Redox active organoiron alkynyl centers with a coordination site: An approach toward molecular plugs. Complexation reactions with (THF)W(CO)5 and (PhCN)2MCl2 (M=Pd, Pt)

www.elsevier.nl/locate/ica

Inorganica Chimica Acta 291 (1999) 403–425

Redox active organoiron alkynyl centers with a coordination site:

an approach toward molecular plugs. Complexation reactions with

(THF)W(CO)₅and (PhCN)₂MCl₂(M=Pd, Pt)

Sylvie Le Stang, Fre´de´ric Paul, Claude Lapinte *

UMR CNRS 6509 Organome´talliques et Catalyse: Chimie et Electrochimie Mole´culaires, Uni6ersite´ de Rennes 1, Campus de Beaulieu,

35042 Rennes Cedex, France

Received 25 January 1999; accepted 5 March 1999

Abstract

The synthesis and full characterization of the functionalized complexes 3a–c [(dppe)Cp*Fe(CꢀC–L)] (L=4-Py, 3-Py, 2-Py;

dppe=1,2-bis(diphenylphosphino)–ethane; Cp*=pentamethylcyclopentadienyl) are reported. These electroactive compounds,

stable under two redox states, behave as typical pyridyl ligands towards simple metal precursors of tungsten, palladium or

platinum complexes. With (THF)W(CO)₅bi-metallic iron–tungsten complexes [(dppe)Cp*Fe(CꢀC–Py)W(CO)₅] (5a–b) were

isolated. After chemical oxidation their corresponding [(dppe)Cp*Fe(CꢀC–Py)W(CO)₅][PF₆] (5a–b⁺) salts were isolated and

characterized as well. With the palladium(II) and platinum(II) dichloride (PhCN)₂MCl₂precursors, neutral, mono- and dicationic

trimetallic complexes of the type {[(dppe)Cp*Fe(CꢀC-4-Py)]₂MCl₂}[PF₆]_n(M=Pd: 8a, Pt: 10a; n=0, 1, 2) were similarly isolated

Keywords: Iron complexes; Tungsten complexes; Alkynyl complexes; Paramagnetic complexes; Polynuclear complexes

pyridyl ligands connected to an unsaturated linker as

ethylene, acetylene, or (poly)ethylene could convey elec-

trons in MV complexes and sometimes lead to strong

electronic [18–20] or magnetic [20–22] interactions be-

tween remote metal centers. Since our group is involved

strongly in the synthesis and study of organometallic

1. Introduction

Among the diverse emerging ﬁelds initiated by the

success of supramolecular chemistry this last decade,

molecular electronics is a fascinating domain which

aroused a lot of fundamental research [1]. The use of

self-organization reactions to synthesize discrete molecu-

lar devices that allow electron transfer reactions to take

place and to be studied at will constitutes a challenging

goal [1–4]. Organization in space of molecular units

possessing a simple pyridyl moiety by means of coordi-

nation to templating metal centers is well known in

supramolecular chemistry [5–10]. Alternatively, the use

of nitrogen-based heterocycles to convey electrons be-

tween metal centers in polynuclear mixed valence (MV)

complexes found its origin in the seminal work of Creutz

and Taube [11,12] and has now considerable precedence

[13–17]. It has also been deﬁnitively established that

architectures

possessing

several

electroactive

‘(dppe)Cp*Fe’ units as models for new nanoscopic

devices [23–28] we reasoned that by judicious substitu-

tion of the electroactive (dppe)Cp*FeCꢀC– unit with a

pyridyl group, we might obtain an interesting family of

organometallic electroactive ligands allowing access to

new polymetallic aggregates (see Scheme 1). With such

a building block, depending on the choice of the central

metal, it should be possible to access polymetallic molec-

ular wires by simple coordination reactions, the pyridyl

ligand constituting a ‘molecular plug’ in this instance. A

prerequisite to such a prospect is however to understand

the electronic changes brought by complexation and to

delineate the factors determining the intramolecular

interaction between the remote iron centers in these

pyridyl-based ligands.

* Corresponding author. Tel.: +33-299-286 361; fax: +33-299-

281 646.

E-mail address: claude.lapinte@univ-rennes1.fr (C. Lapinte)

PII: S0020-1693(99)00135-8

404

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

(dppe) were prepared according to the published proce-

dures and other chemicals were used as received. All the

manipulations were carried out under argon atmo-

sphere using Schlenk techniques or in a Jacomex 532

dry box under nitrogen. Transmittance-FTIR and solid

reﬂectance-near-IR (NIR) spectra were recorded using

a Bruker IFS28 spectrometer, using a Nernst Globar

source and a KBr separator with a DTGS detector

(400–7500 cm⁻¹) or tungsten source and a quartz

separator with a Peltier-effect detector (5200–12500

cm⁻¹). Far-IR spectra were recorded on a Perkin–

Elmer PE1600 spectrometer. Liquid NIR spectra were

recorded on a Cary 5 spectrometer. UV–Vis spectra

were recorded on an UVIKON 942 spectrometer. High

ﬁeld NMR spectra experiments were performed on a

multinuclear Bruker 300 MHz or 200 MHz instrument

(AM300WB and 200DPX). Chemical shifts are given in

parts per million relative to tetramethylsilane (TMS) for

¹H and ¹³C NMR spectra, H₃PO₄for ³¹P NMR spectra.

Cyclic voltammograms were recorded using a PAR 263

instrument. LSIMS analyses were effected at the ‘Cen-

tre Regional de Mesures Physiques de l’Ouest’

(CRMPO) on a high resolution MS/MS ZabSpec TOF

Micromass spectrometer (8 kV). Elemental analyses

were performed at the Center for Microanalyses of the

CNRS at Lyon-Solaise, France.

Scheme 1.

While many pyridyl ligands bearing a ferrocene re-

dox active center have been synthesized with the

prospect of tuning the reactivity of a coordinated metal

center [21,29–35], the few complexes reported to date

and bearing two of these units were never used speciﬁ-

cally for investigating the electron delocalization in

their MV state [29,30,33]; for example, the electrochem-

ical data given for trans-rhenium [29,33] or cis-plat-

inum [30] complexes were suggestive of a weak or

non-existent electronic coupling in their MV state. An

electroactive ‘(dppe)Cp*Fe’ pendant group on the

pyridyl ligand is however expected to give different

results from the ferrocenyl substituent, since the ligand

will now be connected directly to the iron nucleus by a

network of s- and p-bonds. This should result in more

effective electronic coupling between the electroactive

iron center and the pyridyl group¹.

In the following, we will describe (i) the full charac-

terization under the neutral and oxidized states of new

pyridyl ligands 3a–c bearing the organoiron(II) redox

active group, whose syntheses have been communicated

brieﬂy [38]; (ii) their mono-complexation reaction with

a tungsten pentacarbonyl center and the characteriza-

tion of the neutral and oxidized resulting complexes

5a–b/5a–b⁺; (iii) their di-complexation reactions with

palladium or platinum dichloride giving bis-complexes

and the characterization of these complexes under the

different stable redox states 8a/8a^{+ +}/10a/10a^{+ +}and

(iv) the results of our investigations regarding electron

delocalization in their MV state 8a⁺/10a⁺. This will

allow some conclusions to be drawn about the possibil-

ity of using 3a–c as molecular plugs.

2.2. Synthesis of the organoiron ligands (3a–c)

General procedure for the catalytic coupling. In a

Schlenk tube, 400 mg of complex (h₂-dppe)(h₅-

C₅Me₅)Fe–CꢀCH (1; 0.651 mmol), 46 mg of bis

(triphenylphosphine) dichloropalladium complex (0.065

mmol) and 25 mg of copper iodide (0.130 mmol) are

introduced

under

argon.

Subsequently,

the

halogenoaromatic substrate is introduced in 10 ml of

di-iso-propylamine and the mixture is reﬂuxed for 12 h.

The solvent is then trapped cryogenically and the

brown residue is extracted with toluene and ﬁltered on

a celite pad. Evaporation of the toluene and washing

with small portions of n-pentane (2×5 ml) at −40°C

and acetonitrile (5 ml) at −20°C yields the desired

complex 3.

2. Experimental

2.3. (p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-p-(C₅H₄N) (3a)

2.1. General data

The coupling was effected as described with 632 mg

of para-bromo pyridinium hydrochloride salt (1.62

mmol) at 90°C (reﬂux). Approximately 314 mg of the

orange complex 3a (0.45 mmol) is obtained after drying

in vacuo. The pure compound can be isolated as a

bright orange powder by recrystallization from impuri-

ties in toluene–pentane mixtures. Yield 69% (without

recrystallization). Anal. Calc. for C₄₃H₄₃N₁P₂Fe₁: C,

74.68; H, 6.27; N, 2.03; P, 8.96. Found: C, 74.26; H,

6.36; N, 1.99%. ³¹P NMR (81 MHz, CDCl₃) l_P100.6

Reagent grade tetrahydrofuran (THF), toluene, di-

ethylether and n-pentane were dried and distilled from

sodium benzophenone ketyl prior to use. Pentamethyl-

cyclopentadiene and 1,2-(diphenylphosphino)ethane

¹As demonstrated by the much increased electronic interaction

reported for MV complexes with (dppe)Cp*Fe termini relative to

these with ferrocenyl ones. Compare for instance Refs. [23] and [36]

or [26] and [37].

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

405

1

3

(s, 2P, dppe). H NMR (200 MHz, CDCl₃) l_H8.20 (d,

6.73 (dd, 1H, J_HH=7.4 Hz, J_HH=5.1 Hz, C₅H₄N/

2H, ³J_HH=5.8 Hz, C₅H₄N/H_ortho); 7.85 (m, 4H, 2C₆H₄/

H_meta); 6.67 (d, 1H, J_HH=7.9 Hz, C₅H₄N/H_meta); 2.76

3

H_ortho);

7.35–7.00

(m,

16H,

2C₆H₅+2C₆H₅/

(m, 2H, CH_2dppe)_;2.03 (m, 2H, CH_2dppe); 1.45 (s, 15H,

3

H_meta+para+2C₆H₅/H_ortho); 6.62 (d, 2H, J_HH=5.8 Hz,

C₅(CH₃)₅). ¹³C{¹H} NMR (50 MHz, CDCl₃) l_C149.7

1

2

C₅H₄N/H_meta); 2.60 (m, 2H, CH_2dppe); 2.00 (m, 2H,

(s, J_CH=175 Hz, C₅H₄N/C_orthoN); 148.9 (t, J_CP=38

Hz, CꢀC(C₅H₄N)); 148,1 (s, C_ipso); 139.6–127.6 (m,

CH_2dppe); 1.42 (s, 15H, C₅(CH₃)₅). ¹³C{¹H} NMR (50

2

1

2

MHz, CDCl₃) l_C157.7 (t, J_CP=38 Hz, CꢀC(C₅H₄N));

4C₆H₅); 135.2 (s, J_CH=158 Hz, J_CH=7 Hz, C₅H₄N/

_paraN); 125.8 (s, J_CH=163 Hz, J_CH=6 Hz, C₅H₄N/

1

3

149.4 (s, C₅H₄N/C_orthoN); 139.2–127.7 (m, 4C₆H₅);

138.1 (m, C₅H₄N/C_ipso); 125,1 (s, C₅H₄N/C_metaN); 119.6

(s, CꢀC(C₅H₄N)); 88.6 (s, C₅(CH₃)₅); 31.1 (m, CH_2dppe);

C

3

_orthoN); 122.4 (t, J_CP=2 Hz, CꢀC(C₅H₄N)); 117.5 (s,

¹J_CH=162 Hz, J_CH=7,6 Hz, C₅H₄N/C_metaN); 88.6 (s,

1

10.5 (s, J_CH=126 Hz, C₅(CH₃)₅). MS (LSIMS⁺, m-

C₅(CH₃)₅); 31.3 (m, CH_2dppe); 10.7 (s, J_CH=127 Hz,

1

NBA) m/z 692 (M+1, 100%); 589 (M-‘CꢀC(C₅H₄N)’,

C₅(CH₃)₅).

95%); 556 (M–‘Cp*’, 25%).

2.6. Synthesis of the oxidized salts 3a–c⁺

2.4. (p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-m-(C₅H₄N) (3b)

2.6.1. General procedure

The coupling was effected as described with 0.31 ml

of meta-bromo pyridine (3.24 mmol) at 90°C (reﬂux).

Approximately 355 mg of the orange complex 3b (0.51

mmol) are obtained after drying in vacuo. The pure

compound can be isolated as a bright orange powder

by recrystallization from impurities in toluene–pentane

mixtures. Yield 79% (without recrystallization). Anal.

Calc. for C₄₃H₄₃N₁P₂Fe₁: C, 74.68; H, 6.27; N, 2.03; P,

8.96. Found: C, 74.82; H, 6.30; N, 2.01%. ³¹P NMR (81

A total of 0.90 equiv. of [(h⁵-C₅H₅)₂Fe][PF₆] were

added to a solution of the corresponding compound

3a–c in 15 ml dichloromethane. Stirring was main-

tained 1 h at room temperature and the solution was

concentrated in vacuo to ca. 5 ml. Addition of 50 ml of

n-pentane allowed precipitation of a solid. Decantation,

subsequent washing with 3×3 ml portions of toluene

followed by 3×3 ml diethylether and drying under

vacuum yielded the pure 3a–c⁺salts as dark purple

solids.

1

MHz, CDCl₃) l_P100.7 (s, 2P, dppe). H NMR (200

3

MHz, CDCl₃) l_H8.10 (d, 1H, J_HH=4.5 Hz, C₅H₄N/

H_ortho); 8.05 (m, 1H, C₅H₄N/H_ortho); 7.85 (m, 4H,

2C₆H₅/H_ortho); 7.41–7.33 (m, 16H, 2C₆H₅+2C₆H₅/

2.7. [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-p-(C₅H₄N)][PF₆]

(3a⁺)

3

H

_meta+para+2C₆H₅/H_ortho); 7.05 (d, 1H, J_HH=7.8 Hz,

3

C₅H₄N/H_para); 6.97 (dd, 1H, J_HH=7.8 Hz, J_HH=4.4

Hz, C₅H₄N/H_meta); 2.62 (m, 2H, CH_2dppe); 1.99 (m, 2H,

CH_2dppe); 1.42 (s, 15H, C₅(CH₃)₅). ¹³C{¹H} NMR (50

The synthesis was effected as described previously

from 146 mg (0.44 mmol) of [(h⁵-C₅H₅)₂Fe][PF₆] and

339 mg (0.49 mmol) of (h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-p-

(C₅H₄N) (3a). The complex 3a⁺(359 mg) was isolated

in 87% yield (150 mg). Anal. Calc. for

C₄₃H₄₃N₁P₃F₆Fe₁·2H₂O: C, 59.19; H, 5.43; N, 1.61.

