ring. Finally, the homobimetallic com- vs. SCE; their potentials are collected in Table 3. The first

plex 5f shows an absorption spectrum that is very similar to oxidation wave for all compounds is reversible and occurs at

those of complexes 5a and 5c, as expected from the similar potentials close to that of the Fe^II/Fe^IIIcouple in ferrocene

electron-donor abilities of ferrocenyl and p-methoxyphenyl (0.460 V vs. SCE, or 0.530 V vs. decamethylferrocene,

groups (calculated HOMO–LUMO gap of 3.7785 eV for DMFc).^[23]Therefore, this wave is assigned to the Fe^II/Fe^III

5f).

redox couple. The oxidation potentials of the ferrocenyl

At this point, it is useful to consider the electronic transi- units of these compounds show a dependence on the posi-

tions in terms of the molecular orbitals involved. Analysis tion of the 2-aza-1,3-butadiene bridge to which they are

of the electronic transitions by time-dependent DFT meth- attached and also on the electron-donating ability of the

ods (TDDFT) gave essentially identical results for each aromatic group linked to the bridge. Thus, higher potentials

studied compound between experimental and calculated are observed when the ferrocenyl is attached at the 1-posi-

transitions (summarized in Table 2). The agreement be- tion of the bridge and when the substituent of the aromatic

tween calculated optical transitions and the experimentally ring is a nitro group. In contrast to the first oxidation wave,

observed results lends credence to the used theoretical level the second wave is clearly not reversible for 5a–c under the

as well as to the attained geometries and the molecular- conditions of our experiments, since only the anionic peak

orbital model developed to explain such optical transitions. appears in the cyclic voltammogram. Taking into account

In addition, such an agreement confirms the assignation that the p-methoxyphenyl moiety is the most “ferrocene-

previously made to each optical transition.

like” group in terms of its electron-donating capability, as

demonstrated experimentally by the similarity of the UV/

Vis spectra of 5a, 5c, and 5f (vide supra), and also by inde-

pendent non-linear optical measurements,^[24]the 1,4-bis(p-

Electrochemistry

Compounds 5 are all expected to show electroactivity methoxyphenyl)-2-aza-1,3-butadiene 5e was also studied for

due both to the Fe^II/Fe^IIIcouple and the oxidation of the comparison purposes. The CV of this compound shows two

2440

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

irreversible oxidation waves at 1.084 and 1.440 V vs. DMFc sense, the electrochemical behaviors of compound 5e and

that may be assigned to the oxidation of the N atom of the N-p-methoxybenzylidene-p-anisidine are almost identical,

bridge and the p-methoxyphenyl group, respectively. This since the CV displays two irreversible oxidation peaks at

assignment is based on the fact that when the N atom of 1.08 and 1.52 V vs. SCE.

5e is protonated under acidic conditions the first oxidation

The CV of 5f, which contains two ferrocenyl groups, was

peak disappears while the second remains unchanged. On recorded in two different experiments. Firstly, the CV (see

the basis of these observations, the second wave observed a in Figure 3) scanned from 0–1 V shows two closely spaced

in the CVs of compounds 5a–c is also assigned to the oxi- reversible one-electron oxidations corresponding to the oxi-

dation of the 2-aza-1,3-butadiene group, the potential of dation of the ferrocenyl units present in the molecule. Sec-

which varies from 1.090 V to 1.416 V vs. DMFc and also ondly, when the CV was run from 0.0–1.6 V, after the first

shows a dependence on the substituents of the aromatic scan, three oxidation waves were observed: two for the

ring and on the relative position of the ferrocenyl and aro- ferrocenyl units and one for the oxidation of the bridge. The

matic groups. Such an irreversible oxidation wave is not ob- cathodic return wave is, in general, higher and thinner than

served for compound 5d, probably because it appears at a the anodic forward wave, which can be attributed to some

potential more positive than +1.5 V vs. DMFc due to the adsorption of the oxidized compound onto the electrode

strong electron-withdrawing character of the nitro group.

surface in the ferrocenium form. The difference between the

anodic and cathodic peak potentials (ΔE_p) is lower than the

58 mV value expected for a reversible single-electron wave

of an electroactive species in solution at 20 °C (the fully

adsorbed redox species would ideally have a ΔE_pvalue of

zero). However, it is important to emphasize that the results

obtained in the latter experiment could only be reproduced

when the electrode was cleaned before subsequent scans

were run. Otherwise, the wave corresponding to the oxi-

dation of the bridge disappears, indicating that the electrode

behaves as an insulator in the region in which the azadiene

moiety is oxidized. The protonation process of 5f was also

followed by cyclic voltammetry. Upon protonation, the

three waves present in the neutral ligand evolve into two

Table 3. Cyclic voltammogram data for compounds 5a–f in

CH₂Cl₂.

2

Compound R¹

R²

¹E_1/2(Fc⁺/Fc)^[a]E_ap(irrev)^[a]

5a

5b

5c

5d

5e

4-CH₃OC₆H₄Fc

0.530

0.590

1.090

1.190

1.416

–

4-NO₂C₆H₄

Fc

4-CH₃OC₆H₄0.698

4-NO₂C₆H₄

0.704

–

0.560

0.770

4-CH₃OC₆H₄4-CH₃OC₆H₄

1.084, 1.440

5f

Fc Fc

1.116

[a] E_1/2[V] vs. decamethylferrocene.

To gain further insight into the nature of the irreversible waves in the protonated derivative, the latter wave appear-

oxidation wave, the electrochemical behavior of compounds ing at almost the same potential region as that correspond-

5a–d was investigated in the presence of HBF₄. Upon pro- ing to the bridge in the neutral ligand. In order to under-

tonation, the potential shift of the ferrocenyl unit present stand the nature of the unit causing that wave, Osteryoung

in these compounds was found to be strongly dependent on square-wave voltammetry (OSWV) experiments were car-

its position in the 2-aza-1,3-butadiene bridge. When it is ried out by protonation of 5f (see b in Figure 3). While the

placed at the 4-position a ΔE_1/2(L – LH⁺) of 70 mV (5a) neutral ligand exhibits only two oxidation waves at 0.560

and 80 mV (5b) was observed, while when it is placed at and 0.770 V vs. DMFc, stepwise addition of HBF₄resulted

the 1-position the corresponding shifts are 300 mV (5c) and in a clear evolution of the first oxidation wave from 0.560 V

320 mV (5d). Additionally, in all these cases, the irreversible to 0.630 V and the disappearance of the second oxidation

wave associated with the oxidation of the bridge disap- wave, accompanied by concomitant appearance of a new

peared upon protonation, except for 5c, which seems to be oxidation wave at 1.060 V. The peak intensity of the new

due to the oxidation of the methoxy group present in the oxidation wave increases with the concentration of the acid

ring.

added up to until two equivalents. At this point, the original

The species formed during the irreversible oxidation oxidation waves disappear and the new ones reach full de-

waves of 5a–c are probably the radical cations centered velopment. The resulting two oxidation waves, with a 1:1

mainly at the N atom of the 2-aza-1,3-butadiene bridge, ratio and a separation of 0.430 V, are due to the oxidation

species which must be relatively unstable under the experi- of the two ferrocenyl moieties in the protonated complex.

mental conditions used. Although no electrochemical oxi- Again, for reproducibility, it is very important to clean the

dation study of the 2-aza-1,3-butadiene fragment has been electrode after each scan.

reported, this description is based on the fact that the an-

It is worth mentioning that the difference between the

odic oxidation of the closely related N-benzylidene-p-anisi- first two reversible oxidation waves in compound 5f (ΔE_1/2

dine Schiff bases show two irreversible anodic peaks at = 210 mV) is much larger than those observed for most bis-

1.08–1.25 and 1.41–1.61 V vs. SCE. The peak potentials of ferrocenyl compounds with four-membered π-conjugated

the first wave, which show good linearity against the σ⁺bridges, such as 1,4-bis(ferrocenyl)butadiene (129 mV),^[14]

values of the substituent in the benzylidene ring, are as- 1,4-bis(ferrocenyl)butadiyne (120 mV),^[10]1,4-bis(ferro-

cribed to the formation of a radical cation centered at the cenyl)-1,4-diazabutadiene (60 mV),^[15]and 1,4-bis(ferro-

nitrogen atom, whereas the second peak is attributed to the cenyl)-2,3-diaza-1,4-dimethylbutadiene (90 mV).^[16]The ori-

oxidation of the substituent in the aromatic ring.^[25]In this gin of this large ΔE_1/2value must be ascribed to the asym-