Found: C, 59.31; H, 5.17; N, 1.68%.

MHz, CDCl₃) l_C152.0 (s, ¹J_CH=179 Hz, C₅H₄N/

2

C

_orthoN); 147.3 (t, J_CP=40 Hz, CꢀC(C₅H₄N)); 143.1 (s,

¹J_CH=179 Hz, C₅H₄N/C_orthoN); 139.6–127.4 (m,

1

2

4C₆H₅); 136.6 (s, J_CH=160 Hz, J_CH=6 Hz, C₅H₄N/

_metaN); 123.1 (s, J_CH=161 Hz, J_CH=9 Hz, C₅H₄N/

1

2

C

_paraN); 116.5 (s, CꢀC(C₅H₄N)); 88.2 (s, C₅(CH₃)₅); 31.1

2.8. [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-m-(C₅H₄N)][PF₆]

(3b⁺)

1

(m, CH_2dppe); 10.6 (s, J_CH=126 Hz, C₅(CH₃)₅).

2.5. (p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-o-(C₅H₄N) (3c)

The synthesis was effected as described previously

from 50 mg (0.15 mmol) of [(h⁵-C₅H₅)₂Fe][PF₆] and 117

mg (0.17 mmol) of (h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-m-

(C₅H₄N) (3b). Drying under vacuum yielded 3b⁺in

The coupling was effected as described with 0.31 ml

of ortho-bromo pyridine (3.25 mmol) at 90°C (reﬂux).

Approximately 330 mg of the orange complex 3c (0.52

mmol) are obtained after drying in vacuo. The pure

compound can be isolated as a bright orange powder

by recrystallization from impurities in toluene–pentane

mixtures. Yield 73% (without recrystallization). Anal.

Calc. for C₄₃H₄₃N₁P₂Fe₁: C, 74.68; H, 6.27; N, 2.03; P,

8.96. Found: C, 74.59; H, 6.24; N, 1.96%. ³¹P NMR (81

87%

yield

(110

mg).

Anal.

Calc.

for

C₄₃H₄₃N₁P₃F₆Fe₁·H₂O: C, 60.44; H, 5.31. Found: C,

60.17; H, 5.03%.

2.9. [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-o-(C₅H₄N)][PF₆]

(3c⁺)

1

MHz, CDCl₃) l_P101.1 (s, 2P, dppe). H NMR (200

The synthesis was effected as described previously

from 86 mg (0.26 mmol) of [(h⁵-C₅H₅)₂Fe][PF₆] and 200

mg (0.29 mmol) of (h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-o-

(C₅H₄N) (3c). The complex 3c⁺was isolated in 87%

3

MHz, CDCl₃) l_H8.33 (d, 1H, J_HH=5.1 Hz, C₅H₄N/

H_ortho); 7.90 (m, 4H, 2C₆H₅/H_ortho); 7.39–7.19 (m, 17H,

C₅H₄N+2C₆H₅+2C₆H₅/H_meta+para+2C₆H₅/H_ortho);

406

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

yield (190 mg). Anal. Calc. for C₄₃H₄₃N₁P₃F₆Fe₁·

2H₂O: C, 59.19; H, 5.43; N, 1.61. Found: C, 59.74; H,

5.19; N, 1.75%.

C₅H₄N); 115.4 (s, CꢀC(C₅H₄N)); 88.6 (s, C₅(CH₃)₅);

1

30.9 (m, CH_2dppe); 10.5 (s, J_CH=127 Hz, C₅(CH₃)₅).

MS (LSIMS⁺, m-NBA) m/z 1016 (M+1, 3%); 691

(3a, 75%); 589 (3a-‘CꢀC(C₅H₄N)’, 100%); 556 (M–

‘Cp*’, 8%).

2.10. Preparation of the tungsten complexes (5a–b)

2.10.1. [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-p-(C₅H₄N)]W-

(CO)₅(5a)

A

2.11. Preparation of the oxidized tungsten complexes

(5a–b⁺)

solution of (h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-p-

2.11.1. [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-p-(C₅H₄N)W-

(CO)₆]PF₆] (5a⁺)

(C₆H₄N) (3a, 400 mg, 0.58 mmol) and 1.1 equiv. of

W(CO)₆in 200 ml of THF were photolyzed over a

period of 10 min with a mercury-vapor 250 W lamp

under permanent argon bubbling at room temperature.

After evaporation of the solvent, the unreacted W(CO)₆

was sublimed out of the solid residue and the remaining

solid was recrystallized from toluene–pentane before

being washed several times with small fractions of

n-pentane. Drying in vacuo yielded the complex 5a as a

deep orange product (430 mg, 73%). ³¹P NMR (81

MHz, CDCl₃) l_P98.8 (s, 2P, dppe). ¹H NMR (200

To a solution of [(h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-p-

(C₅H₄N)]W(CO)₅(4a, 144 mg, 0.14 mmol) in CH₂Cl₂

(10 ml) at −80°C was added 0.90 equiv. (42 mg, 0.13

mmol) of [Fe(h⁵-C₅H₅)₂][PF₆]. Stirring at this tempera-

ture was maintained 6 h and the desired complex was

precipitated as a solid by addition of 80 ml of n-pen-

tane. Decantation, washings with 2×10 ml of cold

n-pentane and drying under vacuum at room tempera-

ture yielded the pure 4a⁺as a green powder (120 mg,

73%).

3

MHz, CDCl₃) l_H8.21 (d, 2H, J_HH=5.8 Hz, C₅H₄N/

H_ortho); 7.73 (m, 4H, 2C₆H₅/H_ortho); 7.38 (m, 16H,

2C₆H₅+2C₆H₅/H_meta+para+2C₆H₅/H_ortho); 6.41 (d,

2H, ³J_HH=5.8 Hz, C₅H₄N/H_meta); 2.58 (m, 2H,

CH_2dppe); 2.02 (m, 2H, CH_2dppe); 1.41 (s, 15H,

C₅(CH₃)₅). ¹³C{¹H} NMR (50 MHz, CDCl₃) l_C203.0

2.11.2. [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-m-(C₅H₄N)W-

(CO)₆]PF₆] (5b⁺)

The synthesis was conducted as described previously

for 4a⁺from 33 mg (0.10 mmol) of [(h⁵-C₅H₅)₂Fe][PF₆]

and 111 mg (0.11 mmol) of [(h²-dppe)(h⁵-C₅Me₅)Fe–

CꢀC-m-(C₅H₄N)]W(CO)₅(4b). The green complex 4b⁺

(50 mg, 0.04 mmol) was isolated in 39% yield.

1

2

(s, CO); 199.7 (s, J_CW=132 Hz, CO); 172.3 (t, J_CP

=

1

37 Hz, CꢀC(C₅H₄N)) 154.9 (s, J_CH=178 Hz, C₅H₄N/

1

C

_orthoN); 139.9–127.5 (m, 4C₆H₅); 126.3 (s, J_CH=162

Hz, C₅H₄N/C_metaN)); 121.7 (t, ²J_CP=5 Hz,

CꢀC(C₅H₄N)); 89.1 (s, C₅(CH₃)₅); 30.9 (m, CH_2dppe);

2.12. Preparation of (4-t-Bu-C₅H₄N)₂Cl₂Pd (7)

10.3 (q, J_CH=126 Hz, C₅(CH₃)₅). MS (LSIMS⁺, m-

1

NBA) m/z 1016 (M+1, 10%); 692 (3a+1, 95%); 589

(3a–‘CꢀC(C₅H₄N)’, 100%); 556 (M–‘Cp*’, 5%).

To 100 mg of (PhCN)₂PdCl₂(0.261 mmol) in 10 ml

of dichloromethane were added 100 ml of 4-tert-butyl-

pyridine (0.680 mmol). The orange–yellow solution

turned instantaneously clearer and the solution was

further stirred at ambient temperature for 30 min. A

pale yellow complex was then precipitated by addition

of 50 ml n-hexane. The ﬁrst crop of the air-stable

complex 7 (45 mg) was isolated by ﬁltration as a solid,

washed with pentane and dried under vacuum. A sec-

ond microcrystalline crop of 7 (22 mg) was isolated by

further workup of the ﬁltrate after slow concentration

of the solvent. Total yield 58%. Anal. Calc. for

C₁₈H₂₆Cl₂N₂Pd₁: C, 48.29; H, 5.89; N, 6.26. Found: C,

47.80; H, 5.73; N, 6.23%. FT-IR (Nujol/KBr, cm⁻¹) w

1615 (vs, w_8aPy); 1551 (m, w_8bPy). Far-IR (Nujol/ICs,

cm⁻¹) w 366 (m, Cl–Pd–Cl trans). ¹H NMR (200

2.10.2.

[(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-m-(C₅H₄N)]W(CO)₅(5b)

Complex 5b was synthesized by the same procedure

employed for 5a except that (h²-dppe)(h⁵-C₅Me₅)Fe–

CꢀC-m-(C₅H₄N) (3b, 170 mg, 0.25 mmol) was used

instead of (h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-p-(C₅H₄N)

(3a). The deep orange complex 5b (194 mg, 0.19 mmol)

was isolated in 76% yield. ³¹P NMR (81 MHz, CDCl₃)

1

l_P100.2 (s, 2P, dppe). H NMR (200 MHz, CDCl₃) l_H

3

8.21 (d, 1H, J_HHꢀ5 Hz, C₅H₄N/H_ortho); 8.11 (s, 1H,

C₅H₄N/H_ortho); 7.79 (m, 4H, 2C₆H₅/H_ortho); 7.39–7.31

(m, 16H, 2C₆H₅+2C₆H₅/H_meta+para+2C₆H₅/H_ortho);

3

7.04 (d, 2H, ³J_HH=7.9 Hz, J_HHꢀ5 Hz, C₅H₄N/

3

H_ortho); 6.85 (d, 2H, J_HH=5.8 Hz, C₅H₄N/H_meta); 2.47

3

5

MHz, CDCl₃) l_H8.70 (dd, 2H, J_HH=5.6 Hz, J_HH

=

(m, 2H, CH_2dppe); 1.83 (m, 2H, CH_2dppe); 1.41 (s, 15H,

3

C₅(CH₃)₅). ¹³C{¹H} NMR (50 MHz, CDCl₃) l_C203.4

1.4 Hz, C₅H₄N/H_ortho); 7.32 (dd, 2H, J_HH=5.6 Hz,

⁵J_HH=1.4 Hz, C₅H₄N/H_ortho); 1.31 (s, 9H, C(CH₃)₃).

MS (LSIMS⁺, m-NBA) m/z 566 (M–Cl+m-NBA,

5%); 456 (M–Cl+t-Bu-Py, 8%); 448 (M, 5%); 411

(M–Cl-1, 32%); 375 (M–2Cl-2, 13%); 136 (t-Bu-Py,

100%).

1

2

(CO); 199.6 (s, J_CW=131 Hz, CO); 158.7 (t, J_CP=38

Hz, CꢀC(C₅H₄N)); 158.1 (s, ¹J_CH=182 Hz, C₅H₄N/

C

_orthoN)); 148.8 (s, ¹J_CH=182 Hz, C₅H₄N/C_orthoN);

139.7–127.3 (m, 4C₆H₅); 137.6 (s, C₅H₄N); 129.3 (s,

C₅H₄N/C_ipso); 124.9 (s, ¹J_CH=165 Hz, ²J_CH=6 Hz,

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

407

2.13. Preparation of [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-p-

(C₅H₄N)]₂Cl₂Pd (8a)

(Found for ﬁrst pate: C, 59.18; H, 4.94; N, 1.89.)

1

Found: C, 60.71; H, 5.01; N, 1.80%. H NMR (200

3

MHz, CDCl₃) l_H8.30 (d, 4H isomer 1, J_HH=7.0 Hz,

To 200 mg of (h²-dppe)(h⁵-C₅Me₅)FeꢁCꢁC-p-

(C₅H₄N) (3a, 0.289 mmol) and 53 mg of (PhCN)₂PdCl₂

(0.145 mmol) were added 10 ml of dichloromethane.

The deep red solution was stirred at ambient tempera-

ture (20°C) for 30 min and the solvent was evacuated to

dryness. The residual brick red solid was washed three

times with 10 ml of n-pentane portions then extracted

with two fractions (20 and 10 ml) of toluene and ﬁltered

through a celite pad. The toluene was concentrated to

ca. 10 ml and added with the same volume of pentane.

The light red air-sensitive precipitate of 8a that formed

was ﬁltered (140 mg), washed three times with pentane

(5 ml) before being dried in vacuo. Yield 62%. Anal.

Calc. for C₈₆H₈₆Cl₂N₂P₄Fe₂Pd₁: C, 66.23; H, 5.56; N,

C₅H₄N/H_ortho); 8.07 (d, 4H isomer 2, ³J_HH=6.8 Hz,

C₅H₄N/H_ortho); 7.71 (m, 4H isomer 1+2, 8C₆H₄/

H_ortho); 7.33 (m, 68H isomer 1+2, 2C₆H₅+2C₆H₅/

H

_meta+para+2C₆H₅/H_ortho); 6.41 (d, 4H isomer 1,

³J_HH=7.0 Hz, C₅H₄N/H_meta); 6.40 (d, 4H isomer 2,

³J_HH=6.8 Hz, C₅H₄N/H_meta); 2.56 (m, 8H isomer 1+

2, CH_2dppe); 2.02 (m, 8H isomer 1+2, CH_2dppe); 1.41 (s,

30H isomer 1+2, C₅(CH₃)₅). ¹³C{¹H} NMR (50 MHz,

2

CDCl₃) l_C175.5 (t, CꢀC(C₅H₄N) isomer 1, J_PC=38

Hz); 171.4 (t, CꢀC(C₅H₄N) isomer 2, J_PCca. 38 Hz);

151.6 (s, J_CH=180 Hz, J_PtC=132 Hz, C₅H₄N/C_orthoN

isomer 1); 150.8 (s, ¹J_CH=182 Hz, ²J_PtC=137 Hz,

C₅H₄N/C_orthoNisomer 2); 139.0 (s, C₅H₄N/C_ipsoisomer

1+2); 138.4 (m, 2C₆H₅/C_ipsoisomer 1+2); 136.8 (m,

2C₆H₅/C_ipsoisomer 1+2); 134.2 (m, ¹J_CH=162 Hz,

2

1

2

1

1.80. Found: C, 65.07; H, 5.32; N, 1.81%. H NMR

(200 MHz, CD₂Cl₂) l_H8.13 (d, 4H, ³J_HH=5.5 Hz,

C₅H₄N/H_ortho); 7.72 (m, 8H, 2C₆H₄/H_ortho); 7.33 (m,

32H, 2C₆H₅+2C₆H₅/H_meta+para+2C₆H₅/H_ortho); 6.47

(d, 4H, ³J_HH=5.5 Hz, C₅H₄N/H_meta); 2.60 (m, 4H,

CH_2dppe); 2.05 (m, 4H, CH_2dppe); 1.40 (s, 30H,

C₅(CH₃)₅). ¹³C{¹H} NMR (50 MHz, CD₂Cl₂) l_C173.3

(t, CꢀC(C₅H₄N), ²J_PC=37 Hz); 151.3 (s, ¹J_CH=178

4C₆H₅isomer 1+2); 129.8 (s, J_CH=160 Hz, 2C₆H₅/

1

C_paraisomer 1+2); 127.9 (m, ¹J_CH=161 Hz, 4C₆H₅

1

3

isomer 1+2); 126.5 (s, J_CHnot obs., J_PtCnot obs.,

1

3

C₅H₄N/C_metaNisomer 1); 126.0 (s, J_CH=167 Hz, J_PtC

not obs., C₅H₄N/C_metaNisomer 2); 122.1 (s,

CꢀC(C₅H₄N) isomer 1); 121.5 (s, CꢀC(C₅H₄N) isomer

1); 89.2 (s, C₅(CH₃)₅isomer 1); 89.2 (s, C₅(CH₃)₅isomer

1

Hz, C₅H₄N/C_orthoN); 139.3 (s, C₅H₄N/C_ipso); 138.3 (m,

2); 30.9 (m, CH_2dppeisomer 1+2); 10.4 (s, J_CH=127

1

2C₆H₅/C_ipso); 137.0 (m, 2C₆H₅/C_ipso); 134.2 (m, J_CH

=

Hz, C₅(CH₃)₅isomer 1+2). MS (LSIMS⁺, m-NBA,

low resolution mode) m/z 2305 (M+3a–Cl+2, 2%);

1875 (M–Cl+PPh₃+2, 5%); 1648 (M+2, 1%); 1477

(M+PPh₃-dppe-Cl, 2%); 1250 (M-dppe+2, 1%); 957

(M–3a+2, 1%); 886 (M–3a–2Cl+1, 3%); 691 (3a,

100%); 589 (3a–‘CꢀC(C₅H₄N)’, 60%).