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2441

P. Molina, J. Veciana et al.

FULL PAPER

the metal and the organic bridge oxidation potentials is rel-

atively small, it is possible, upon oxidation of the ferrocenyl

group, to induce optically an electron transfer from the thi-

enyl group to the Fe^IIIcenter.^[26]This IET has an associated

low-energy ligand-to-metal charge transfer (LMCT) band

in the near-IR, which is very similar to those observed in

symmetrical mixed-valence complexes. Moreover, it has

been reported recently that the electrochemical oxidation of

Fe^IIcenters of mono- or bis(ferrocenylethynyl)oligothioph-

enes, in which the potential difference between the revers-

ible oxidation of the ferrocenyl group and the irreversible

oxidation of the thienyl ring varies from 0.38 V to 1.12 V,

yields the corresponding radical cations, which also exhibit

oligothiophene-to-Fe^IIIcharge-transfer LMCT bands in the

near-IR region.^[27]Taking into account that in compounds

5a–d the potential differences between the two successive

oxidation waves are in the range of 0.3–0.7 V (Table 3), and

that these values are comparable to those of oligothi-

ophene-substituted ferrocenes, the oxidized ferrocenyl de-

rivatives 5a^·+–d^·+were considered suitable candidates to

study the optically induced IET from the 2-aza-1,3-butadi-

ene bridge to the Fe^IIIcenter. In favor of such an expecta-

tion is the recently reported fact that ferrocenyl derivatives

containing a -CH₂-N(R)-CH₂- bridge undergo intramolecu-

lar electron transfer between the Fe^IIIcenter and the nitro-

gen atom in the electrochemically oxidized state to give an

intermediate containing an Fe^IIcenter and a nitrogen with

a radical cation character.^[28]

The generation of the oxidized species derived from 5a–

d was performed electrochemically and monitored by ab-

sorption spectroscopy. Stepwise coulometric titrations were

performed on ca. 2×10^–3m solutions of complexes 5a–d in

CH₂Cl₂, with [(nBu)₄N] PF₆(0.15 m) as supporting electro-

lyte, and the absorption spectra were regularly recorded for

different average number (0 Յ n Յ 1) of removed electrons.

The UV/Vis–near-IR spectroscopic data of 5a^·+–d^·+are col-

lected in Table 4. The spectra of 5a^·+–d^·+show three absorp-

tion bands, one in the UV region in the range of 371–

418 nm of strong intensity, another in the visible region be-

tween 518 and 685 nm, with a weaker intensity, and an even

weaker additional band in the near-IR region between 973–

1219 nm. The latter band is not observed for 5d^·+. The high-

energy absorptions in the UV appear at similar wavelengths

and have comparable intensities to those shown by the cor-

responding neutral complexes 5a–d and, therefore, they are

assigned to localized π-π* transitions at the 2-aza-1,3-buta-

diene group. The absorptions appearing in the visible region

are similar to the bands shown by other oxidized ferrocenyl

derivatives and, consequently, are assigned to Cp Ǟ Fe^III

ligand-to-metal charge transfer (LMCT) transitions.^[21]Fi-

nally, the absorption band in the near-IR, which is not ob-

served in the neutral complexes and is likely to take place

after the creation of an electronic vacancy in the HOMO,

is assigned to a 2-aza-1,3-butadiene Ǟ Fe^IIILMCT transi-

tion. This assignment is based on the following considera-

tions: (i) during the course of the oxidations of 5a–c these

Figure 3. a) Cyclic voltammogram of compound 5f (1 mm) in

CH₂Cl₂/[(nBu)₄N]ClO₄scanned at 0.1 Vs^–1from –0.5 V to 0.9 V

(dash) and from –0.5 V to 1.6 V (solid). b) OSWV of compound

5f (1 mm) in CH₂Cl₂/[(nBu)₄N]ClO₄recorded at 0.1 mVs^–1before

(solid) and after (dash) addition of 1 equiv. of HBF₄. Decamethyl-

ferrocene (DMFc) was used as an internal standard.

metry of the bridge linking the two ferrocenyl groups. Thus,

the distinct position of the N atom with respect to the two

ferrocenyl groups should perturb the two ferrocenyl moie-

ties to a different extent.

Optically Induced Intramolecular Electron Transfer

in 5a^·+–d^·+

It has been demonstrated that in electrochemically active low-energy bands increase continuously in intensity, achiev-

ferrocenylthiophene derivatives, in which the gap between ing a maximum when the oxidation process is complete (see

2442

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

Figure 4); (ii) defined isosbestic points at 385 and 487 nm in 5c^·+) that is mainly localized on this N atom, as pointed

are maintained during the course of the oxidations (see Fig- out by theoretical calculations. Such calculations also show

ure 4); and (iii) similar low-energy transitions have been ob- an absolute energetic minimum on the potential energy hy-

served in the near-IR region in other Fe^IIIand Ru^IIIcom- persurface essentially corresponding to the respective neu-

plexes with conjugated oligothiophene ligands^[26,27]and tral top after the appropriate geometrical reorganization to

with a -CH₂-N(R)-CH₂- bridge.^[28]Therefore, the absorp- redistribute the electronic density. Only a rough correlation

tion bands in the near-IR shown by 5a^·+–c^·+correspond to with the experimental absorption values was claimed from

an optically induced electron transfer from the 2-aza-1,3- time-dependent DFT calculations performed on some of

butadiene to the Fe^IIIcenter, as schematically shown in Fig- these structures (see Table 4), due to the reported inaccu-

ure 5 using the Marcus–Hush model.^[4]We have assumed racy of these types of methods in particular cases.^[29]Al-

that during this optical transition one electron from the lo- though there is only a rough correlation between the experi-

calized pair of nonbonding electrons on the N atom of the mental and calculated values of absorption bands for oxid-

2-aza-1,3-butadiene bridge is transferred to the Fe^IIId_πor- ized complexes 5a^·+–d^·+, the results of the time-dependent

bital of the ferrocene moiety. This assumption is supported DFT calculations confirmed the assignations of optical

by the presence of a high-energy filled orbital in oxidized transitions previously given for 5a^·+–d^·+, based on experi-

complexes 5a^·+–d^·+(e.g. HOMO–2 in 5a^·+and HOMO–3 mental facts.

Table 4. Calculated and experimental electronic absorptions of oxidized species 5a^+·–d^+·and 5f^+·.

Compound Calculated

λ_max[nm] (oscillator strength)

Experimental^[a]

λ_max[nm] (10^–3ε, m^–1cm^–1)

418.8 (0.1538), 421.1 (0.1415), 486.6

(0.0560), 527.7 (0.0002), 566.0 (0.0014),

947.5 (0.0210) 418 (8.5), 481 (3.8), 519 (2.8), 575

(1.9), 618 (0.9)

975 (1.0)

5a^+·

618.0 (0.3043)

–

259 (16.7), 330 (10.24), 390(sh), 518

(2.29), 685 (0.70)

1219 (0.30)

5b^+·

5c^+·

5d^+·

269.6 (0.3447), 286.9 (0.1123), 360.0

(0.3045), 414.9 (0.2880), 556.8 (0.2648)

–

974.5 (0.0275) 259 (9.8), 293 (7.3), 383(7.2), 536 (2.2) 973 (0.13)

–

263 (19.7), 286 (sh), 371(25.2), 549

(4.2)

–

349.3 (0.1272), 374.2 (0.0753), 425.8

(0.0428), 549.5 (0.0101)

846.0 (0.0027), 327 (12.8), 369 (sh), 429 (sh), 547 (4.9) 764 (0.33), band

5f^+·

1249.4 (0.0017)

A; 1250 (0.64)

band B

–

297 (sh), 365 (11.4), 558 (3.7)

764 (0.60), band

A; 971 (0.36);

band C

5f²⁺

[a] In CH₂Cl₂.

Figure 4. Evolution of Vis-near-IR spectra during the course of the oxidation of compound 5c. The average number of removed electrons

is given on each spectrum. Inset: NIR region enlargement.