1

162 Hz, 4C₆H₅); 129.7 (s, J_CH=160 Hz, 2C₆H₅/C_para);

129.6 (s, ¹J_CH=160 Hz, 2C₆H₅/C_para); 127.8 (m,

¹J_CH=162 Hz, 4C₆H₅); 125.7 (s, ¹J_CH=166 Hz,

C₅H₄N/C_metaN); 121.6 (s, CꢀC(C₅H₄N)); 89.1 (s,

1

C₅(CH₃)₅); 30.8 (m, CH_2dppe); 10.2 (s, J_CH=127 Hz,

C₅(CH₃)₅). MS (LSIMS⁺, m-NBA, low resolution

mode) m/z 2217 (M+2+3a-Cl, 2%); 1787 (M+2-

Cl+PPh₃, 1%); 1561 (M+2, 1%); 1488 (M–2Cl, 1%);

797 (M–3a–2Cl, 3%); 691 (3a, 100%); 589 (3a–

‘CꢀC(C₅H₄N)’, 60%).

2.15. Preparation of the di-oxidized complexes

(M=Pd, 8a^{+ +}; Pt, 10a^{+ +}

)

A total of 0.95 equiv. of [(h⁵-C₅H₅)₂Fe][PF₆] was

added to a solution of 110 mg of the corresponding

compound 8a or 10a in 5 ml dichloromethane. Stirring

was maintained 30 min at room temperature and the

solution was concentrated in vacuo to ca. 1 ml. Addi-

tion of 40 ml of n-pentane allowed precipitation of a

dark solid. Decantation and subsequent washing with

3×3 ml portions of toluene followed by 3×3 ml

diethylether, drying under vacuum yielded the desired

8a⁺(75%) or 10a⁺(50%) salt as a dark brown

product.

2.14. Preparation of [(p²-dppe)(p⁵-C₅Me₅)Fe–CꢀC-p-

(C₅H₄N)]₂Cl₂Pt (10a)

The mixture of platinum isomers 10a was isolated

similarly to 8a, but after 45 min reaction time, starting

from 285 mg of (h²-dppe)(h⁵-C₅Me₅)Fe–CꢀC-p-

(C₅H₄N) (3a, 0.410 mmol) and 92 mg of (PhCN)₂PtCl₂

(0.195 mmol). The light red precipitate that formed was

ﬁltered (180 mg), dried in vacuo, before being dissolved

in 5 ml of dichloromethane and ﬁltered through an

alumina/celite pad which was subsequently rinsed with

three fractions of 5 ml of dichloromethane. After evac-

uation of the solvent, the red residue was washed with

pentane (5 ml), then three times with acetonitrile (3 ml)

and three times with diethylether (5 ml) before being

dried in vacuo, giving 70 mg of product 10a as an

air-sensitive light red powder. Yield 22%. Anal. Calc.

for C₈₆H₈₆Cl₂N₂P₄Fe₂Pt₁: C, 62.63; H, 5.26; N, 1.70.

3. Results and discussion

3.1. Synthesis and characterization of the ligands 3a–c

and their oxidized states 3a–c⁺

Recently we have reported the use of the Sonogashira

coupling between the (dppe)Cp*Fe–CꢀCH complex 1

408

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

bond (see Table 1)². The two ring-stretch absorptions

expected for the pyridyl ring (w₈pseudosymmetry A_g/

B_1gmodes) [42–44] were identiﬁed for 3b and 3c,

whereas only one was observed for 3a, the second

appearing possibly as a very weak intensity absorption

shoulder. Although the extinction coefﬁcient of the

these vibrational modes has not been estimated [45],

their high relative intensity in the spectra of all com-

plexes would be in accordance with the electron releas-

ing nature of the iron-alkynyl substituent³[46–49].

Curiously⁴, the IR absorption corresponding to the

triple bond of 3a–c presents a ﬁne structure with three

maxima for each compound, whereas only a single

peak would be expected for this oscillator based on

symmetry considerations (A% mode). Since this ﬁne

structure is maintained in solution, it is therefore not

Scheme 2.

[39] and various haloaromatic units [40]. In particular

with 2-,3- and 4-bromopyridines, such a reaction al-

lowed us to isolate with good yields (70–80%) the

desired pyridyl-functionalized complexes 3a–c (Scheme

2) [38]. These complexes were characterized fully by

means of usual spectrometric and voltammetric analy-

ses. The purity of the isolated complexes was assessed

by correct C, H, N elemental analyses and, in the case

of 3a, the corresponding molecular ion was observed in

LSIMS⁺. Consistently, the fragmentation pattern ob-

served on the MS spectrum is reminiscent of the ones

obtained for related complexes reported previously

[41].

²Slight changes can be stated with previously published values [38].

The current values are the deﬁnitive ones.

³Worth noting, is the fact that all these w₈frequencies are shifted

to lower wavenumbers with regard to the expected theoretical posi-

tions for purely inductive organic substituents [46–49]. The character-

istic frequencies found here for the substituted pyridyl moiety are

reminiscent of 4-chloro, 2-chloro and 3-cyano substituents for 3a, 3c

and 3b, respectively. This shift may be diagnostic of the presence of

the electron-releasing organometallic iron-alkynyl substituent which

strongly interacts in a p-fashion with the pyridyl fragment [46,50,51].

⁴Such a feature was not reported for the analogous (L)₂CpRu–

CꢀC-4-Py (L=phosphine) complexes [52].

Infrared absorptions characteristic of the pyridyl

unit were diagnostic of the success of the coupling

reaction in the complexes 3a–c, while a strong alkynyl

stretching mode evidenced the presence of the triple

Table 1

Selected IR frequencies (cm⁻¹) for compounds 3a–c/3a–c⁺

Compound w_CꢀCin Nujol/CH₂Cl₂

w_8a(Py) in Nujol/ w_8b(Py) in Nujol/ w_PFor w_CONujol/CH₂Cl₂

6

CH₂Cl₂

3a

2058(vs)/2055(vs) 2029(vs)/2030(vs)

1582(vs)/1584(vs) 1563(sh) ^a/n.o. ^b

2015(vs)/2015(vs)

1991(vw)/1989(vw)

2045(vs)/2057(vs) 2014(vs)/2032(s)

2003(vw)/2005(vw)

2057(m)/2063(m) 2041(sh)/2047(sh)

2022(vs)/2028(vs)

3a⁺

3b

1584(s)/1586(s)

1569(s)/1570(s)

1575(m)/1575(m) 1559(w)/1558(w) 840(vs)/846(vs)

1577(vs)/1580(vs) 1545(m)/1546(m)

1575(w) ^a/n.o. ^b

1548(w)/1551(w)

840(vs)/846(vs)

3b⁺

3c

3c⁺

5a

1993(vw)/1991(vw)

2017(m)/2015(s)

1580(s)/1581(s)

1600(m)/1599(s)

1559(m)/1557(m) 839(vs)/847(vs)

n.o. ^b

2070(m)/2069(m) 1940(vs)/1930(vs) 1918(vs)

1896(s)/1884(s)

2069(s)/2071(m) 1922(vs)/1928(vs) 1887(sh)/

1893(sh) 840(vs)/846(vs)

5a⁺

5b

1976(m)/1971(sh)

2029(m)/2030(m)

n.o. ^b

1585(s)/1587(m)

n.o. ^b/1603(m)

1575(m)/1575(m) 1555(m)/1555(m) 2067(m)/2069(m) 1927(vs)/1926(vs) 1907(vs)

1877(s)/1884(s)

1585(w)/1586(m) 1570(w)/1569(m) 2071(m)/2072(m) 1922(vs)/1928(vs) 1879(vs)/

1891(vs) 840(vs)/846(vs)

5b⁺

8a

2064(sh)/2067(sh) 2011(s)/2014(s)

2030(m)/2055(sh) 2034(w) 2020(sh)

2069(sh)/2071(sh) 2010(s)/2011(s)

2030(m)/2055(sh) 2034(w) 2020(sh)

1598(s)/1599(s)

1601(s)/1608(s)

1600(s)/1602(s)

1604(s)/1606(s)

n.o. ^b

8a⁺⁺

10a ^c

10a⁺⁺

1573(sh)/n.o. ^b

n.o. ^b

836(vs)/846(vs)

832(vs)/846(vs)

1574(sh)/n.o. ^b

^aTentative assignment.

^bNot observed.

^cSame for both isomers in the 4000–400 cm⁻¹range, see text.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

409

3

atoms of the dppe appears as a triplet with a J_PC

coupling constant of ca. 38 Hz. In ³¹P NMR, as

expected, one single signal is observed for the chemi-

cally equivalent dppe phosphorus atoms and the

weak differences of shift among 3a, 3b and 3c indi-

cate that differences in substitution of the pyridyl

ring are only weakly sensed by the phosphorus nu-

clei.

ascribable to packing effects in the solid state. Addi-

tionally, no conformers, isomers or oligomers can ex-

plain it, as conﬁrmed by NMR spectra (see later). We

attribute it to Fermi coupling with other combination

modes or overtones of similar symmetry involving ad-

jacent atoms. Identiﬁcation of the modes involved in

this coupling has not been attempted. Fermi coupling

with the triple bond stretch is not unusual and has

already been observed in related compounds

[40,41,53]. The structure of the various complexes was

further conﬁrmed by apparition of ad hoc proton and

carbon resonances in the ¹H NMR and ¹³C NMR

spectra, respectively, of the complexes (Tables 2a and

2b). For instance, observation of an AB system for

Cyclic voltammetry (CV) of the complexes 3a–c

(Table 3) in the range 1.4 to −1.6 V presents for all

compounds a reversible redox process around 0 V

relative to SCE. Such a process is typical of an iron-

centered oxidation [28]. Relative to the complex

(dppe)Cp*Fe–Cl [54], the alkynyl-pyridine ligand be-

haves as an attractor, rendering the oxidation of the

iron nucleus more difﬁcult [28,40,41]. Similar obser-

vations have been made for pyridyl-substituted fer-

rocenes [21,29,30,34]. This oxidation is easier

following the substitution order meta\ortho\para,

which corresponds to the electron releasing power of

the pyridyl ring at the functionalized position [45,55].

If we consider that the solvatation energies involved

in dichloromethane will be the same for 3a–c, this

1

3a in the H NMR spectrum indicated para-substitu-

tion of the pyridyl moiety. This was also evidenced

by observation of three characteristic types of pyridyl

carbon in ¹³C NMR, whereas more signals, indicative

of respectively meta- and ortho-substitution were ob-

served for 3b and 3c. The quaternary alkynyl carbon

atoms of 3a–c were also evidenced by characteristic

¹³C NMR signals. For all samples, the a carbon reso-

nance due to coupling with the two phosphorus

Table 2a

Selected ¹H and ³¹P{¹H} NMR data (ppm) for compounds 3–10 in CDCl₃

Compound H_ortho/Nof pyridyl ring (³J_HH

)

H_meta/Nof pyridyl ring (³J_HH

)

H_para/Nof pyridyl ring (³J_HH

)

H of Cp* P of dppe

3a

3b

3c

5a

8.20 (d, 5.8 Hz)

8.10 (d,4.4 Hz) 8.05 (s)

8.33 (m, 5.1 Hz)

8.21 (d, 5.8 Hz)

8.21 (d, ca. 5 Hz); 8.11 (s)

8.26 (d, 5.5 Hz)

6.62 (d, 5.8 Hz)

1.42

1.45

1.41

1.42

1.37

1.41

100.6

100.7

101.1

98.8

100.2

100.0

99.8 ^c

99.6

6.97 (dd, 4.4 and 7.8 Hz)

6,67(m, 7.95) 6.73 (m, 7.35) ^c

6.41 (5.8 Hz)

7.04 (7.9 Hz)

6.36 (d, 5.5 Hz)

7.05 (d, 7.8 Hz)

n.o. ^a

5b

6.85 (dd, ca. 5 and 7.9 Hz)

8a

10a ^b

10a ^d

8.30 (d, 7.0 Hz)

8.07 (d, 6.8 Hz,

6.41 (d, 7.0 Hz)

6.40 (d, 6.8 Hz)

^aNot observed.

^bPossibly trans-isomer, see text.

^cTentative assignment.

^dPossibly cis-isomer, see text.