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2443

P. Molina, J. Veciana et al.

FULL PAPER

via a single oscillator having the same frequency in both

initial and final states.^[4]Originally developed for the in-

terpretation of intervalence charge transfer (IV-CT) bands

in extended solids, the Marcus–Hush model has been ex-

tended and applied to IV-CT in mixed-valence organome-

tallic^[30,31]and organic compounds,^[1g,32]as well as to

LMCT and metal-to-ligand charge transfer (MLCT) pro-

cesses in charge-localized asymmetric complexes.^[33,34]In

accordance with such precedents we have also used the

Marcus–Hush model to study the IET of asymmetric com-

plexes 5a^·+–d^·+. As schematized in Figure 5, such species

exist in two states having the charge/hole localized either on

the metal (M) or a particular region of the organic ligand

(L), which are separated by the relative free enthalpy

ΔG°_LM. The electronic coupling between the two electroac-

tive centers (M and L) is measured by the V_LMterm and

the optical transition energy of the LMCT band at its maxi-

mum (hν_max; in cm^–1), corresponds to the sum of ΔG°_LM

and the vertical reorganization energy λ_LM(i.e. hν_max

ΔG°_LM+ λ_LM).

=

The spectral data of LMCT bands of 5a^·+–d^·+, summa-

rized in Table 5, can be used to determine the electronic

coupling parameter, V_LM, and the reorganization energy,

λ_LM, for each oxidized compound. This was accomplished

by deconvoluting the LMCT bands^[35]using the classical

procedure (see Figure S2 in the Supporting Information)

assuming that the effective electron-transfer distance is the

distance between the Fe and N atoms. Such distances can

be determined from the crystal structures of 5b, 5d, and 5f

(5.2 and 3.9 Å for the complexes with the ferrocenyl groups

attached to the 4- and 1-positions, respectively). The rela-

tive free enthalpies of the two vibronically localized elec-

tronic states for compounds 5a–c can be estimated, in a

first approximation, from the difference of electrochemical

potentials corresponding to the oxidation of the 2-aza-1,3-

butadiene center (L) and to the ferrocenyl center (M) of

neutral complexes. Due to the experimental impossibility of

measuring the L⁰/L⁺¹couple of the 2-aza-1,3-butadiene

when it is linked to a neutral ferrocenyl group, the corre-

sponding oxidation potential was assumed to be 40 mV less

positive than the anionic peak of compound 5e, observed at

+1.084 V, since the p-methoxyphenyl and ferrocenyl groups

Figure 5. Marcus–Hush model used to describe IET in asymmetric

oxidized complexes 5a^·+–d^·+(top) and in oxidized homobimetallic

complex 5f^+·(bottom).

The Marcus–Hush model can be used to describe the have similar electron-donor capabilities but the first group

classical intramolecular electron-transfer process when the is not electrochemically active at potentials lower than

electron is coupled between a donor and an acceptor group +1.4 V (vide supra). An inspection of the resulting V_LMval-

Table 5. LMCT and IV-CT band parameters, obtained from the spectral deconvolution, and calculated IET parameters for radical cations

5a^+·–d^+·and 5f^+·.

R¹

R²

ΔG⁰

hν_max

λ_LM

ε_max

Δν_1/2

f

^[a]·10³

d

[Å]

V_LM

LM

[cm^–1]

[m^–1cm^–1]

[cm^–1]

5a^+·4-CH₃OC₆H₄

Fc

4150

3660

2790

2740

3900

1110

10260

8200

10280

6110

4540

7490

–

1000

300

130

2640

2090

3800

–

12.1

2.9

2.3

–

5.2

3.9

–

5.2

9.1

650

280

380

5b^+·

5c^+·

5d^+·

5f^+·

4-NO₂C₆H₄

Fc

4-CH₃OC₆H₄

4-NO₂C₆H₄

Fc

–

13090^[b](band A)

9190^[b]

330^[b]

3450^[b]

5.2^[b]

480^[b]

8000^[c](band B)

6890^[c]

640^[c]

2500^[c]

7.4^[c]

260^[c]

[a] Total oscillator strength obtained by f = 4.6×10^–9ε_max·Δν_1/2. [b] LMCT band corresponding to an electron transfer between L and

M units. [c] IV-CT band corresponding to an electron transfer between M and MЈ units.

2444

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

ues for complexes 5a–c (Table 5) shows that electronic

couplings between the ferrocenyl and the 2-aza-1,3-butadi-

ene ligand are moderate, such that all these complexes be-

long to class II.^[8]In addition, the electronic coupling of

complexes 5a–c is influenced by the position to which the

ferrocenyl is attached, as well as on the electron-donor

capability of the substituent on the phenyl group. It should

be noted from the spectral analysis that the charge-transfer

transition oscillator strengths clearly increase for complexes

5a–d. Complex 5d apparently does not show any LMCT

band. This can be ascribed to the presence of a nitro group

linked to the 4-position of the bridge, which drains some

electron density from the 2-aza-1,3-butadiene chain, thus

lowering the effective overlap between diabatic orbitals and

decreasing the probability of the transition. This effect has

indeed been predicted theoretically for other electron-with-

drawing groups, such as the cyano group.^[36]

Intramolecular Electron Transfer in the Mixed-Valence

Compound 5f^+·

Oxidized species derived from 5f were generated electro-

chemically and their formation followed by absorption

spectroscopy. Thus, a stepwise coulometric titration was

performed on a 2×10^–3m solution of 5f in CH₂Cl₂, with

[(nBu)₄N]PF₆(0.15 m) as supporting electrolyte, and ab-

sorption spectra were regularly recorded for a different

average number (0 Յ n Յ 2) of removed electrons. Figure 6a

shows the evolution of the spectra during the oxidation of

5f in the 0 Յ n Յ 1 range, with an increase of the Cp Ǟ

Fe^IIILMCT band appearing at 547 nm, characteristic of a

ferrocenium ion, and a decrease of the band that corre-

sponds to a localized excitation of a ferrocene group at

478 nm. Along with the changes of these bands, the appear-

ance and maintenance of two well-defined isosbestic points

at 425 and 501 nm was observed. Nevertheless, the most

interesting feature is that during the oxidation of 5f two

new weak and broad bands centered at 764 nm (band A)

and at 1250 nm (band B) appear, and these increase contin-

uously in intensity until one electron has been removed (i.e.

when the formation of 5f^+·is complete). As also shown in

Figure 6, on removing more electrons (1 Յ n Յ 2) the inten-

Figure 6. Evolution of Vis-near-IR spectra upon oxidation of 5f:

a) oxidation of 5f until the maximum (n = 1) of the mixed-valence

species 5f^+·is achieved. Inset: Expanded spectra of the visible re-

gion showing the two isosbestic points; b) oxidation until the maxi-

mum (n = 2) of the fully oxidized species 5f²⁺is achieved; c) decon-

volution of the NIR spectrum of the 5f^+·PF₆^–·salt. The experimen-

tal spectrum was deconvoluted by means of three Gaussian func-

tions (dashed lines) using spectral intensity times wavenumber vs.

wavenumber. The sum of the dashed lines (dotted line) closely

matches the experimental spectrum (solid line).

The presence of several near-IR bands in the spectrum

sity of band A continues to increase until its intensity has of a mixed-valence complex is not uncommon, and their

doubled, while band B decreases until it disappears (when occurrence is generally explained by one of three different

the dication 5f²⁺is completely formed). At the same time, possible causes.^[38,39]One could be the presence of a strong

another new weak and broad band at 971 nm (band C) in- spin-orbit coupling effect, which becomes important only

creases until the fully oxidized species 5f²⁺is formed. for complexes containing third-row transition metals.^[38]

A

Table 4 collects all the spectroscopic data for oxidized com- second source could be the presence of a double-exchange

plexes 5f^+·and 5f^·2+. The deconvolution of the experimental mechanism,^[39]a mechanism that becomes more probable

spectra of 5f^+·, performed on spectral intensity times wave- as the bridge length and the level of π-orbitals increases.

number vs. wavenumber, assuming Gaussian shapes (Fig- Finally, such multiple near-IR bands might be caused by

ure 6c), allows an accurate determination of the position, the presence of a bridge with accessible redox state levels,

width, and intensity of bands A and B, which are gathered as has been recently proposed to explain the rich absorption

in Table 5.^[35,37]These spectral parameters are relevant for spectrum of certain mixed-valence compounds with redox-

the characterization of intramolecular electron transfer in active bridges.^[1g]Based on the behaviors of 5a^+·–c^+·, the

5f^+·

latter source seems to be the most probable origin for the

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2445

P. Molina, J. Veciana et al.

FULL PAPER

two near-IR bands of 5f^+·. Accordingly, and also based on higher in energy than the HOMO in 1,3-butadiene, but

the spectral evolution observed during the oxidation of 5f, 1.587 eV higher than HOMO–1 in the 2-aza-1,3-butadiene,

the band A is assigned to a 2-aza-1,3-butadiene Ǟ Fe^IIIthe HOMO being that containing the electron lone-pair at

LMCT transition, which implies an electron transfer from the N atom. (See Figure S3 in the Supporting Information).

the N atom of the 2-aza-1,3-butadiene bridge to the Fe^IIIThis leads to a poorer, although still significant, overlap

site linked at the 4-position of the bridge, whereas band B between the diabatic orbitals of the two ferrocenes with the

is attributed to the bridge-mediated superexchange between orbital of the bridge involved in the superexchange mecha-

the two coupled iron sites (M and MЈ) − an IET between nism.

the two iron sites.

For the mixed-valence 5f^+·the free energy difference be-

In order to estimate the effectiveness of both IET path- tween the two electronic states having the charge localized

ways we used the Marcus–Hush model, as schematized in on each Fe site is 3.2 kcalmol^–1. This relatively large value

Figure 5, where the reorganization energies of both transi- suggests that, at 300 K, the population of the state with the

tions, λ_LMand λ_MMЈ, were determined from the absorption positive charge located on the iron attached to the 4-posi-

maxima of the bands A and B. For the free-energy differ- tion of the 2-aza-1,3-butadiene bridge is dominant, with

ence between the two electronic states of 5f^+·having the just a very small population (Ͻ0.1%) of the state with the

charge localized on each Fe site we used, in a first approxi- charge on the other iron site. In addition, we can make a

mation, the value of ΔG°_MMЈ(1113 cm^–1) obtained from the crude estimation of the thermal activation energy barrier

difference of the first oxidation potentials of 5c and 5f. For for the interconversion of such two states, using the Mar-

ΔG°_LMwe used a value of 3904 cm^–1, which was determined cus–Hush model.^[40,41]The resulting thermal energy barrier

by employing the same approximations as for 5a^+·–c^+·(vide of about 7.4 kcalmol^–1suggests a rate constant of about

supra). The spectral parameters of bands A and B were also 10⁸s^–1at 300 K and hence any spectroscopic technique with

used to determine the effective electronic coupling parame- a time scale faster than 10⁹s^–1should show the valence-

ters V_LMand V_MMЈ, taking as effective electron-transfer localized mixed-valence complex 5f^+·, with the positive

distances those measured between the N and the two Fe charge trapped on the Fe site attached at the 4-position of

atoms in the crystal structure of 5f. Both effective electronic the bridge, at this temperature. This effect would be even

couplings are moderate, indicating that the mixed-valence more pronounced at temperatures below 100 K, so that

5f^+·also belongs to class II.^[8]The relative values of V_LMwith spectroscopic techniques such ⁵⁷Fe Mössbauer and

and V_MMЈindicate that the IET may proceed through the EPR the valence should appear to be localized, as is indeed

two different pathways. The IET mediated by the bridge the case (see Supporting Information).

through a superexchange is apparently less effective than

that in which the electron jumps into the bridge. It is also

of interest to compare the effective electronic coupling be-

Conclusion and Perspectives

tween the two iron sites of 5f^+·with that of the mixed-val-

The presence of low-energy LMCT and IV-CT bands in

the oxidized forms 5a^+·–d^+·and 5f^+·of mono- and bis-ferro-

cenyl derivatives attached to a 2-aza-1,3-butadiene bridge

indicates the existence of optically induced IETs between

the metal units and the organic linkers. This important re-

sult demonstrates that the introduction of a heteroatom,

like the N atom, in a conjugated organic bridge is an at-

tractive strategy for promoting IET through an alternative

pathway. This finding may have implications in the design

of molecular electron-transfer materials and, in particular,

in the development of new molecular devices that act as

electronic relays which promote IET over large distances.

Efforts in this direction are in progress in our laboratories

using symmetrical compounds with two electroactive units

attached to oxidizable, symmetrical, and conjugated organic

bridges containing two heteroatoms.

ence compound derived from 1,4-diferrocenyl-1,3-butadi-

ene, i.e., [Fc(CH=CH)₂Fc]^+·.^[14]The latter compound has a

V_MMЈof 430 cm^–1, which is larger than that of 5f^+·, suggest-

ing that the replacement of an sp²C atom by an N atom in

the 1,3-butadiene bridge results in a decrease of the interac-

tion between both ferrocenes, probably because of a poorer

overlap between the diabatic orbitals of the two ferrocenes

with the orbital of the bridge involved in the superexchange

mechanism. Indeed, in both types of systems the HOMO of

the neutral molecule is the bridge-centered superexchange

involved MO, which can be considered as being formed by

a combination of the higher energy MO of π-type at the

bridge and two 3e₂Ј-based MOs of both ferrocene subunits

(Figure 7). At the level of theory used, the latter is 0.953 eV

Experimental Section

General Procedures: All reactions were carried out under N₂and

using solvents which were dried by routine procedures. All melting

points were determined on a Kofler hot-plate melting point appara-

tus and are uncorrected. IR spectra were determined as Nujol

emulsions or films on a Nicolet Impact 400 spectrophotometer.

UV/Vis–near IR spectra were recorded on a Varian Cary 5 E spec-

Figure 7. Superexchange-responsible MO (HOMO) in compound

5f at a 0.050 isovalue.

2446

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

trophotometer. ¹H and ¹³C NMR spectra were recorded on a

Bruker AC200, 300, or 400 spectrometer (200, 300, or 400 MHz

sponding neutral minimized structures with the appropriate charge

and multiplicity constraints. Structure 5b^+·was not calculated due

to systematic failures in the SCF convergence. Similarly, 5f^+·was

used as the initial geometry for the fully oxidized 5f²⁺. Time-de-

pendent DFT methods (TDDFT), as contained within the

Gaussian 98 program suite,^[48]performed as an SPE calculation at

the same level as the minimized structures, were employed to study

the excited states and model the UV/Vis–NIR electronic transitions

in both neutral and oxidized species.

1

for H). Chemical shifts are referenced to signals of tetramethylsi-

1

lane in the case of H and ¹³C NMR spectra and to 85% aqueous

phosphoric acid in the case of ³¹P NMR spectra. The EI and FAB⁺

mass spectra were recorded on a Fisons AUTOSPEC 500 VG spec-

trometer, using 3-nitrobenzylalcohol as matrix. EPR spectra were

obtained with an X-Band Bruker ESP 300E spectrometer equipped

with a TE₁₀₂microwave cavity, a Bruker variable temperature unit,

a field frequency lock system Bruker ER 033 M; line positions were

determined with an NMR Gaussmeter Bruker ER 035 M. The

modulation amplitude was kept well below the line width, and the

microwave power was well below saturation. Crystallographic mea-

surements were made at 233(2) K on a Bruker diffractometer

equipped with an area detector positioned at the window of a rotat-

ing anode generator using Mo-K_αradiation (λ = 0.71073 Å). Möss-

bauer spectra were recorded on an ES-Technology MS-105 spec-

trometer, with a 100 MBq ⁵⁷Co source, in a rhodium matrix. Solid

samples were prepared by grinding with boron nitride. Spectra were

referenced to natural iron at 298 K. Parameters were obtained by

fitting the data with Lorentzian lines; errors Ͻ0.01 mms^–1. Micro-

analyses were performed on a Perkin–Elmer 240C instrument. The

cyclic voltammetric measurements were performed on a QUICEL-

TRON potentiostat/galvanostat controlled by a personal computer

and driven by dedicated software. Cyclic voltammetry was per-

formed with a conventional three-electrode configuration con-

sisting of platinum working and auxiliary electrodes and an SCE

General Procedure for the Preparation of Diethyl N-Arylidene- and

[(Ferrocenylmethylidene)aminomethyl]phosphonates 3: A mixture of

diethyl aminomethylphosphonate (0.78 g, 4.64 mmol), an equi-

molar amount of the appropriate aldehyde 2, and p-toluenesulfonic

acid (0.01 g, 0.058 mmol) in dry benzene (50 mL) was stirred at

room temperature for 4 h. Then, the flask was fitted with a Dean–

Stark trap and a reflux condenser and heated overnight, atr reflux

temperature, with constant removal of water. The reaction mixture

was allowed to cool to room temperature and the solvent was re-

moved under vacuum to give the corresponding diethyl phosphona-

tes 3 as colored oils, which were used without further purification

in the next step.