Table 2b

Selected ¹³C NMR data (ppm) for compounds 3-10 in CDCl₃

Compound

C_a(²J_CP

)

C_b(³J_CP

)

C_ipso(Py)

C_ortho/N(Py)

Other C_Py

3a

3b

3c

5a

157.7 (38 Hz)

147.3 (40 Hz)

148.9 (38 Hz)

172.3 (37 Hz)

158.7 (38 Hz)

173.3 (37 Hz)

175.5 (38 Hz)

171.4 (ca. 38 Hz)

119.6 (n.o. ^a)

116.5 (n.o. ^a)

122.4 (2 Hz)

121.7 (5 Hz)

115.4 (n.o. ^a)

121.6 (n.o. ^a)

122.1 (n.o. ^a)

121.5 (n.o. ^a)

138.1

n.o. ^a

148.1

138.5

129.3

139.3

139.0

138.4

149.4

152.0 143.1

149.7

154.9

158.1 148.8

151.2

125.1

136.6 123.1

135.2 125.8 117.5

126.3

137.6 124.9

125.7

5b

8a ^b

10a ^c

10a ^d

151.6

150.8

126.5

126.0

^aNot observed.

^bIn CD₂Cl₂.

^cPossibly trans-isomer, see text.

^dPossibly cis-isomer, see text.

410

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

Table 3

a

Cyclic voltammetry for compounds 3–10 in CH₂Cl₂

Compound

E₀(DE_p, i_pa/i_pc) (V)

Fe(IV)/Fe(III) and other Fe(III)/Fe(II)

3a

3b

3c

5a

5b

8a

10a

1.41 ^b(i)

1.34 ^b(i)

1.31 ^b(i)

1.19 ^b(i)

1.45 ^b(i)

B1.44 ^b(i)

1.18 ^b(i)

−0.03 (0.09, 1)

−0.11 (0.09, 1)

−0.08 (0.09, 1)

0.06 (0.09, 1)

−0.01 (0.07, 1)

0.04 (0.10, 1)

0.07 (0.08, 1)

Scheme 3.

empty p* orbital of the pyridyl ring (possibly the LUMO

in these compounds) is observed at very cathodic

potentials⁵.

^aConditions: [n-Bu₄N][PF₆], 0.1 M, 20°C relative to SCE cali-

brated with ferrocene at 0.460 V, Pt electrode, sweep rate 0.100 V

s⁻¹

^bOxidation peak potential.

.

These complexes possess a Fe(III) redox state (3a–c⁺)

in an accessible potential range as suggested by the

observation of a reversible wave (Table 4) [28]. Accord-

ingly, this mono-oxidized state is stable and can success-

fully be isolated after chemical oxidation using

ferricinium hexaﬂuorophosphate (Scheme 3).

Table 4

UV–Vis spectra in CH₂Cl₂for compounds 3–10 ^a

Compound

Absorptions (nm) (10⁻³· o in M⁻¹cm⁻¹

)

The purple oxidized compounds have been character-

ized by IR and UV–Vis spectrometry and elemental

analysis. The cyclic voltammograms of the salts 3a–

c⁺are superimposable on the voltammograms of their

neutral precursors 3a–c, assessing their identity. These

radicals are quite stable in solution at room temperature

under an inert atmosphere. For instance the UV–Vis

spectrum of 3a⁺shows no marked changes after 12 h.

The triple bond stretching mode in 3a–c⁺appears

extremely weak in the IR spectra, and is barely visible

(see Table 1). However, no other vibrational mode,

ascribable for instance to a vinylidene stretch, is present⁶.

Consequently, the triple bond in these organometallic

radicals seems preserved, but the IR absorption of its

stretching motion is weak and shifted toward lower

wavenumbers. This shift, indicative of a decreased bond

order, is difﬁcult to evaluate with precision due to the

multiple absorption pattern present in the neutral com-

plexes. Such shifts upon oxidation are typical of iron

alkynyl complexes [23,24,27,28,39,53] and were earlier

attributed to the increased weight of the vinylidene

limiting resonant structure upon oxidation [27]. They will

be discussed in more detail elsewhere [41]. The weakness

of intensity can be explained by the fact that the polarity

of this bond is decreased strongly after oxidation of the

iron center, resulting in a very weak dipole transition

moment for the alkynyl stretching mode in IR. Similar

inﬂuences of particular substituents on triple bond

3a

260 (sh, 23.2); 403 (7.9)

272 (sh, 27.3); 312 (sh, 15.9); 339 (sh, 6.4);

414 (1.9); 521 (1.6); 591 (sh, 1.1)

3a⁺

3b

265 (sh, 25.4); 363 (10.7); 401 (sh, 8.3)

273 (sh, 27.4); 307 (sh, 16.3); 354 (sh, 5.5);

450 (1.8); 559 (2.0); 634 (1.9)

265 (sh, 23.6); 374 (9.0); 391 (sh, 8.9)

257 (sh, 30.8); 272 (sh, 29.4); 306 (sh, 19.5);

348 (sh, 6.5); 408 (sh, 2.1); 557 (1.6); 632 (1.5)

325 (sh, 10.2); 415 (sh, 14.1); 462 (18.5)

319 (sh, 15.1); 384 (15.6); 500 (sh, 1.9); 698

(sh, 0.7)

3b⁺

3c

3c⁺

5a

5a⁺

5b

284 (sh, 19.1); 379 (14.4)

5b⁺

274(sh, 31.3); 322(sh, 15.3); 355(sh, 8.9); 393

(sh, 5.9); 441(sh, 2.3); 512(sh, 1.3); 612(sh, 0.8)

257(sh, 115.0); 264(88.9); 328(1.4); 389(1.3)

271 (sh, 39.4); 407 (sh, 18.5); 461 (29.7)

267 (sh, 113.9); 342 (sh, 53.0); 456 (42.0)

266 (sh, 82.0); 349 (49.9); 421 (sh, 6.9); 512 (4.6);

593 (sh, 2.1)

7

8a

8a⁺

8a⁺⁺

10a ^b

263 (sh, 75.0 ^c); 297 (sh, 31.0 ^c); 415 (sh, 14.2 ^c); 490

(24.6 ^c)

10a^{+ b}

266 (sh, 128.2 ^c); 354 (sh, 45.2 ^c); 491 (23.7 ^c)

267 (sh, 97.0 ^c); 355 (53.0 ^c); 431 (sh, 6.5 ^c);

507 (4.0 ^c); 605 (sh, 1.8 ^c)

10a^{++ b}

^aCH₂Cl₂, absorbance mode, 20°C (Cary 5).

^bMixture of isomers, see text.

^cQualitative value only.

ordering suggests that electronic conjugation between the

iron nucleus and the electronegative nitrogen atom of the

cycle may inﬂuence strongly the energy level of the

HOMO, stabilizing it in the more favored para case 3a.

Additionally, irreversible redox events are observed at

high potentials which could be ascribed to the generation

of an unstable iron(IV) complex [56]. No irreversible

process ascribable to a ligand-based reduction of the

⁵Note that neither free pyridine, dppe, nor substituted pyridine as

4-t-Bu-pyridine or 2-picoline exhibit any redox process between 1.4

and 1.6 V.

⁶Such complexes could be formed by radical abstraction from the

solvent, however cyclic voltammograms exhibit a reversible process in

the range −50 to −150 mV which indicates that no iron(II)

vinylidene complexes are present. These complexes usually give rise to

irreversible oxidations [57].

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

411

formulae of these complexes could nevertheless be con-

ﬁrmed by LSIMS⁺spectrometry.

The presence of the tungsten pentacarbonyl center was

evidenced by apparition of the three intense carbonyl

stretching modes (2 A₁and E) expected for an M(CO)₅

C₄₆fragment in FTIR [60]. The most intense of these

modes is split in the spectra of 5a and 5b due to solid

state effects (Table 1)⁸. The presence of the triple bond

in 5a and 5b was evidenced by the presence of an

absorption in the 2020–2030 region of their spectra,

however, the ﬁne structure present in 3a and 3b was no

longer apparent. These absorptions are shifted ca. 15–20

cm⁻¹relative to their position in the spectra of the free

3a and 3b, suggesting a slight decrease in the triple bond

order upon coordination (Scheme 5). A similar shift upon

complexation to W(CO)₅was reported for the related Ru

analogs [52]. A shift toward higher wavenumbers is also

visible for the stretching modes of the pyridyl ring. This

shift is more pronounced for complex 5a than for 5b (ca.

15 versus 5 cm⁻¹), suggesting a more important interac-

tion between the heterocycle and the tungsten complex

in 5a. Regarding the other absorptions, the spectra of the

complexes 5a and 5b are similar those ones of the

corresponding free ligands 3a and 3b, the only difference

being the presence of weak absorptions around 750 cm⁻¹

arising from carbonyl bending motions on the tungsten

center.

Scheme 4.

stretch intensities have already been reported for un-

charged organic alkynes [58,59]. Oxidation of 3a–c also

induces in a very slight shift (ca. 5 cm⁻¹) toward higher

wavenumbers of the most intense w₈ring-stretch mode of

the pyridyl ring, and a slightly more intense shift (ca. 10

cm⁻¹) for the other. It seems therefore that the oxidation

of the iron nucleus does not markedly affect the bonding

within the heteroaromatic ring in 3a–c/3a–c⁺. Finally,

the presence of the hexaﬂuorophosphate anion can also

be inferred from the speciﬁc absorptions around 840 nm.

The new iron(II) alkynylpyridine 3a–c complexes

isolated present features characteristic of substituted

pyridines possessing an electron-releasing substituent.

Their oxidized state with a paramagnetic iron(III) center

3a–c⁺are stable and can be isolated. On the whole, these

new iron-pyridylalkynyls resemble their homologs with-

out any heteroatom in the aromatic cycle, but present the

advantage of having a Lewis-basic site amenable to

complexation [28,40,41]. Compounds very similar to 3a

were reported recently by Wu et al. with ruthenium,

however, these were only characterized in their neutral

state [52].

1

In the H NMR spectra of 5a and 5b, complexation

induces slight shifts on the ortho- or meta-protons of the

pyridyl rings relative to 3a and 3b, respectively, but leaves

the other resonances quite unaffected (Table 2a) [34]. The

shift of the ortho-N protons in 5a relative to 3a contrasts

with the larger shifts reported habitually [31,64]. These,

as the unusual downﬁeld shift of the meta-N protons

might tentatively be attributed to ring current shieldings

of the dppe phenyl groups, despite the latter being quite

far apart. In ³¹P NMR, the phosphorus atoms of dppe

appear weakly shielded after complexation. An opposite

trend was previously observed for the related ruthenium

analogs mentioned previously [52]. The ¹³C NMR spec-

trum clearly indicates the presence of the tungsten

pentacarbonyl center in 5a and 5b. Two carbonyl reso-

nances appear at 203.5 ppm (¹J_WC=150 Hz) and 199.6

ppm (¹J_WC=131 Hz) for 5a or 203.3 ppm (¹J_WC=150

Hz) and 199.6 ppm (¹J_WC=131 Hz) for 5b, surrounded

by the characteristic ¹⁸³W satellites [34]. The observation

of only two different types of carbonyl resonances with

5b indicates that the asymmetry induced by the meta-sub-

stituted 3b ligand is not detected on the other ligands

coordinated to tungsten, possibly due to ﬂuxionality in

solution. The triple bond is also characterized by typical

resonances (Table 2b). In comparison with 3a or 3b, the

quaternary C_aatom has been deshielded by 15 ppm in

5a and 11 ppm in 5b. This is remarkable when consider-

ing that the complexation site is quite far apart, and

shows that complexation of the nitrogen is sensed near

3.2. Complexation reactions of the ligands 3a–b with

tungsten

In order to test the coordinating ability of these

‘electroactive ligands’ toward a metal center, we have

used tungsten hexacarbonyl in a ﬁrst instance, since this

metal is known to give stable complexes with pyridine

[31,32,34,52,60,61]. Moreover the W(CO)₅center consti-

tutes an interesting spectroscopic probe via its carbonyl

ligands, allowing comparison with existing data to be

made. Anticipating steric complications with the ortho-

substituted compound 3c, we have limited our investiga-

tions to 3a and 3b.

Synthesis of the neutral complexes 5a–b was accom-

plished easily by reacting 3a–b and tungsten hexacar-

bonyl (4) under UV irradiation (Scheme 4). After

photochemical displacement of one carbonyl from the

tungsten center, the desired orange–red complexes 5a–b

were isolated in good yields after sublimation of the

unreacted tungsten hexacarbonyl. Unfortunately, com-

plexes 5a and 5b were contaminated by very small

amounts of another tungsten-containing complex (less

than 3%) that we were unable to separate⁷. The elemental

⁷Based on the spectroscopic similarities observed, we believe these

purities to be the corresponding di-substituted tetracarbonyl complexes

([Fe]–CꢀC–Py)₂W(CO)₄[60].

⁸Some very weak bands belonging to the theoretically forbidden

B₁stretching mode [62] appear as shoulders near 1970 cm⁻¹in 5a

[34,63].

412

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

Scheme 5.

the iron atom. The carbon atoms in ortho position to

the nitrogen are affected as well, however less

markedly. The other ¹³C resonances for the pyridyl ring

carbon atoms are in accordance with reported data for

pyridyl-W(CO)₅complexes [34,64].

Considering the limiting resonant structures depicted

in Scheme 5 [27], the data can be interpreted by an

increased weight of the vinylidene-like limiting structure

II upon complexation, stabilizing the HOMO in the

neutral complexes and rendering thereby the iron center

less oxidizable. This would also be fully consistent with

the decrease in triple bond order induced by

complexation.

Cyclic voltammetry indicates noticeable changes in

the redox potential of 5a or 5b relative to the uncoordi-

nated ligands 3a or 3b, respectively (Table 3). The

Fe(II)/Fe(III) potential is shifted anodically, whereas

the second irreversible redox event appears shifted ca-

thodically relative to the one observed for 3a–b. The

shift for the Fe(II)/Fe(III) potential indicates that com-

plexation with the tungsten pentacarbonyl fragment has

rendered the oxidation of the iron center more difﬁcult

by 0.08–0.10 V regardless the substitution pattern of

the heterocycle. Similar trends upon complexation, but

with less pronounced shifts, have been observed with

ferrocene-functionalized 4-pyridyl acetylenes [31], while

comparable shifts were reported for the related analogs

of 3a/5a with ‘(L)₂CpRu–CꢀC’ (L=Phosphine) sub-

stituents [52]. This can be explained by an increased

‘electron deﬁciency’ of the pyridyl moiety after coordi-

nation of the nitrogen free doublet to the tungsten. The

other redox process, attributed previously to a Fe(III)/

Fe(IV) oxidation, should also be more difﬁcult now and

consequently shifted toward higher potentials relative

to 3a–b. The opposite trend is observed, thus we be-

lieve that the second potential observed corresponds

now to a new redox process centered on tungsten and

that the Fe(III)/Fe(IV) potential is no longer visible in

the potential range investigated [31,32,34,52,60,63–65].

Note also that, despite the fact that coordination of the

tungsten has rendered apparently the pyridyl moiety

more electron accepting, the pyridyl-based one-electron

reduction is still not observed for 5a–b⁹.