Diethyl [(p-Methoxybenzylidene)aminomethyl]phosphonate (3a):

Yield: 95%. IR (Nujol): ν = 1641, 1609, 1577, 1524, 1444, 1422,

˜

1396, 1316, 1252, 1166, 1102, 969 cm^–1. ¹H NMR (CDCl₃): δ =

1.34 (t, ⁴J_H,P= 6.5 Hz, 6 H), 3.84 (s, 3 H), 4.07 (d, ¹J_H,P= 16.8 Hz,

2 H),4.19–4.09 (m, 4 H), 6.91 (d, J = 8.6 Hz, 2 H), 7.68 (d, J =

4

8.6 Hz, 2 H), 8.22 (d, J_H,P= 4.7 Hz, 1 H) ppm. ¹³C NMR

reference electrode. The experiments were carried out with a 10^–3

m

3

(CDCl₃): δ = 16.5 (d, J_P,C= 5.6 Hz, CH₃), 55.3 (CH₃O), 57.4 (d,

solution of sample in dry CH₂Cl₂containing 0.1 m [N(nC₄H₉)₄]-

ClO₄as supporting electrolyte. Deoxygenation of the solutions was

achieved by bubbling nitrogen for at least 10 min and the working

electrode was cleaned after each run. The cyclic voltammograms

were recorded with a scan rate increasing from 0.05 to 1.00 Vs^–1.

The square-wave voltammograms were recorded before and after

addition of sequential additions of an aliquot of 0.1 equiv. of

2.5×10^–2m solutions of HBF₄·Et₂O in CH₃CN. ΔE_s= 4 mV, ΔE_p

= 25 mV, and f = 15 Hz. Decamethylferrocene (DMFe) (–0.07 V

vs. SCE) was used as an internal reference both for potential cali-

bration and for reversibility criteria. Oxidations were performed by

electrolysis in a three-electrode cell under argon using dry CH₂Cl₂

as solvent and 0.15 m [N(n-C₄H₉)₄]ClO₄as supporting electrolyte.

The progress of the oxidation was followed coulometrically (or

chronoamperometrically) by 263A of EG&PAR potentiostat-gal-

vanostat. The reference electrode and the counter electrode were

separately immersed in the solvent containing the supporting elec-

trolyte and isolated from the bulk solution by a glass frit. The

working electrode was a platinum grid. UV/Vis–near IR absorption

spectra were regularly recorded by transferring a small aliquot of

the solution contained in the electrochemical cell into a UV quartz

cell for a different average number of removed electrons.

2

¹J_P,C= 152.6 Hz, NCH₂), 62.4 (d, J_P,C= 6.7 Hz, OCH₂), 114.0

3

(CH_Ar), 128.7 (q), 130.0 (CH_Ar), 162.0 (q), 164.9 (d, J_P,C

=

16.0 Hz, CH=N) ppm. ³¹P NMR (CDCl₃): δ = 22.68 ppm. EIMS:

m/z (%) = 285 (100) [M⁺], 152 (82), 148 (61), 125 (90), 121 (80), 97

(44), 91 (19), 78 (25).

Diethyl [(p-Nitrobenzylidene)aminomethyl]phosphonate (3b): Yield:

85%. IR (Nujol): ν = 1654, 1604, 1523, 1454, 1404, 1360, 1266,

˜

1

4

1097, 1016, 966, 847 cm^–1. H NMR (CDCl₃): δ = 1.34 (t, J_H,P

=

6.9 Hz, 6 H), 4.01–4.24 (m, 6 H), 7.91 (d, J = 8.0 Hz, 2 H), 8.26

4

(d, J = 8.0 Hz, 2 H), 8.40 (d, J_H,P= 5.1 Hz, 1 H) ppm. ¹³C NMR

3

1

(CDCl₃): δ = 16.4 (d, J_P,C= 5.7 Hz, CH₃), 57.7 (d, J_P,C

=

2

152.7 Hz, NCH₂), 62.7 (d, J_P,C= 6.6 Hz, OCH₂), 123.8 (CH_Ar),

3

130.5 (CH_Ar), 140.9 (q), 149.2 (q), 163.2 (d, J_P,C= 22.0 Hz,

CH=N) ppm. ³¹P NMR (CDCl₃): δ = 21.77 ppm. EIMS: m/z (%)

= 300 (36) [M⁺], 163 (60), 152 (96), 125 (82), 108 (100), 90 (95), 81

(42).

Diethyl [(Ferrocenylmethylidene)aminomethyl]phosphonate (3c):

Yield: 94%. IR (Nujol): ν = 1642, 1473, 1448, 1398, 1241, 1166,

˜

1

4

1035, 972, 828, 778 cm^–1. H NMR (CDCl₃): δ = 1.38 (t, J_H,P

=

3.3 Hz, 6 H), 3.92 (d, 2 H, ²J_H,P= 18 Hz), 4.08–4.25 (m, 4 H), 4.23

(s, 5 H), 4.36 (t, J = 1.8 Hz, 2 H), 4.64 (t, J = 1.8 Hz, 2 H), 8.19

(d, ⁴J_H,P= 4.8 Hz, 1 H) ppm. ¹³C NMR (CDCl₃): δ = 16.3 (d, ³J_P,C

Theoretical Calculations: Geometry optimizations were performed

with the Spartan’02 package (Spartan’02, build 119, Wavefunction

Inc., Irvine, CA). For neutral compounds 5a–f a preliminary opti-

mization was done at the semi-empirical PM3(d) level and the ob-

tained geometries were then refined at the DFT level of theory,^[42]

which has proved to be remarkably successful in reproducing exper-

imental ground-state geometries and rotational barriers for ferro-

cenes.^[43]The B3LYP functional^[44](Becke’s three-parameter hybrid

functional^[45]with the Lee–Yang–Parr^[46]correlation functional)

and the 3-21G* basis set were used. Harmonic frequency calcula-

tions verified the nature of the stationary points as minima (all real

frequencies).^[47]The geometry of oxidized radical-cation species

5a^+·, 5c^+·, 5d^+·, and 5f^+·were optimized starting from the corre-

1

2

= 5.7 Hz, CH₃), 57.2 (d, J_P,C= 150.7 Hz, NCH₂), 62.4 (d, J_P,C

=

6.5 Hz, OCH₂), 68.5 (2×CH, Cp), 69.1 (5×CH, Cp), 70.6 (2×CH,

Cp), 79.7 (d, ⁴J_P,C= 9 Hz, q, Cp), 166.2 (d, ³J_P,C= 19.5 Hz, CH=N)

ppm. ³¹P NMR (CDCl₃): δ = 22.77 ppm. EIMS: m/z (%) = 363

(86) [M⁺], 298 (100), 224 (39), 260 (48), 121 (63), 81 (25), 65 (9),

56 (24).

General Procedure for the Preparation of 1,4-Disubstituted-2-aza-

1,3-butadienes 5: n-Butyllithium (1.6 m in hexane, 3 mL) was added

dropwise to a solution of the appropriate diethyl phosphonate 3

(4.64 mmol) in dry THF (20 mL) at –78 °C under nitrogen. A solu-

tion of an equimolar amount of the appropriate aldehyde 4

(4.64 mmol) in dry THF (10 mL) was added dropwise to this stirred

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2447

P. Molina, J. Veciana et al.

FULL PAPER

1

solution and the mixture was stirred for 1.5 h. Then, the reaction

mixture was allowed to reach room temperature and was heated at

reflux temperature overnight. After the solution had cooled to

room temperature, the solvent was evaporated under reduced pres-

sure and the crude product was stirred with water (50 mL) and then

extracted with dichloromethane (4×50 mL). The combined organic

layers were dried with anhydrous Na₂SO₄and, after filtration, the

solution was concentrated to dryness to give a crude product which

was crystallized from dichloromethane/diethyl ether (1:10).

813 cm^–1. H NMR (CDCl₃): δ = 3.80 (s, 3 H), 3.84 (s, 3 H), 6.86

(d, J = 8.0 Hz, 2 H), 6.91 (d, J = 14.1 Hz, 1 H), 6.94 (d, J = 8.0 Hz,

2 H), 7.40 (d, J = 8.0 Hz, 2 H), 7.44 (d, J = 14.1 Hz, 1 H), 7.76 (d,

J = 8.0 Hz, 2 H), 8.25 (s, 1 H) ppm. ¹³C NMR (CDCl₃): δ = 55.2,

55.3, 114.0, 114.1, 127.8, 129.2, 129.4, 130.0, 140.4, 159.2, 159.7,

161.8 ppm. EIMS: m/z (%) = 267 (100) [M⁺], 252 (44), 236 (28),

160 (69), 121 (35), 91 (28). C₁₇H₁₇NO₂(267.33): calcd. C 76.38, H

6.41, N 5.24; found C 76.20, H 6.67, N 5.12.

1,4-Bis(ferrocenyl)-2-aza-1,3-butadiene (5f): Yield: 87%; m.p. 253–

4-Ferrocenyl-1-(p-methoxyphenyl)-2-aza-1,3-butadiene (5a): Yield:

90%; m.p. 175–176 °C. IR (Nujol): ν = 1605, 1511, 1456, 1395,

255 °C. IR (Nujol): ν = 1649, 1580, 1102, 1023, 1041, 1003, 966,

˜

829, 810 cm^–1. H NMR (CDCl₃): δ = 4.15 (s, 5 H), 4.22 (s, 5 H),

1

˜

1240, 1162, 1107, 1074, 1046, 1024, 990, 985, 830, 741 cm^–1. ¹H

NMR (CDCl₃): δ = 3.86 (s, 3 H), 4.15 (s, 5 H), 4.28 (t, J = 1.6 Hz,

2 H), 4.43 (t, J = 1.6 Hz, 2 H), 6.70 (d, J = 14.0 Hz, 1 H), 6.94 (d,

J = 8.0 Hz, 2 H), 7.18 (d, J = 14.0 Hz, 1 H), 7.75 (d, J = 8.0 Hz,

2 H), 8.18 (s, 1 H) ppm. ¹³C NMR (CDCl₃): δ = 55.3 (CH₃O), 67.0

(2×CH, Cp), 69.0 (2×CH, Cp), 69.3 (5×CH, Cp), 81.5 (q, Cp),

114.2 (CH_Ar), 128.3 (CH=CH), 129.4 (q), 129.9 (CH_Ar), 139.7

(CH=CH), 158.0 (q), 161.5 (CH=N) ppm. EIMS: m/z (%) = 345

(100) [M⁺], 280 (77), 278 (10), 121 (31), 56 (23). C₂₀H₁₉FeNO

(345.22): calcd. C 69.58, H 5.55, N 4.06; found C 69.35, H 5.40, N

4.15.

4.27 (s, 2 H), 4.40 (s, 2 H), 4.45 (s, 2 H), 4.72 (s, 2 H), 6.60 (d, J =

13.2 Hz, 1 H), 7.05 (d, J = 13.2 Hz, 1 H), 8.11 (s, 1 H) ppm. ¹³C

NMR (CDCl₃): δ = 66.80 (2×CH, Cp), 68.61 (2×CH, Cp), 69.00

(2×CH, Cp), 69.20 (10×CH, Cp), 71.00 (2×CH, Cp), 80.81 (q,

Cp), 82.11 (q, Cp), 126.32 (CH=), 140.32 (=CH-N), 159.82

(CH=N) ppm. EIMS: m/z (%) = 423 (100) [M⁺], 358 (60), 212 (32),

186 (28), 121 (64), 56 (25). C₂₃H₂₁Fe₂N (423.12): calcd. C 65.29, H

5.00, N 3.31; found C 65.40, H 4.88, N 3.50.

–

Preparation of the Mixed-Valence Compound 5f⁺I₃: A sample of

this mixed-valence compound was prepared by adding a solution

of iodine (0.030 g, 0.12 mmol) in dry benzene (5 mL) to a solution

of 1,4-diferrocenyl-2-aza-1,3-butadiene (0.1 g, 0.24 mmol) in the

same solvent (50 mL). The reaction mixture was stirred at room

temperature under nitrogen for 1 h and the dark-purple microcrys-

tals formed were collected by filtration and washed with three por-

tions of benzene and one portion of diethyl ether. The solid ob-

tained was dried in a dessicator overnight. FAB MS: m/z (%) = 424

(100) [M⁺+ 1]. C₂₃H₂₁Fe₂I₃N (803.82): calcd. C 34.33, H 2.61, N

1.74; found C 34.23, H 2.38, N 1.58.

4-Ferrocenyl-1-(p-nitrophenyl)-2-aza-1,3-butadiene (5b): Yield: 50%;

m.p. 223–224 °C. IR (Nujol): ν = 1600, 1514, 1454, 1335, 1182,

˜

1

1109, 1030, 957, 745, 698 cm^–1. H NMR (CDCl₃): δ = 4.12 (s, 5

H), 4.32 (t, J = 1.6 Hz, 2 H), 4.42 (t, J = 1.6 Hz, 2 H), 6.87 (d, J

= 13.0 Hz, 1 H), 7.21 (d, J = 13.0 Hz, 1 H), 7.89 (d, J = 8.1 Hz, 1

H), 8.22 (s, 1 H), 8.23 (d, J = 8.1 Hz, 1 H) ppm. ¹³C NMR (CDCl₃):

δ = 67.5 (2×CH, Cp), 69.6 (5×CH, Cp), 69.9 (2×CH, Cp), 80.4

(q), 124.2 (CH_Ar), 128.5 (CH_Ar), 134.0 (CH=CH), 138.6

(CH=CH), 142.3 (q), 148.5 (q), 154.7 (CH=N) ppm. EIMS: m/z

(%) = 360 (100) [M⁺], 330 (19), 295 (23), 265 (21), 192 (34), 165

(30), 121 (62). C₁₉H₁₆FeN₂O₂(360.19): calcd. C 63.36, H 4.48, N

7.78; found C 63.54, H 4.35, N 7.69.

Supporting Information: Crystal data of 5b, 5d, and 5f, optical ab-

sorption data of oxidized species and the procedure for their analy-

sis with the Marcus–Hush model, and ⁵⁷Fe Mössbauer and EPR

spectra are provided.

1-Ferrocenyl-4-(p-methoxyphenyl)-2-aza-1,3-butadiene (5c): Yield:

85%; m.p. 102–103 °C. IR (Nujol): ν = 1645, 1605, 1565, 1514,

˜

1250, 1176, 1107, 1038, 969, 826 cm^–1. ¹H NMR (CDCl₃): δ = 3.82

(s, 3 H), 4.22 (s, 5 H), 4.47 (t, J = 1.6 Hz, 2 H), 4.73 (t, J = 1.6 Hz,

2 H), 7.34 (d, J = 14.0 Hz, 1 H), 7.40 (d, J = 8.0 Hz, 2 H), 7.82 (d,

J = 14.0 Hz, 1 H), 7.88 (d, J = 8.0 Hz, 2 H), 8.21 (s, 1 H) ppm.

¹³C NMR (CDCl₃): δ = 55.2 (CH₃O), 68.7 (2×CH, Cp), 68.9

(5×CH, Cp), 71.1 (2×CH, Cp), 80.4 (q), 114.1 (CH_Ar), 127.5

(CH_Ar), 127.6 (CH=CH), 129.3 (q), 141.0 (CH=CH), 159.0 (q),

161.5 (CH=N) ppm. EIMS: m/z (%) = 345 (100) [M⁺], 318 (40),

279 (24), 278 (14), 186 (43), 121 (66), 56 (39). C₂₀H₁₉FeNO

(345.22): calcd. C 69.58, H 5.55, N 4.06; found C 69.33, H 5.28, N

4.21.

Acknowledgments

We gratefully acknowledge the financial support of the DGI, Min-

isterio de Ciencia y Tecnología, Spain (projects MAT2003-04699

and CTQ2004-02201), the Fundación Séneca (CARM) (project no.

PB/72/FS/02), and DGR, Catalunya (2001 SGR00362). The Bio-

technology and Biological Sciences Research Council, UK, is also

thanked for funding (D.J.E.).

[1] a) D. Segal, A. Nitzan, W. B. Davis, M. R. Wasielewski, M. A.

Ratner, J. Phys. Chem. B. 2000, 104, 3817–3829; b) M. N. Pad-

dow-Row, in Electron Transfer in Chemistry (Ed.: V. Balzani),

Wiley-VCH, Weinheim, 2001, pp. 179–271; c) L. DeCola, P.

Belser, in Electron Transfer in Chemistry (Ed.: V. Balzani),

Wiley-VCH, Weinheim, 2001, pp. 97–136; d) K. Kilsa, J. Ka-

janus, A. N. Macpherson, J. Martensson, B. Albinsson, J. Am.

Chem. Soc. 2001, 123, 3069–3080; e) J.-P. Launay, Chem. Soc.

Rev. 2001, 30, 386–397; f) M. D. Newton, Chem. Rev. 1991, 91,

767–792; g) S. F. Nelsen, R. F. Ismagilov, D. F. Powell, J. Am.