3.3. Oxidation of the tungsten complexes 5a–b

The observation of a reversible redox Fe(II)/Fe(III)

system suggested again that 5a or 5b might be stable in

the oxidized state. Accordingly, we attempted the isola-

tion of the oxidized salts 5a⁺and 5b⁺by chemical

oxidation of 5a and 5b using the ferricinium hex-

aﬂuorophosphate salt. However, since 5a⁺proved to

evolve slowly in solution, giving the corresponding

vinylidene¹⁰by hydrogen atom abstraction from the

solvent (Scheme 6), oxidation of 5a and 5b and isola-

tion of 5a⁺and 5b⁺had to be conducted at low

temperature. As solids, the dark green oxidized com-

plexes were quite stable at room temperature. They

were characterized by infrared, UV–Vis spectrometry

and CV. Solution spectrometric measurements in solu-

tion on 5a⁺and 5b⁺were recorded immediately after

their solubilization. The voltammograms of 5a–b⁺

were identical to those performed with the starting

complexes, while IR and UV–Vis spectrometry clearly

indicated that the isolated complexes were different

structurally.

In IR, a shift in the triple bond stretching mode to

lower wavenumbers of ca. 40 cm⁻¹relative to the

neutral complexes 5a–b is observed for 5a–b⁺. Com-

parison with 3a–b⁺is also instructive and indicates

that complexation of the tungsten center decreases fur-

⁹Such a reduction was reported to be near the solvent cutoff (1.8

V) for the analog of 5a with a less electron-releasing ferrocenyl group

[31], but, accordingly with our results, was not detected with the more

electron-releasing ‘(PPh₃)₂CpRu–CꢀC’ substituents [52].

¹⁰The formation of this product was proposed based on IR, ¹H

and ³¹P NMR data. Further work is in progress to characterize

deﬁnitively this evolution product of 5a⁺

.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

413

Scheme 6.

by reason of its kinetic instability in solution, 5a⁺

evolved during the course of the reaction and only the

corresponding complexed vinylidene was detected

among the products. This constitutes however an indi-

rect evidence for the intermediacy of 5a⁺and validates

our initial hypothesis.

ther the bond order of the acetylenic linker, thereby

enhancing the weight of the limiting resonant structures

II⁺and III⁺in the description of the molecule (Scheme

5). Regarding H-atom abstraction, the increased reactiv-

ity of 5a–b⁺in solution, relative to 3a–b⁺, and its

propensity to form the vinylidene follow the same lines

(Scheme 6). This suggests a pronounced radical charac-

ter on the b-acetylenic carbon after complexation and

favors of the increased weight of a vinylidene-like limit-

ing structure such as II⁺in the Lewis description of

5a–b⁺. On the other hand, the carbonyl stretching

modes are quite unaffected, indicating that the oxida-

tion is only sensed weakly by the tungsten atom. As a

consequence the triple bond stretching mode is masked

by the strong carbonyl vibrations in 5b, and appears

only as a shoulder in 5a. As stated for the oxidized free

ligands 3a–c⁺, the w₈pyridyl ring-stretching modes are

only sensitive slightly to oxidation. Here also, only a

slight shift toward higher wavenumbers is detected for

both modes with 5b⁺. The presence of the PF₆⁻counter

ion was evidenced by the apparition of a typical absorp-

tion around 840 nm. Below 1500 cm⁻¹, the spectra of

5a⁺and 5b⁺are very similar to the ones of the free

oxidized ligands 3a⁺or 3b⁺except for two weak

absorptions around 600 cm⁻¹, originating presumably

from carbonyl bending motions of W(CO)₅.

Thus, 3a–b react as typical pyridines and coordinate

strongly the tungsten pentacarbonyl center. Complexa-

tion results in characteristic modiﬁcations of the spec-

troscopic features and of the oxidation potential of the

iron–pyridyl alkynyl ligand. Notably, the redox poten-

tial shift may have some general character [21,29,31–

33], and may indicate that complexation is sensed

effectively by the iron terminus, which can be rational-

ized, along with IR data, by an increased weight of the

limiting vinylidene-type limiting structures II and II⁺in

Scheme 5. In the resulting complexes 5a–b the

organoiron(II) ligand can still be oxidized and the

corresponding oxidized complexes 5a–b⁺can be iso-

lated. This illustrates the potential of the compounds

3a–b to behave as redox ligands toward metal centers.

As mentioned previously for 3a, our results on 5a are

consistent with the data reported very recently by Wu et

al. on the ruthenium analogs. The corresponding oxi-

dized tungsten complexes were however not investigated

[52].

The oxidized complexes 5a–b⁺might be accessed as

well from the oxidized ligands 3a–b⁺(Scheme 6). In

order to check this hypothesis, we have attempted to

synthesize the complex 5a⁺by reaction of the pre-oxi-

dized ligand 3a⁺with (THF)W(CO)₅. Unfortunately,

3.4. Complexation reactions of the ligands 3a with

palladium or platinum centers

Having established that the 3a–b family behaves as

usual ligands with tungsten in the two redox states, we

414

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

Scheme 7.

Coordination of the ligand is further indicated by the

¹H and ¹³C NMR spectra where the pyridyl signals of

attempted to interconnect two 3a molecules using a

templating metal center with two coordinating sites. A

simple and easy choice for that purpose was to use

bis-benzonitrile chloride complexes of palladium (6)

and platinum (9) [66]. Indeed these complexes which

are prepared easily from the corresponding chloride salt

and benzonitrile [67,68] were reported to form bis-

pyridyl complexes [69–71]. In order to dispose of a

suitable reference complex and to check the efﬁciency

of the synthetic approach, we reacted a 2-fold excess of

4-tert-butyl-pyridine with 6 in dichloromethane at am-

bient temperature. An instantaneous reaction took

place, yielding the desired complex (4-t-Bu-Py)₂PdCl₂

(7). Complex 7 was isolated as an air-stable light yellow

1

the free ligand 3a have been shifted. Since H and ³¹P

NMR proved to be very sensitive to traces of oxidized

products with 3a (see later), the samples of 8a for NMR

were freshly washed by small portions of acetonitrile to

1

remove any traces of oxidized products. In H NMR,

the shift of the downﬁeld and the upﬁeld shifts for

ortho- and meta-protons, respectively, of the pyridyl

ligand are reminiscent of those stated for tungsten.

Here as well, the downﬁeld shift of the pyridyl ortho-

protons of this complex relative to 3a is quite weak (ca.

0.2 ppm) in comparison to values reported usually for

palladium pyridyl complexes [70,73,74]. Similarly to 5a,

they may be ascribed to shielding effects occasioned by

the phenyl groups of the dppe ligands [75]. In ¹³C

NMR, complexation is indicated by an upﬁeld shift of

ca. 15 ppm for the C_aof the alkyne, also reminiscent of

the one reported for the tungsten complex 5a. Finally,

and consistently with the data reported for 5a, the

phosphorus atoms of the dppe are slightly more

shielded upon complexation (0.6 ppm) in the ³¹P NMR

spectrum.

In principle, whereas complexes of various metal/lig-

and stoichiometries could have been formed, palladium

chloride forms usually neutral bis-pyridyl complexes

(A1–A2; Scheme 8). Dicationic tetra-pyridyl complexes

(B; Scheme 8) are also known [74,76], but are formed

generally under harsher conditions with an excess lig-

and present¹¹. The elemental analysis obtained for 8a is

quite close to the percentages expected for a di-pyridyl

1

product and characterized by IR, H NMR, LSIMS⁺

and elemental analysis (see experimental part).

Repeating the reaction with two molar equivalents of

3a in dichloromethane (Scheme 7) led to an instanta-

neous color change, and a new red complex 8a could be

isolated by precipitation (ca. 60%). Its IR spectrum,

despite being quite similar to that of 3a, presented a

slight shift in the triple bond stretching frequency to-

ward lower wavenumbers (ca. 20–25 cm⁻¹) and also

an opposite shift of the single observable w₈mode of the

pyridyl toward higher wavenumbers (ca. 10–15 cm⁻¹

)

[72]. Notably, the former absorption has lost its ﬁne

structure, only a slight shoulder around 2065 cm⁻¹

being visible. The very similarity of the changes relative

to free 3a with those reported previously for 5a is

indicative strongly of complexed 3a ligands in the new

compound. Importantly, the disappearance of the typi-

cal w_CNabsorption of complexed benzonitrile around

2250 cm⁻¹also indicates that this ligand is no longer

present in the coordination sphere of the palladium.

¹¹In the case of platinum, the dicationic Pt tetra-pyridyl complex is

converted to the neutral dipyridyl by heat or acidic media [77].

Scheme 8.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

415

red product (10a) in ca. 60% yield. Its IR spectrum was

identical to that of 8a and featured no absorption of

complexed benzonitrile (Table 1). LSIMS⁺of 10a gave

a spectrum with a very similar ion distribution to 8a.

Again, the molecular ion corresponding to a bis-adduct

with the expected theoretical isotopic pattern was deﬁn-

itively identiﬁed by high resolution mass measurements

at 1646.3460 amu relative to PEG. However at this stage,

apart from one main product, the ¹H, ³¹P and ¹³C NMR

spectra were indicative of the presence of many other

compounds amounting to roughly 25% of the mixture¹³,

with no free ligand 3a present. Thus, a more complex

puriﬁcation procedure was used and a cleaner product

containing essentially the previous most abundant spe-

cies, along with a second one (ca. 25%) was isolated with

a diminished global yield (20%). We were not able to

Scheme 9.

complex (much closer than for any other stoichiometry),

with correct C/H/N ratios. Moreover the ¹H NMR data

[74,78] and the isolated yield are also more consistent

with this formulation. Additional support for such a

structure comes from the LSIMS⁺spectrum of the

product. Indeed; (i) a molecular ion corresponding

actually to a neutral bis-pyridyl complex from its mass

and theoretical isotopic pattern is detected and deﬁni-

tively identiﬁed by high resolution measurements

(1559.3082 amu) relative to polyethylene glycol (PEG);

(ii) no di- or mono-charged ion corresponding to a

tetra-pyridyl adduct is detected; and (iii) most of the

other ions (C) arising from the ionization process in the

matrix are also present in the LSIMS⁺spectrum of the

neutral bis-pyridyl adduct 7 previously characterized as

a bis-adduct¹². Finally, additional evidence for this

stochiometry was also obtained with the di-oxidized

complexes 8a^{+ +}(see Section 3.6).

Coming now to the geometry of complex 8a, we favor

strongly the trans arrangement (A2; Scheme 8), since for

such bis-pyridyl dichloro palladium square planar com-

plexes, cis-isomers are known to be unstable and to

convert readily to the more stable trans-isomers at room

temperature, even in the solid state [66,79]. In order to

have additional data, we have undertaken a far-IR

measurement in order to detect palladium-chloride vibra-

tions. While a cis-complex (C₂₆geometry) should present

two IR-active stretching modes, a trans-complex (D₂₆

geometry) has only one IR-active absorption. The 400–

250 cm⁻¹region of the spectrum of 6a was only resolved

poorly and presented many weak absorptions, but a

single absorption at 365 cm⁻¹is consistent with our

structural hypothesis (A2; Scheme 8) [80–82].

1

separate these compounds, but their H, ³¹P and ¹³C

NMR features are reminiscent of these observed for 8a

(Tables 2a and 2b). Notably, in ¹H NMR, only one signal

for the Cp* is observed for the mixture, but two sets of

pyridyl signals and two singlets in ³¹P NMR. As for the

tungsten complex 5a, the slight downﬁeld shift (ca. 0.2

ppm) of the pyridyl ortho-protons of the major complex

relative to 3a is unusual [70,74]. It might also be ascribed

to shielding effects occasioned by the phenyl groups of

the dppe ligands in the cis structure [75]. The ¹³C NMR

data are consistent with published data regarding the

pyridyl platinum complexes [86]. These compounds are

believed to be the cis- (most abundant compound) and

trans-isomers (other compound) of 10a. Such a hypoth-

esis is likely, since both A1; (Scheme 8) and (A2)

structures are known to be stable in the case of platinum,

the cis-structure being the kinetic isomer [77]. Weak

absorptions ascribable to platinum–chloride vibration

modes for the cis- and trans-isomers of 10a could be

detected around 339 cm⁻¹and would be consistent with

a mixture of mainly cis- (two IR-active modes; 340 and

325 cm⁻¹) and trans-isomers (one IR-active mode: 345

cm⁻¹) [82,87]. As a logical consequence, the reported

UV–Vis data for this mixture will only have a qualitative

value since the extinction coefﬁcients measured do not

correspond to a single compound.

In the case of the bis-benzonitrile dichloroplatinum

precursor 9 (Scheme 9) [68,83], the reaction was very

similar in appearance and led to the isolation of a deep

Cyclic voltammetry of 8a and 10a establishes also that

complexation has occurred since a shift of the Fe(II)/

Fe(III) reversible oxidation potential toward more posi-

tive values is observed (Table 3 and Fig. 1). The

concomitant decrease in current intensity is indicative of

decreased diffusion coefﬁcients for 8a and 10a isomers

relative to 3a [88]. As for 5a, the shifts in the Fe(II)/

Fe(III) oxidation potentials are indicative of a decrease

¹²Notably the relatively intense signal corresponding to a mono–

chloro tris-pyridyl cation observed at 2217 amu (Scheme 8; C). This

ion is possibly formed in the NBA matrix by reaction of the very

reactive [L₂PdCl]⁺fragment (not detected) with traces of 3a arising

from decomposition of the complex. Indirect support for such a

hypothesis comes from the observation of a [L₂(PPh₃)PdCl]⁺ion at

1787 amu. The triphenylphosphine in 8a can only result from the

catalyst used in the synthesis of 3a and was not detected by NMR. It

can therefore only have been present as traces as well. This indicates

that the formation of such [L₂(PPh₃)PdCl]⁺or [L₃PdCl]⁺ions is

greatly favored in the matrix during the ionization process. Addition-

ally, a structure such as C would not be compatible with NMR data.

¹³Similar reactivity differences between related platinum(II) and

palladium(II) precursors with Lewis bases possessing a triple bond

have been reported already [75]. Apart from the cis- and trans-iso-

mers, we believe that p-ligated platinum alkynyl oligomers may be

formed in the medium [84,85].

416

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

anodic shift was less pronounced (ca. 55 mV). Notewor-

thy and in contrast to our complexes, their dioxidized

complex was not stable enough to be isolated. If the

interaction between the platinum dichloride center and

the pyridyl ligands is of the same magnitude in all these

complexes, the larger shift for the mixture of platinum

isomers 10a stated here may be related, either to the

different substitution pattern of the pyridyl ring (para-

vs. meta-) or to a more effective s-ligation of the pyridyl

to the iron nucleus. Considering that the shifts upon

complexation to tungsten of 3a and 3b were quite similar,

we rather favor the last explanation.