Chem. Soc. 1998, 120, 1924–1925; h) W. B. Davis, W. A. Svec,

M. A. Ratner, M. R. Wasielewski, Nature 1998, 396, 60–53; i)

D. S. Weiss, J. R. Cowdery, R. H. Young, in Electron Transfer

in Chemistry (Ed.: V. Balzani), Wiley-VCH, Weinheim, 2001,

pp. 379–471; j) H. Bassler, in Semiconducting Polymers (Eds.:

G. Hadziioannou, P. F. van Hutten), Wiley-VCH, Weinheim,

2000, pp. 365–410.

1-Ferrocenyl-4-(p-nitrophenyl)-2-aza-1,3-butadiene (5d): Yield: 40%;

m.p. 182–183 °C. IR (Nujol): ν = 1561, 1508, 1438, 1337, 1225,

˜

1177, 1123, 1107, 1033, 750, 696 cm^–1. ¹H NMR (CDCl₃): δ = 4.24

(s, 5 H), 4.56 (t, J = 1.6 Hz, 2 H), 4.78 (t, J = 1.6 Hz, 2 H), 6.87

(d, J = 13.4 Hz, 1 H), 7.52 (d, J = 8.0 Hz, 2 H), 7.55 (d, J =

13.4 Hz, 1 H), 8.18 (d, J = 8.0 Hz, 2 H), 8.35 (s, 1 H) ppm. ¹³C

NMR (CDCl₃): δ = 69.4 (2×CH, Cp), 69.6 (5×CH, Cp), 72.1

(2×CH, Cp), 79.6 (q), 124.2 (CH_Ar), 125.2 (CH=CH), 126.6

(CH_Ar), 143.7 (q), 146.4 (q), 146.5 (CH=CH), 165.9 (CH=N) ppm.

EIMS: m/z (%) = 360 (100) [M⁺], 333 (22), 192 (22), 186 (34), 165

(27), 120 (63), 56 (33). C₁₉H₁₆FeN₂O₂(360.19): calcd. C 63.36, H

4.48, N 7.78; found C 63.19, H 4.29, N 7.62.

1,4-Bis(p-methoxyphenyl)-2-aza-1,3-butadiene (5e): Yield: 85%;

m.p. 170–172 °C. IR (Nujol): ν = 1647, 1609, 1553, 1510, 1454,

˜

1423, 1373, 1311, 1280, 1255, 1174, 1111, 1024, 986, 943, 875,

2448

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Radical Cations Derived from Ferrocenyl-Substituted 2-Aza-1,3-butadienes

FULL PAPER

[2]

Hendrickson, H. B. Gray, J. Am. Chem. Soc. 1971, 93, 3603–

3612.

S. Barlow, H. E. Bunting, C. Ringham, J. C. Green, G. U. Bub-

litz, S. G. Boxer, J. W. Perry, S. R. Marder, J. Am. Chem. Soc.

1999, 121, 3715–3723.

N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96, 877–910.

V. Alain, A. Fort, M. Barzoukas, T.-H. Chen, S. R. Blanchard-

Desce, S. R. Marder, J. W. Perry, Inorg. Chim. Acta 1996, 242,

43–49.

M. Masui, H. J. Ohmori, J. Chem. Soc., Perkin Trans. 2 1972,

1882–1887.

a) Y. B. Zu, M. O. Wolff, Chem. Mater. 1999, 11, 2995–3001; b)

Y. B. Zu, D. B. Millet, M. O. Wolf, S. J. Rettig, Organometallics

1999, 18, 1930–1938.

a) U. T. Müller-Westerhoff, Angew. Chem. Int. Ed. Engl. 1986,

25, 702–717; b) E. E. Bunel, P. Campos, J. Ruz, L. Valle, I.

Chadwick, M. Santa Ana, G. Gonzalez, J. M. Manriquez, Or-

ganometallics 1988, 7, 474–476; c) E. E. Bunel, L. Valle, N. L.

Jones, P. J. Carroll, M. Gonzalez, N. Muñoz, J. M. Manriquez,

Organometallics 1988, 7, 789–791; d) H. Atzkern, B. Huber,

F. H. Köhler, G. Müller, R. Müller, Organometallics 1991, 10,

238–244; e) J. Hiermeier, F. H. Köhler, G. Müller, Organometal-

lics 1991, 10, 1787–1793.

a) A. C. Benniston, V. Goulle, A. Harriman, J. M. Lehn, B.

Marczinke, J. Phys. Chem. 1994, 98, 7798–7804; b) M. Haga,

M. M. Ali, S. Koseki, K. Fujimoto, A. Yoshimura, K. Nozaki,

T. Ohno, K. Nakajima, D. J. Stufkens, Inorg. Chem. 1996, 35,

3335–3347; c) M. E. Elliot, D. L. Derr, S. Ferrere, M. D. New-

ton, Y. P. Liu, J. Am. Chem. Soc. 1996, 118, 5221–5228; d) C.

Pataoux, C. Coudret, J. P. Launay, C. Joachim, A. Gourdon,

Inorg. Chem. 1997, 36, 5037–5049; e) J.-P. Sauvage, J.-P. Collin,

J.-C. Chambron, S. Guillerez, C. Coudret, Chem. Rev. 1994, 94,

993–1019; f) N. S. Hush, Coord. Chem. Rev. 1985, 64, 135–157;

g) A. C. Ribou, J. P. Launay, K. Takhashi, T. Nihira, S. Tarut-

ani, C. W. Spangler, Inorg. Chem. 1994, 33, 1325–1329.

a) C. Creutz, Prog. Inorg. Chem. 1983, 30, 1–73; b) R. J.

Crutchley, Adv. Inorg. Chem. 1994, 41, 273–325; c) S. Barlow,

D. O’Hare, Chem. Rev. 1997, 97, 637–669; d) N. S. Hush, Prog.

Inorg. Chem. 1967, 8, 391–444.

a) C. Lambert, G. Nöll, J. Schelter, Nat. Mater. 2002, 1, 69–

73; b) P. J. Low, M. A. J. Paterson, H. Puchmann, A. E. Goeta,

J. A. K. Howard, C. Lambert, J. C. Cherryman, D. R. Tackley,

S. Leeming, B. Brown, Chem. Eur. J. 2004, 10, 83–91.

W. H. Morrison, S. Krogsrud, D. N. Hendrickson, Inorg.

Chem. 1973, 12, 1998–2004.

a) F. Delgado-Pena, D. R. Talham, D. O. Cowan, J. Or-

ganomet. Chem. 1983, 253, C43–C46; b) M. F. Moore, S. R.

Wilson, D. N. Hendrickson, U. T. Mueller-Westerhoff, Inorg.

Chem. 1984, 23, 2918–2920; c) M. F. Moore, S. R. Wilson,

M. J. Cohn, T. Y. Dong, U. T. Mueller-Westerhoff, D. N. Hend-

rickson, Inorg. Chem. 1985, 24, 4559–4565.

M. Robin, P. Day, Adv. Inorg. Radiochem. 1967, 10, 247–422.

a) T. Y. Dong, T. J. Ke, S. M. Peng, S. K. Yeh, Inorg. Chem.

1989, 28, 2103–2106; b) B. Floris, P. Tagliatesta, J. Chem. Res.,

Synop. 1993, 42–43.

C. Le Vanda, K. Bechgaard, D. O. Covan, J. Org. Chem. 1976,

41, 2700–2704.

I. Motoyama, M. Watanabe, H. Sano, Chem. Lett. 1978, 513–

516.

C. Jutz, Tetrahedron Lett. 1959, 1–4.

a) J. Lukasser, H. Angleitner, H. Schottenberger, H. Kopacka,

M. Schweiger, B. Bildstein, K. H. Ongania, K. Wurst, Organo-

metallics 1995, 14, 5566–5578; b) A. Tárraga, P. Molina, J. L.

López, A. Espinosa, D. J. Evans, Tetrahedron Lett. 2002, 43,

4717–4720.

A. C. Ribou, J. P. Launay, M. L. Sachtleben, H. Li, C. W.

Spangler, Inorg. Chem. 1996, 35, 3735–3740.

[22]

[23]

[24]

[3]

[25]

[26]

[27]

[28]

Y. B. Zu, M. O.Wolf, J. Am. Chem. Soc. 2000, 122, 10121–

10125.

a) I. Yamaguchi, T. Sakano, H. Ishii, K. Osakada, T. Yamam-

oto, J. Organomet. Chem. 1999, 584, 213–216; b) M. Horie, T.

Sakano, K. Osakada, H. Nakao, Organometallics 2004, 23, 18–

20; c) M. Horie, Y. Suzaki, K. Osakada, J. Am. Chem. Soc.