Fig. 1. Cyclic voltammograms of compounds 3a, 8a and 10a in

dichloromethane at 20°C (0.006 mmol, 0.1 M tetrabutylammonium

hexaﬂuorophosphate, scan rate: 0.1 V s⁻¹).

3.5. Stability of the palladium and platinum complexes

Many transition metal complexes of pyridine are

known to be labile, especially palladium and platinum

complexes which can isomerize over time. In order to

build molecular wires by coordination reactions, thermo-

dynamically stable complexes have however to be formed

[77,79,89]. Consequently, we have tried to probe the

stability of 8a and of the isomer mixture 10a in solution

by ¹H or ³¹P NMR measurements¹⁵. In ¹H NMR the Cp*

protons were barely distinguishable, leaving the weak

meta-pyridyl proton as the most diagnostic signal.

First, we have dissolved the complexes 8a and 10a in

in the electron density around the iron nuclei and along

with the IR data consistent with a more cumulenic

character entering in the Lewis description of the com-

plexed neutral ligands (Scheme 5). For 8a, the observa-

tion of a single wave for the two Fe(II)/Fe(III) processes

occurring within the same molecule indicates that these

redox event are not resolved and therefore must be

separated by less than 0.1 V. Similarly, only one redox

wave is observed with the mixture of platinum isomers

10a. This indicates additionally that the stereochemistry

of the complexes is only sensed weakly by the iron nuclei

in isomers 10a, and further evidences that the iron–iron

interaction will be weak. However, the difference in the

electronic nature of the Lewis acid interacting with the

pyridyl ligand is clearly sensed by the iron center. Thus,

the shift is more pronounced with platinum than with

palladium or tungsten, suggesting that platinum is a

stronger Lewis acid than palladium and tungsten in these

complexes and gives rise to a stronger interaction with

the pyridyl group.

A second irreversible event is observed with the

platinum complex 10a at 1.44 V and may be attributed

to the Fe(III)/Fe(IV) irreversible oxidation of iron center,

by analogy with 3a. A similar event can be detected in

the solvent edge with the palladium complex 8a by

formation of degradation products. Thus, complexation

seems to have only a weak effect on this irreversible

oxidation in the case of 10a. Alternatively, these waves

could also be attributed to the oxidation of the templat-

ing metal center, by analogy with 5a¹⁴. On the other hand

no oxidation was detected in this potential range for 6,

7 or for 9.

1

deuterated pyridine. For 8a, the H and ³¹P NMR data

were indicative of instantaneous ligand exchange and

subsequent quantitative release of free ligand 3a in

solution, but more than 4 days were required to reach

the same result with 10a, only ca. 30% of 3a being

dissociated after 30 min. Therefore the redox active

ligand 3a is labile in these compounds and this reaction

is more favored with the palladium center than with the

platinum center accordingly with existing data [91]. With

8a, attempts to displace this ligand with an excess (30

equivs.) of 4-tert-butyl pyridine resulted mostly in the

expected result and formation of 7, and also in the

formation of additional new and non-identiﬁed com-

plexes, possibly mixed-tetrapyridyl adducts (B; Scheme

8), resulting from the large excess of pyridine in solution.

Complexes 8a or 10a in CDCl₃show no indication of

free dissociated ligand (however small amounts would

possibly not have been detected) and no quantitative fast

exchange process (relative to the NMR time scale) that

would allow the observation of a mean species taking

place, as indicated by the fact that added free ligand can

be observed distinctly from 8a, with both sets of signals

apparently not shifted relative to pure samples in the

same solvent¹⁶. These observations suggest that the

Comparison with the electrochemical data reported by

Carugo et al. with 3-ferrocenyl substituted pyridines

coordinated in a cis-fashion to platinum dichloride (see

D in Scheme 12) is also instructive [30]. In their case the

iron centered oxidations were not resolved, but the

¹⁵For ³¹P NMR experiments, very pure samples of complex had to

be used in the absence of air since ³¹P signals were rapidly broadened

by minute amounts of oxidized product (resulting for instance in no

apparent signal for 8a when 5% of 8a⁺is present in the medium) [90].

¹⁶The added free ligand 3a in excess is consumed slowly (over days)

to form new unidentiﬁed complexes, presumably also tetra-pyridyl

adducts.

¹⁴This would be consistent with a more difﬁcult oxidation for the

palladium complex 8a than for the platinum isomers 10a.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

417

Comparison of the triple bond stretching frequency

in the di-oxidized complexes with that of the oxidized

ligand 3a⁺indicates that a strengthening of the

acetylenic bond has taken place upon complexation,

while the reverse trend had been stated for the neutral

complexes 8a and 10a relative to 3a. Likewise, this

contrasts with the IR data obtained from the oxidized

tungsten complexes 5a⁺(Table 1). Apparently, from

these stretching frequencies, a higher bond order is

present in the oxidized complexes 8a^{+ +}or 10a^{+ +}

relative to 3a⁺for the acetylenic linkage. This suggests

that upon coordination to dichloro-palladium or -plat-

inum complexes, oppositely to what was stated previ-

ously with tungsten pentacarbonyl, the weight of the

vinylidene-like limiting resonance structure (II⁺) in

Scheme 5 has decreased. This would result in a more

iron-centered radical and is consistent with the higher

chemical stability stated for the oxidized complexes

8a^{+ +}or 10a^{+ +}relative to 5a⁺.

stability constants of 8a or 10a are signiﬁcantly larger

than these of the mono- or tris-pyridyl adducts,

which are possibly the intermediates responsible for

the exchange process. In conclusion, while the ligand

3a remains labile in 7a and 10a, the stability con-

stant of these complexes is sufﬁciently large to ren-

der the bis-adduct the dominant structure in solu-

tion.

From 8a, the redox-active pyridyl ligand 3a can be

displaced very cleanly within minutes using an excess (5

equivs.) of triphenylphosphine, to give quantitatively

the known complex (PPh₃)₂PdCl₂(Scheme 10). This

constitutes an easy chemical means to ‘plug off’ the

redox ligand in the adduct 8a and thereby ‘switch off’

any communication.

3.6. Oxidation of the palladium (8a) and platinum (10a)

complexes

Similarly to the tungsten complexes, the di-cationic

complex palladium 8a^{+ +}was isolated after chemical

oxidation using 2 equiv. of ferricinium hexaﬂuorophos-

phate (Scheme 11). The complex 8a^{+ +}is stable in

solution for hours, as demonstrated by paramagnetic

NMR experiments [90], and the reaction can be per-

formed at 20°C. This new dark brown compound gives

an identical cyclic voltammogram to its neutral precur-

sor 8a, but in solution, the compound gives a much

less intense coloration than the neutral complex. It was

also characterized by IR and UV–Vis spectrometry.

Very similar results were obtained using the mixture of

isomers 10a, yielding presumably a mixture of the

The di-oxidized complex 8a^{+ +}can also be quanti-

tatively made by reaction of 2 equiv. of the oxidized

ligand 3a⁺and 1 equiv. of the palladium precursor 6

(Scheme 11). The resulting 8a^{+ +}was identiﬁed ﬁrmly

by comparison of CV, UV–Vis and IR spectra with

those of an authentic sample of 8a^{+ +}synthesized by

chemical oxidation of 8a. This paramagnetic com-

pound has been characterized by ¹H NMR spec-

troscopy as well [90]. As for 8a, an excess of

4-tert-butyl pyridine displaces readily 3a⁺from the

palladium center in 8a^{+ +}. Using this reaction, by

monitoring the speciﬁc absorption of the dioxidized

palladium complex at 349 nm (Table 4), we could

titrate a solution of 3a⁺of known concentration by

UV–Vis using a solution of (PhCN)₂Cl₂Pd (6) in

dichloromethane. This transition grew upon successive

additions of aliquots of 6 and reached a plateau when

0.5 equiv. total had been added. This demonstrates

deﬁnitively that our palladium complexes possess the

2:1 ligand–metal stoichiometry.

corresponding di-oxidized isomers 10a^{+ +}

.

Again, in the 4000–400 cm⁻¹range, the IR spectra

of 8a^{+ +}and 10a^{+ +}are identical, both in the solid

(Nujol dispersion) and in the liquid state, which indi-

cates that the presence of cis- and trans-isomers and

other minor compounds of similar structure is not

detected from the IR-active modes inﬂuenced by com-

plexation. The absorption due to the triple bond

stretching frequency appears quite weakly and is nar-

rower than for the neutral precursors 8a and 10a, but

remains clearly visible (see Table 1). Its frequency is

shifted slightly toward higher wavenumbers (ca. 15

cm⁻¹), and a ﬁne structure similar to the one observed

for 3a is visible in solution. Thus, the triple bond in

these organometallic radicals is retained, and seems

even reinforced by oxidation when the pendant pyridyl

is complexed to a late transition metal like palladium

or platinum. The single w₈pyridyl ring-stretching mode

observable is quite insensitive to oxidation. Below 1500

cm⁻¹, the spectra of 8a^{+ +}and 10a^{+ +}are very simi-

When mixing 3a⁺and the platinum precursor 9 the

reaction was less clean and overall much more slug-

gish. This result is not surprising considering the reac-

tion of neutral 3a in similar conditions. Once oxidized,

the ligand appears much less nucleophilic relative to its

reduced state and does not displace easily the ben-

zonitrile ligand. Moreover, 3a⁺is also more labile

than 3a on Pt. For instance in CV, with the plati-

num complex 10a^{+ +}decoordination of the ligand

3a⁺is apparent within minutes in the presence of

a

slight excess of 4-tert-butyl pyridine or benzonitrile.

For complexes 8a and 10a possessing now two 3a

ligands, a mono-oxidized MV state does exist. These

MV complexes can in principle be generated upon

mixing of equimolar amounts of the neutral and di-

lar to the spectrum of 3a⁺and the speciﬁc absorption

of the hexaﬂuorophosphate anion (around 840 cm⁻¹

is present.

)

418

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

Scheme 10.

the shift is signiﬁcant owing to the experimental preci-

sion of the measure (2 cm⁻¹). Consequently the MV

complex can be characterized by a speciﬁc IR signa-

ture in the solid state, while the presence of small

amounts of both 8a^{+ +}and 8a cannot be excluded

owing to the proximity of their signals. From the

isolated dark solid sample of 8a⁺, after drying in

vacuo, the red neutral complex 8a can be extracted by

trituration in toluene, leaving a dark red–brown

residue of 8a^{+ +}. This conﬁrms that equilibrium (1) is

readily taking place in solution.

oxidized precursors. Their concentration in solution

will depend on the comproportionation constant (K_c)

of the equilibrium which expression is given in Eq. (1)

[28]. From electrochemical data, we know that this

thermodynamic constant is weak, since DE° expressed

in volts (the difference between the two successive

iron-based oxidations occurring within the same

molecule) was weak (B0.1 V). However, even in the

case when this interaction is minimum, the statistic

distribution generates 1 equiv. of MV in solution, for

0.5 equiv. of neutral and di-oxidized complexes.

These results indicate that the MV complex 8a⁺

can be generated quantitatively by displacement of the

comproportionation equilibrium, when the solvent is

evacuated under vacuum. Very similar results were

obtained with the mixture of isomers 10a⁺. IR spec-

trometry, while indicating that the MV complex pre-

sents speciﬁc absorptions, gives however no clear

indication on the electron transfer rate, since no vibra-

tions typical of the Fe(III) and Fe(II) sites are visible

in the investigated range. The stated slight changes in

the triple bond stretching frequency may be related to

other factors than electronic delocalization be-

tween the iron redox sites, such as electrostatic inter-

actions in the crystal lattice with the hexaﬂuorophos-

phate ion. It is noteworthy that, only one w_CꢀC

stretching mode is apparent on the spectrum while

two should theoretically be observed for a localized

K

c

Ox–Ox+Red–Red = 2Ox–Red

0.059 log(K_c)=DE°

(1)

Using this property, we have generated in situ the

MV complex 8a⁺in the cell using a solution of the

di-oxidized 8a^{+ +}and neutral 8a complexes twice as

concentrated in solution and recorded the UV–Vis

and NIR spectra. The IR spectra of the deep red

solution obtained after mixing is a summation of the

spectra of 8a^{+ +}and 8a. However, more interestingly,

the IR spectrum of the resulting dark brown solid

after evaporation of the solvent presents a triple bond

stretching frequency which cannot result from the su-

perposition of the neutral and dicationic precursor

spectra (Fig. 2). This absorption is weakly shifted (7

cm⁻¹) toward lower wavenumbers relative to 8a and

Scheme 11.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

419

nonsymmetrical complex of C₂symmetry. This is how-

ever not in contradiction with a mainly localized struc-

ture for the MV complex. Conversely, an averaged

symmetrical structure due to fast electron-delocalization

relative to the IR time scale should give rise to a single

mean IR signal relative to the absorption of the neutral

and di-oxidized complexes, which is not the case here.

Moreover, since the w_CꢀCmode of the oxidized part in the

MV complex may have a much weaker intensity than the

other one, as stated for the all oxidized complexes

investigated here, it might be masked by the latter.

The spectrum of the oxidized 3a–c⁺complexes is very

different from their neutral precursors (Table 4 and Fig.

4) Again intense absorptions above the solvent edge

(around 260 nm) were observed, while the intense MLCT

visible transition has disappeared. The former are ascrib-

able to LC transitions and present many shoulders

extending into the near UV–Vis range. In appearance,

the whole manifold of far-UV transitions has been shifted

bathochromically of ca. 40 nm relative to 3a–c. Addi-

tionally, in the visible range, a succession of two to three

weak transitions starting around 600 nm overlap with the

background absorption resulting from the far-UV tran-

sitions. However weak, these transitions explain the dark

purple color of the compounds in solution. They might

be ascribed to ligand-to-metal charge transfer (LMCT)

transitions. Such transitions are likely to take place after

creation of an electronic vacancy in HOMO subsequent

to the oxidation (SHOMO)¹⁷. The bathochromic shifts

of the far-UV transitions ascribed to LC p–p* transitions

in the neutral 3a–b precursors indicate either a stabiliza-

tion of the empty ligand-based orbitals (p*) or destabi-

lization of the occupied ones (p) upon oxidation.

Disappearance of the MLCT transition is possibly due

to a hypsochromic shift of the latter which is now masked

by the previously mentioned far-UV absorptions. Such

a hypsochromic shift may be produced by stabilization

of the single occupied HOMO (SHOMO) upon oxidation

(no longer inter-electronic repulsion). Comparison of the

spectra of the three oxidized complexes 3a–c⁺indicates

a slight bathochromic shift of the three low energy

transitions in the order meta\ortho\para, (Fig. 4)

suggesting again the inﬂuence of resonance-like interac-

tions between the heteroatom of the aromatic cycle and

the iron center. A more precise attribution of the

observed transitions would require a detailed electronic

picture of these complexes [41].