2004, 126, 3684–3685.

M. Turki, C. Daniel, S. Zalis, A. Vicek, J. van Slageren, D. J.

Stufkens, J. Am. Chem. Soc. 2001, 123, 11431–11440.

M. C. B. Colbert, J. Lewis, N. J. Long, P. R. Raithby, M.

Younus, A. J. P. White, D. J. Williams, N. N. Payne, L. Yel-

lowlees, D. Beljonne, N. Chawdhury, R. H. Friend, Organome-

tallics 1998, 17, 3034–3043.

[4]

[5]

[29]

[30]

[31]

Y. Zhu, O. Clot, M. O. Wolf, G. P. A. Yap, J. Am. Chem. Soc.

1998, 120, 1812–1821.

[6]

[7]

[32]

[33]

C. Lambert, G. Nöll, J. Am. Chem. Soc. 1999, 121, 8434–8442.

a) P. Desjardins, G. P. A. Yap, R. J. Crutchley, Inorg. Chem.

1999, 38, 5901–5905; b) C. E. B. Evans, M. L. Naklicki, R. J.

Crutchley, Inorg. Chem. 1995, 34, 1350–1354.

T. Ito, N. Imai, T. Yamagushi, T. Hamaguchi, C. H. Londer-

gan, C. P. Kubiak, Angew. Chem. Int. Ed. 2004, 43, 1376–1381

and references cited therein.

I. R. Gould, D. Noukakis, L. Gómez-Jahn, R. H. Young, J. L.

Goodman, S. Farid, Chem. Phys. 1993, 176, 439–456.

C. Joachim, Thesis, Université Paul Sabatier, Tolouse, France,

1990, pp. 146, 211, and 214.

[34]

[35]

[36]

[37]

[8]

[9]

The proportion of mixed-valence species at half-oxidation (P),

[10]

[11]

½

where P = K_c/(2 + K_c^1/2) is generally used to compute the

intensity of corrected spectra. The comproportionation con-

stant, K_c, which is related to the thermodynamic stability of

the mixed-valence compound, of the equilibrium of compro-

portionation, [Fe^II-L-F^II] + [Fe^III-L-F^III] w 2[Fe^II-L-F^III], was

calculated for 5f from the electrochemical data, giving a value

of 3.6×10³. With such a high value of K_c, P is almost 100%

so that the dependency on K_cis very small and it is not neces-

sary to correct the intensity of the spectrum.

a) E. M. Kober, K. A. Goldsby, D. N. S. Narayane, T. J. Meyer,

J. Am. Chem. Soc. 1983, 105, 4303–4309; b) P. A. Lay, R. H.

Magnuson, H. Taube, Inorg. Chem. 1988, 27, 2364–2371; c)

W. M. Laidlaw, R. G. Denning, J. Chem. Soc., Dalton Trans.

1994, 1987–1994.

[12]

[13]

[38]

[14]

[15]

[16]

B. Bildstein, M. Malaun, H. Kopacka, M. Fontani, P. Zanello,

Inorg. Chim. Acta 2000, 300–302, 16–22.

a) A. G. Osborne, M. W. da Silva, M. B. Hursthouse, K. M. A.

Malik, G. Opromolla, P. J. Zanello, J. Organomet. Chem. 1996,

516, 167–176; b) V. A. Sauro, M. S. Worketin, Can. J. Chem.

2002, 80, 250–256.

a) A. Tárraga, P. Molina, D. Curiel, M. D. Velasco, Organome-

tallics 2001, 20, 2145–2152; b) A. Tárraga, P. Molina, D. Cur-

iel, M. D. Velasco, Tetrahedron Lett. 2002, 43, 8453–8457.

S. K. Davidsen, G. W. Phillips, S. F. Martin, Org. Synth. 1993,

8, 451–458.

P. Séller, J. D. Dunitz, Acta Crystallogr., Sect. B 1982, 38, 1741.

J. L. López, A. Tárraga, A. Espinosa, M. D. Velasco, P. Mol-

ina, V. Lloveras, J. Vidal-Gancedo, C. Rovira, J. Veciana, D. J.

Evans, K. Wurst, Chem. Eur. J. 2004, 10, 1815–1826.

a) G. L. Geoffroy, M. S. Wrighton, Organometallic Photochem-

istry, Academic Press, New York, 1979; b) Y. S. Sohn, D. N.

[39]

[40]

a) D. E. Richardson, H. Taube, J. Am. Chem. Soc. 1983, 105,

40–51; b) J. Halpern, L. E. Orgel, Discuss. Faraday Soc. 1960,

29, 32–41.

According to Hush and Marcus, for symmetrical weakly cou-

pled class II mixed-valence complexes the activation barrier to

[17]

[18]

electron transfer is given by ΔG_MMЈ* = λ_MMЈ/4 – V_MMЈ

+

(V_MMЈ²/λ_MMЈ); see ref.^[41]. With the parameters given in Table 5

for 5f^+·, a value of ΔG_MMЈ* + ΔG°_MMЈ= 7.4 kcalmol^–1is ob-

tained for the 4-Fe³⁺+ 1-Fe²⁺Ǟ 4-Fe²⁺+ 1-Fe³⁺process. The

corresponding first-order rate constant (k_th)can be calculated

from k_th= k·ν_n·exp(–ΔG_MMЈ*/RT) where k = 1, since the reac-

tion is adiabatic, and the nuclear frequency factor, ν_n, is

2.00×10⁺¹³s^–1, as determined from ν_n= [2(V_MMЈ)²/h] · [(π³/

λ_MMЈRT)^1/2]. This gives a value for k_thof 8.8×10⁺⁷s^–1at

300 K.

[19]

[20]

[21]

Eur. J. Inorg. Chem. 2005, 2436–2450

www.eurjic.org

2449

P. Molina, J. Veciana et al.

FULL PAPER

[41] a) R. A. Marcus, J. Chem. Phys. 1956, 24, 966–978; b) R. A.

Marcus, J. Chem. Phys. 1965, 43, 679; c) R. A. Marcus, N.

Sutin, Inorg. Chem. 1975, 14, 213–216; d) C. Creutz, M. D.

Newton, N. Sutin, J. Photochem. Photobiol. Sect. A 1994, 82,

47–59.

[42] a) E. J. Baerends, D. E. Ellis, P. Ros, Chem. Phys. 1973, 2, 41–

51; b) W. Kohn, A. D. Becke, R. G. Parr, J. Phys. Chem. 1996,

100, 12974–12980.

[43] a) A. Berces, T. Ziegler, L. Fan, J. Phys. Chem. 1994, 98, 1584–

1595; b) A. Berces, T. Ziegler, Top. Curr. Chem. 1996, 182, 41–

85; c) U. Hohm, D. Goebel, S. Grimme, Chem. Phys. Lett.

1997, 272, 328–334; d) C. A. Morrison, S. F. Bone, D. W. H.

Rankin, H. E. Robertson, F. Parsons, R. A. Coxall, S. Fraser,

J. A. S. Howell, P. C. Yates, N. Fey, Organometallics 2001, 20,

2309–2320.

[44] L. J. Bartolotti, K. Fluchick, in Reviews in Computational

Chemistry (Eds.: K. B. Lipkowitz, B. D. Boyd), VCH Publish-

ers, New York, 1996, vol. 7, pp. 187–216.

[47] J. W. McIver, A. K. Komornicki, J. Am. Chem. Soc. 1972, 94,

2625–2633.

[48] Gaussian 98, Revision A.9, M. J. Frisch, G. W. Trucks, H. B.

Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.

Zakrzewski, J. A. Montgomery Jr., R. E. Stratmann, J. C. Bu-

rant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin,

M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.

Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J.

Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma,

D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman,

J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu,

A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L.

Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A.

Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.

Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gor-

don, E. S. Replogle, J. A. Pople, Gaussian, Inc., Pittsburgh PA,

1998.

[49] R. Gleiter, R. Seeger, Helv. Chim. Acta 1971, 54, 1217–1220..

Received: November 30, 2004

[45] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.

[46] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.

2450

www.eurjic.org

Eur. J. Inorg. Chem. 2005, 2436–2450

Article Doi

Synthesis and characterization of radical cations derived from mono- And biferrocenyl-substituted 2-Aza-1,3-butadienes: A study of the influence of an asymmetric and oxidizable bridge on intramolecular electron transfer

DOI: 10.1002/ejic.200400992

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejic.200400992

Full text of DOI:10.1002/ejic.200400992

Products guided by the article

R&D Labs maybe for 560133-91-7

Relevant to this article

Hot Product