3.7. UV–Vis and NIR spectroscopy of the ligands and their

complexes

In order to gain further insight into the electronic

structure of these complexes and on the modiﬁcations

brought about by complexation or oxidation, we inves-

tigated their UV–Vis and NIR spectra, while UV–Vis

will provide complementary information on the energetic

levels of the molecular orbitals (MOs). Additionally,

NIR spectra of compounds 5a–b⁺, both in solution or

in the solid may allow observation of a intervalence

transition (IVCT), diagnostic of intramolecular electron

exchange.

Complexes 3a–c exhibit a broad absorption around

400 nm (Fig. 3) which explains the bright orange color

of the complexes, while other more intense absorptions

start above 300 nm. Their maximum could not be

determined due to the solvent absorption edge, but such

high energy processes can safely be ascribed to ligand

centered (LC) p–p* electronic transitions. Absorptions

centered in the near UV–Vis range were much more

interesting for us. Whereas the broad absorption ob-

served at 403 nm with the para complex 3a shows only

one maximum, it may be composed of two overlapping

absorptions in the case of 3b and 3c as suggested by the

presence of a slightly apparent shoulder. Considering a

principally metal- or ligand-localized orbitalar descrip-

tion for these complexes, these absorptions arise possibly

from metal-to-ligand charge transfer (MLCT) processes,

since such transitions are usually observed for related

complexes [41,52]. Molecular calculations and related

data suggest that the HOMO is principally metal-based

(strong iron d-orbital character) in these complexes

[27,50,54,92–96]. The occurrence of two transitions

could be rationalized considering the existence of two

empty p* alkynyl-based orbitals close in energy, resulting

from interaction of the degenerate p* levels of the alkynyl

fragment with those of the pyridyl fragment [50]. It is

worth noting that these MLCT transitions are shifted

gradually toward higher energies in the order paraBor-

thoBmeta, again suggesting the inﬂuence of conjugation

between the iron center and the nitrogen heteroatom in

the energy on the HOMO–LUMO gap. These complexes

are completely inactive in the NIR range (see Table 5).

Fig. 2. FTIR of palladium complexes 8a, 8a⁺and 8a^{+ +}in Nujol

mull.

¹⁷Similar transitions were observed for other related iron–alkynyls

after oxidation [26,27,92]. Alternatively, regarding their weak extinc-

tion coefﬁcients, these could also correspond to forbidden d–d ligand

ﬁeld (LF) transitions.

420

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

Fig. 3. UV–Vis spectra of compounds 3a, 3b and 3c in

Fig. 4. UV–Vis spectra of oxidized compounds 3a⁺, 3b⁺and 3c⁺in

dichloromethane at 20°C.

In the NIR range, all these complexes present a very

weak absorption around 1900 nm (5200 cm⁻¹), absent

in their neutral precursors (see Table 5 and Fig. 7).

While all our data on these complexes are indicative of

an iron-localized radical, it is therefore very unlikely

that this absorption would correspond to an IVCT

transition between iron and pyridyl. It is more likely

LF transitions speciﬁc to iron(III) radicals. Similar

absorptions were observed previously for related

chromium radicals [97].

Complexation affects markedly the UV–Vis spectra

of 5a or 5b (Table 4 and Fig. 5). Again, strong absorp-

tions are present near the solvent edge (LC transitions).

Additionally, for 5a two overlapping transitions are

present at the border of the visible range, as for com-

pounds 3a–b. The lowest energy transition is probably

the charge transfer band observed in 3a, shifted

bathochromically ca. 55 nm with its intensity enhanced.

Similar bathochromic shifts of MLCT transitions of

(L)₂CpRu–CꢀC-4-Py (L=Phosphine) upon coordina-

tion of W(CO)₅to the pyridyl ring have been reported

[52]. The assignment of the other absorptions seems

linked with the introduction of the tungsten center in

the compounds. Thus, the shoulder at 405 nm could be

a LF centered transition on tungsten (¹A₁¹E), its

high extinction coefﬁcient originating from overlap

with the iron-to-alkynyl MLCT transition, while the

shoulder at 325 nm could possibly be a tungsten-to-

pyridine p* charge transfer band [60,63,65]. This makes

sense considering that the iron alkynyl is a powerful

electron-releasing substituent which might blue-shift

this transition appearing at ca. 350 nm for (CO)₅W(Py)

in iso-octane. Similar shifts have been reported else-

where for related complexes and were explained by the

fact that electron-releasing substituents in the 4-posi-

tion on the pyridyl ring destabilize the lowest p* empty

pyridine/alkyne-based orbital [31]. The high extinction

coefﬁcient of this transition may originate from overlap

with the ligand centered transitions observed in the

far-UV. For 5b only one broad absorption is present in

the visible range, apart the intense p–p* ligand-based

far-UV transition in the solvent edge. Again its large

extinction coefﬁcient may result from the overlap with

other absorptions similar to those observed for 5a

Table 5

NIR spectra for compounds 3–10 (cm⁻¹

)

a

Compound

Solution (m in M⁻¹cm⁻¹

)

Solid ^b

c

w_max(m92)

w_max(m)

w_1/2

3a⁺

5165 (70)

5342 (70)

5364 (65)

5061 (75)

5133 (60)

5123 (75 ^d)

5123 (135)

5071 (55 ^d,e

5230

5280

5200

5060

5180

5090

5000

1720

2100

1000

1940

1640

1720

2000

1720

2000

3b⁺

3c⁺

5a⁺

5b⁺

8a⁺

8a⁺⁺

10a⁺

10a⁺⁺

)

5071 (100 ^e)

^aCH₂Cl₂, absorbance mode, 20°C (Cary 5).

^bKBr, diffuse reﬂexion mode, 20°C (Brucker IFS 28), results 950

cm⁻¹

.

^cNo signiﬁcance here, given to complete the data (9100 cm⁻¹).

^dEstimated from the total number of moles of complex present in

solution.

^eQualitative value only since mixture of isomers, see text.

Fig. 5. UV–Vis spectra of neutral and oxidized complexes 5a, 5b and

5a⁺, 5b⁺in dichloromethane at 20°C.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

421

(iron-to-alkynyl CT and tungsten-centered LF or tung-

sten-to-pyridine p* CT transitions) [60]. In conclusion,

for complexes 5, a global bathochromic shift of the

iron-based MLCT transitions observed for the corre-

sponding complex 3 seems to take place, which is less

important for the meta- (5b) than for the para-substi-

tuted complex 5a (ca. 20 nm), and appearance of new

tungsten-based transitions (MLCT and LF). Consider-

ing the LUMO as mainly ligand based, the low-lying

MLCT transitions attributed to an iron-centered pro-

cess, would be consistent with a further decrease in the

HOMO–LUMO gap in 5a–b relative to 3a–b, respec-

tively. Accordingly with published data [98], the tung-

sten center might primarily stabilize one of the p*

ligand/pyridine empty orbitals (the LUMO), and con-

comitantly destabilize slightly one of his own d-orbitals,

which would however not be the HOMO of the system.

This is also consistent with electrochemical data of the

complexes 5a–b.

Fig. 6. UV–Vis spectra of neutral complexes of palladium (8a) and

platinum (10a), as well as mono- (8a⁺) and di-oxidized (8a^{+ +}

)

complexes of palladium in dichloromethane at 20°C.

sors (see Fig. 7). Owing to their energy and extinction

coefﬁcients, these transitions greatly resemble the ones

detected for the free oxidized ligands 3a–c⁺, slightly

blue-shifted upon complexation to the tungsten center in

5a⁺and 5b⁺(Table 5). These are also likely to be LF

transitions speciﬁc to iron(III)-based radicals. No other

transition of low intensity, ascribable for instance to an

W(0)Fe(III) IVCT transition is observed, suggesting

thereby that the radical remains located on the iron atom

in the oxidized complexes 5a–b⁺[53,99].

In summary, after complexation of 3a–b with the

tungsten pentacarbonyl a similar global behavior can be

observed upon oxidation regarding the iron-based tran-

sitions, i.e. shift of the MLCT transition and appearance

of low energy (possibly LF) transitions, whereas the

tungsten-based LF transition seem only slightly affected.

This again is in accordance with an oxidation centered

on the iron nucleus and consistent with data reported

very recently, on the neutral ruthenium analogs

(L)₂CpRu–CꢀC-4-Py (L=Phosphine) [52].

The UV–Vis spectra of the palladium and platinum

neutral complexes 8a and 10a are quite similar and are

both distinct from the spectrum of the free ligands 3a

(Table 4 and Fig. 6). As always observed, strong absorp-

tions start near the solvent edge and extend in the

far-UV range, featuring many shoulders. Only one main

transition is present in the visible range. From the

spectrum of the model complex 7 and from published

data for (Py)₂PtCl₂[100], we can infer that the visible

region of the spectrum is dominated by the electronic

transitions centered on iron, since the bis-pyridyl

dichloro palladium(II) and platinum(II) fragments do

not absorb in this region. The lowest platinum-centered

transitions have been reported to be respectively a

spin-forbidden LF transition in the case of the cis-isomer

and a MLCT dp* transition for the trans-isomer,

both below 360 nm and presenting very weak extinction

coefﬁcients (resp. 21 and ca. 3400). Thus, if present,

these transitions are obscured probably by the other

ones. The most intense transition observed is

UV–Vis spectrometry of 5a–b⁺indicated a change in

the electronic structure upon oxidation (Table 4 and Fig.

5). As stated for oxidation of the free ligands 3a–b, these

changes consist mainly in a bathochromic shift of the

far-UV absorptions of 5a–b and a concomitant hyp-

sochromic shift of the broad visible absorptions. In the

visible range, some very low intensity absorptions are

present as shoulders and account for the dark green

color of the oxidized complexes. For 5a⁺, based on

extinction coefﬁcient considerations, we believe that only

the iron-centered MLCT transition of 5a has been

shifted and appears now as the shoulder at 325 nm, while

the intense LF centered transition on tungsten (¹A₁¹E)

assigned previously at the shoulder at 415 nm in neutral

5a was only shifted slightly and is observed as the peak

at 384 nm. This corroborate previous reports where this

transition was demonstrated to be only slightly sensitive

to the nature of the substituent present on the pyridyl

ring [60]. The two absorptions observed as shoulders at

500 and 698 nm could correspond roughly to absorp-

tions of comparable extinction coefﬁcients present in

3a⁺, the latter one being red-shifted ca. 90 nm (Fig. 5).

For 5b⁺, a similar interpretation of the observed ab-

sorptions can be made; the iron-centered MLCT transi-

tion of 5b would have been shifted hypsochromically and

would give the shoulder at 322 nm, while the tungsten-

based LF (¹A₁¹E) would be little affected by oxidation

and would appear as the shoulder observed at 393 nm.

Again, the three less energetic absorptions of the visible

spectrum exhibiting very low extinction coefﬁcients

would correspond to the three absorptions of lowest

extinction coefﬁcients observed for 3b⁺, slightly red

shifted ca. 10–40 nm (Fig. 4). Alternatively, one of these

(i.e. the shoulder observed at 441 nm) could also corre-

spond to a tungsten-based LF (¹A₁³E) transition [60].

In the NIR range, all the oxidized complexes present

a very weak absorption, absent in their neutral precur-

422

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

probably the charge transfer band of 3a, shifted

bathochromically by ca. 55 nm with its intensity en-

hanced, as stated for the complex 5a. In this case, the

presence of two iron alkynyl fragments in 8a also

contribute signiﬁcantly to its magnitude. Reminiscent

of the data reported for 3a–c, a shoulder might be

discerned, perhaps indicating a second MLCT transi-

tion in the same energy range. In the case of the

platinum analog 10a, a more important shift (ca. 85

nm) is observed, suggesting a stronger interaction rela-

tive to 8a, consistent with the electrochemical data. The

presence of the cis- and trans-isomers and other minor

compounds in 10a is not apparent markedly, emphasiz-

ing that mainly iron-centered transitions are observed

in the visible range which are not inﬂuenced by the

stereochemistry around the platinum center. In conclu-

sion, for the complexes 8a and 10a the visible spectrum

seems to be dominated by the iron–alkynyl MLCT

transitions. As for the tungsten complex 5a, these are

shifted bathochromically relative to 3a; however, in

contrast to 5a, new MLCT and LF palladium- or

platinum-based transitions are not observed.

The UV–Vis spectrum of the di-oxidized dark brown

8a^{+ +}or 10a^{+ +}complexes relative to their neutral

precursors shows the characteristic features already

noted in the case of 5a/5a⁺, i.e. a hypsochromic shift

of the MLCT iron-based transition and appearance of

very low intensity absorptions in the mid-visible range

at ca. 510–600 nm (Table 4 and Fig. 6). Again, the

intense absorptions above the solvent edge present nu-

merous shoulders that extend into the near UV–Vis

range. As for 8a and 10a the observation of a similar

spectrum in the case of platinum (mixture of isomers)

and palladium (single complex) further indicates that

the visible spectrum is probably dominated by the

iron-alkynyl electronic transitions. Relative to their

neutral precursors, the supposed shift of the iron-based

MLCT transition maximum is larger for the mixture of

platinum isomers (135 nm) than for the palladium one

(110 nm), suggesting that the electronic changes in the

former upon oxidation are more pronounced.

5a⁺and this would be inaccordance with its attribution

as an iron-centered LF transition, since two ligands are

present in the complexes 8a^{+ +}and 10a^{+ +}. No other

transition of low intensity, ascribable for instance to an

M(II)Fe(III) transition could be detected (M=Pd,

Pt). These results conﬁrm that more or less similar

electronic modiﬁcations take place within the ligand

upon complexation with palladium or platinum centers

reminiscent of those observed previously with tungsten.

Such changes appear therefore to be diagnostic of

coordination of the lone pair to a transition metal

center and are readily detectable by UV measurements.

Much published data on related complexes indicate

that these changes upon complexation have some gen-

eral character and may be attributed here to the stabi-

lization of one of the lowest lying p*

pyridine/acetylene-based orbital which is probably the

LUMO of the system [21,29,31–33].

In the UV–Vis range, the spectrum of the MV

complex 8a⁺generated in the cell using double

amounts of the di-oxidized 8a^{+ +}and neutral 8a com-

plexes consists of the superposition of the main peaks

observed for neutral and dicationic precursor spectra,

no extra absorption attributable to a new species being

present (Fig. 6).

In the NIR range, only a very weak absorption,

similar to the one detected for the di-oxidized species,

but of reduced extinction coefﬁcient is observed. This

makes sense considering that a halved extinction coefﬁ-

cient should be observed for the MV relative to the

di-oxidized complex and brings us to conclude that no

IVCT transition, even with a weak extinction coefﬁcient

(50–100), is present in the NIR spectrum of the MV

state. Clearly, no other transition eventually at-

tributable to an Fe(II)Fe(III) or IVCT transition is

present. Similar results were obtained from mixtures of

the platinum isomers 10a and 10a^{+ +}. Accordingly

with the IR data reported already, these MV complexes

can be classiﬁed as localized class-I MV complexes in

the classiﬁcation of Robin and Day [101].

Similarly to the tungsten complex 5a, a very weak

absorption at respectively 1951 nm (5123 cm⁻¹) and

1972 nm (5171 cm⁻¹) is present in the di-oxidized

complexes 8a^{+ +}and 10a^{+ +}and absent in the neutral

complexes 8a and 10a (Table 5 and Fig. 7). It corre-

sponds possibly to the weak transition observed for the

3a⁺radical slightly blue shifted upon complexation, as

stated previously in the case of the tungsten complex

5a⁺. Interestingly, the magnitude of the hypsochromic

shift for 5a⁺, 8a^{+ +}and 10a^{+ +}relative to 3a⁺follow

the same ordering than the shift stated for the Fe(II)/

Fe(III) redox potentials and might be related to the

strength of the bond between the pyridyl group and the

complexing metal center. The extinction coefﬁcient of

this transition is roughly doubled relative to 3a⁺or

Fig. 7. NIR spectra of the neutral (3a) and oxidized (3a⁺) para-

pyridyl ligand and of its oxidized complexes with tungsten (5a⁺),

palladium (8a^{+ +}) and platinum (10a^{+ +}) in dichloromethane at

20°C.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

423

Scheme 12.

triphenylphosphine, thereby allowing the possibility to

chemically ‘switch on and off’ the formation of 8a.

As a consequence, these trimetallic complexes can

not be envisioned as molecular wire-like compounds in

their mono-oxidized state. While this conclusion is

quite similar to the one reached by Carugo et al. with D

(Scheme 12) [30], our results point out that intra-ligand

communication was not the limiting factor for the

electronic interaction. Indeed, the electronic modiﬁca-

tions in the energy of the iron-centered HOMO are

diagnostic of a good interaction between the iron center

and the latter. As a consequence, the weak electronic

coupling ﬁnds its origin in ineffective electronic interac-

tion between the pyridine–nitrogen and the platinum-

or palladium-based molecular orbitals, as suggested by

the results of Miller et al. with rhenium [29]. Further

work is currently in progress to conﬁrm this hypothesis.

In itself, the platinum center is not an insulating linker

since recent results in the group of Vahrenkamp with

the complex E (Scheme 12) allowed observation of a

4. Conclusions

We have demonstrated that the new organoiron-

functionalized pyridyl compounds 3a–c could be iso-

lated in fair yields from the terminal iron alkynyl

complex 1 by use of a Sonogashira-derived coupling

reaction. These compounds are stable under two oxida-

tion states and can be seen as pyridines with an elec-

tron-releasing substituent. Upon chemical oxidation,

thermally stable organoiron(III) radicals 3a–c⁺are

isolated. Complexes 3a–b are good ligands for different

metal centers such as tungsten, palladium or platinum

and, depending on the coordination geometry, they

may be disposed in space using transition metal com-

plexes as stereochemical connectors. Coordination of

the lone pair of the dangling pyridyl group to a metal

results in characteristic modiﬁcations of the electronic

properties of the ligands, as for instance (i) a shift of

the oxidation potential toward more positive values; (ii)

a bathochromic shift of the iron(II)-centered MLCT

transition, indicating a decrease of the HOMO–LUMO

gap. These changes illustrate further the good electronic

communication within the ligand. From our electro-

chemical (CV) and spectroscopic (UV–Vis) data with

5a, 8a and 10a, the following order can be proposed

regarding the strength of the interaction between the

new metal center and pyridine in the neutral state:

PdBW5Pt. The electronic distribution in the oxidized

complexes is less clear, since IR and CV indicate oppo-

site inﬂuences for platinum and tungsten regarding the

Lewis structure of coordinated oxidized ligand. The

electronic communication through the central square

planar palladium or platinum dichloride core was more

particularly investigated using the corresponding MV

complex. CV and NIR were indicative of a negligible

electronic coupling between the iron centers in the

strong IVCT transition (1280 nm; m=270 M⁻¹cm⁻¹

)

in the MV complexes¹⁸[103,104]. Thus, the unpaired

electron can in principle travel through the platinum

orbitals and the poor electronic interaction in trans-

10a⁺relative to in E has to be rationalized by consider-

ing that the cyanide ligand coordinates the platinum

center more effectively for electron transfer than the

para-alkynyl pyridyl ligand¹⁹. Nevertheless, we have

shown here that two electroactive ligands as 3a could

be coordinated on the same metal center and that the

resulting complexes were kinetically stable under sev-

eral redox states. Moreover, in the case of palladium,

the neutral ligand 3a can be easily ‘plugged off’ from

the coordination sphere of the complex by addition of

¹⁸Note however that only one IVCT transition was observed for

this complex and was attributed to a Fe(II)Fe(III) e⁻-transfer,

none corresponding to a Pt(II)Fe(III) e⁻-transfer [102].

¹⁹The fact that electronic coupling occurs through six bonds in E

and 14 bonds in 8a⁺or 10a⁺will evidently also have a strong

inﬂuence on its magnitude.

424

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

[23] N. Le Narvor, L. Toupet, C. Lapinte, J. Am. Chem. Soc. 117

(1995) 7129.

[24] N. Le Narvor, C. Lapinte, Organometallics 14 (1995) 634.

[25] V. Guillaume, V. Morvan-Mahias, C. Lapinte, J. Organomet.

Chem. (1997) 203.

[26] T. Weyland, C. Lapinte, G. Frapper, M.J. Calhorda, J.-F.

Halet, L. Toupet, Organometallics 16 (1997) 2024.

[27] T. Weyland, K. Costuas, A. Mari, J.-F. Halet, C. Lapinte,

Organometallics 17 (1998) 5569.

[28] F. Paul, C. Lapinte, Coord. Chem. Rev. 178/180 (1998) 427.

[29] T.M. Miller, K.J. Ahmed, M.S. Wrighton, Inorg. Chem. 28

(1989) 2347.

complexes, showing that adiabatic electron transfer is

weak or non-existent through such inorganic connec-

tors (Class-I MV). However, these results do not imply

that through-bond intramolecular electron transfer can-

not take place competitively with intermolecular elec-

tron exchange in these molecules and current work is

being undertaken to clarify this point and to see if other

transition-metal complexes would constitute better

connectors.

[30] O. Carugo, G. De Santis, L. Fabbrizzi, M. Licchelli, A.

Monochino, P. Pallavicini, Inorg. Chem. 31 (1992) 765.

[31] J.T. Lin, S.-S. Sun, J.J. Wu, L. Lee, K.-J. Lin, Y.F. Huang,

Inorg. Chem. 34 (1995) 2323.

[32] J.T. Lin, J.J. Wu, C.-S. Li, Y.S. Wen, K.-J. Lin, Organometal-

lics 15 (1996) 5028.

[33] J.T. Lin, S.-S. Sun, J.J. Wu, Y.-C. Liaw, K.-J. Lin, J.

Organomet. Chem. 517 (1996) 217.

[34] S. Sakanishi, D.A. Bardwell, S. Couchman, J.C. Jeffery, J.A.

Mac Cleverty, M.D. Ward, J. Organomet. Chem. 528 (1997) 35.

[35] J.T. Lin, M.-F. Yang, C. Tsai, Y.S. Wen, J. Organomet. Chem.

564 (1998) 257.

Acknowledgements

We wish to thank D. Lentz for preliminary experi-

mental work, L. Balois (UMR 7513) for far-IR mea-

surements, P. Je´han and S. Sinbandhit (C.R.M.P.O.)

for LSIMS and NMR measurements, respectively. Fi-

nally, we are grateful to Standa Industries for ﬁnancial

support.

[36] C. Levanda, C. Bechgaard, D.O. Cowan, J. Org. Chem. 41

(1976) 2700.

[37] H. Fink, N.J. Long, A.J. Martin, G. Opromolla, A.J.P. White,

D.J. Williams, P. Zanello, Organometallics 16 (1997) 2646.

[38] S. Le Stang, D. Lenz, F. Paul, C. Lapinte, J. Organomet.

Chem. 572 (1998) 189.

[39] N.G. Connelly, M.P. Gamasa, J. Gimeno, C. Lapinte, E.

Lastra, J.P. Maher, N. Le Narvor, A.L. Rieger, P.H. Rieger, J.

Chem. Soc., Dalton Trans. (1993) 2775.

[40] R. Denis, T. Weyland, F. Paul, C. Lapinte, J. Organomet.

Chem. 545/546 (1997) 615.

[41] R. Denis, F. Paul, C. Lapinte, work in progress.

[42] C.H. Kline Jr., J. Turkevich, J. Chem. Phys. 12 (1944) 300.

[43] L. Corrsin, B.J. Fax, R.C. Lord, J. Chem. Phys. 21 (1953) 1170.

[44] J.K. Wilmshurst, H.J. Bernstein, Can. J. Chem. 35 (1956) 1183.

[45] A.R. Katritzky, A.M. Monro, J.A.T. Beard, D.P. Dearnaley,

N.J. Earl, J. Chem. Soc. (1958) 2182.

[46] A.R. Katritzky, J. Chem. Soc. (1958) 4162.

[47] A.R. Katritzky, A.R. Hands, R.A. Jones, J. Chem. Soc. (1958)

3165.

[48] A.R. Katritzky, A.R. Hands, J. Chem. Soc. (1958) 2202.

[49] A.R. Katritzky, J.N. Gardner, J. Chem. Soc. (1958) 2198.

[50] D.L. Lichtenberger, S.K. Renshaw, R.M. Bullock, J. Am.

Chem. Soc. 115 (1993) 3276.

[51] R. Chukwu, A.D. Hunter, B.D. Santasiero, Organometallics 10

(1991) 2141.

References

[1] J.M. Lehn, Supramolecular Chemistry-Concepts and Perspec-

tives, VCH, Weinheim, 1995.

[2] A. Harriman, R. Ziessel, Chem. Commun. (1996) 1707.

[3] A. Harriman, J.-P. Sauvage, Chem. Soc. Rev. (1996) 41.

[4] M.D. Ward, Chem. Ind. 15 (1996) 568.

[5] P.J. Stang, B. Olenyuk, Acc. Chem. Res. 30 (1997) 502.

[6] W.T.S. Huck, A. Roher, A.T. Anilkumar, R.H. Fokkens,

N.M.M. Nibbering, F.C.J.M. van Vegel, D.N. Reinhoudt, New

J. Chem. 2 (1998) 165.

[7] T. Habicher, J.-F. Nierengarten, V. Gramlich, F. Diederich,

Angew. Chem., Int. Ed. Engl. 37 (1998) 1916.

[8] M. Fujita, M. Aoyagi, F. Ibukoro, K. Ogura, K. Yamaguchi, J.

Am. Chem. Soc. 120 (1998) 611.

[9] F. Ibukuro, T. Kusukawa, M. Fujita, J. Am. Chem. Soc. 120

(1998) 8561.

[10] P.J. Stang, Chem. Eur. J. 4 (1998) 19.

[11] C. Creutz, H. Taube, J. Am. Chem. Soc. 91 (1969) 3988.

[12] C. Creutz, H. Taube, J. Am. Chem. Soc. 95 (1973) 1086.

[13] J.-P. Sauvage, J.-P. Colin, J.-C. Chambron, S. Guillerez, C.

Coudret, Chem. Rev. 94 (1994) 993.

[14] M.D. Ward, Chem. Soc. Rev. 24 (1995) 121.

[15] R. Ziessel, A. Harriman, Sci. Chim. Lett. Dep. Sci. C.N.R.S. 55

(1996) 11.

[16] V. Grosshenny, A. Harriman, M. Hissler, R. Ziessel, Platinum

Metals Rev. 40 (1996) 72.

[17] V. Grosshenny, A. Harriman, M. Hissler, R. Ziessel, Platinum

Metals Rev. 40 (1996) 26.

[52] I.-Y. Wu, J.T. Lin, J. Luo, S.-S. Sun, C.-S. Li, K.J. Lin, C.

Tsai, C.-C. Hsu, J.-L. Lin, Organometallics 16 (1997) 2038.

[53] F. Coat, M.-A. Guillevic, L. Toupet, F. Paul, C. Lapinte,

Organometallics 16 (1997) 5988.

[54] C. Roger, P. Hamon, L. Toupet, H. Rabaaˆ, J.-Y. Saillard, J.-R.

Hamon, C. Lapinte, Organometallics 10 (1991) 1045.

[55] H.H. Jaffe´, J. Chem. Phys. 20 (1952) 1554.

[56] M. Tilset, J.-R. Hamon, P. Hamon, Chem. Commun. (1998)

765.

[18] J.E. Sutton, H. Taube, Inorg. Chem. 20 (1981) 3125.

[19] J.A. Thomas, C.J. Jones, T.A. Hamor, J.A. Mac Cleverty, F.

Mabbs, D. Collison, C. Harding, Mol. Cryst. Liq. Cryst. 234

(1993) 103.

[20] S.L.W. Mac Whinnie, J.A. Thomas, T.A. Hamor, C.J. Jones,

J.A. Mac Cleverty, D. Collison, F.E. Mabbs, C.J. Harding, L.J.

Yellowlees, M.C. Hutchings, Inorg. Chem. 35 (1996) 760.

[21] M.M. Bhadhade, A. Das, J.C. Jeffery, J.A. Mac Cleverty,

J.A.N. Badiola, M.D. Ward, J. Chem. Soc., Dalton Trans.

(1995) 2769.

[57] N. Le Narvor, C. Lapinte, J. Chem. Soc., Chem. Commun.

(1993) 357.

[58] T.B. Grindley, K.F. Johnson, A.R. Katritzky, H.J. Keogh,

R.D. Topsom, J. Chem. Soc., Dalton Trans. (1974) 273.

[59] T.B. Grindley, K.F. Johnson, A.R. Katritzky, H.J. Keogh, C.

Thirkettle, R.D. Topsom, J. Chem. Soc., Dalton Trans. (1974)

282.

[22] A.J. Amoroso, A.M.W.C. Thompson, J.P. Maher, J.A. Mac

Cleverty, M.D. Ward, Inorg. Chem. 34 (1995) 4828.

S. Le Stang et al. / Inorganica Chimica Acta 291 (1999) 403–425

425

[60] M.S. Wrighton, H.B. Abrahamson, D.L. Morse, J. Am. Chem.

Soc. 98 (1976) 4105.

[84] J. Fornies and E. Lalinde, J. Chem. Soc., Dalton Trans. (1996)

2587.

[61] S.W. Kirtley, Comprehensive Organometallic Chemistry, vol. 3,

5th Edition, Pergamon, New York, 1982, p. 1271.

[62] R.M. Kolodziej, A.J. Lees, Organometallics 5 (1986) 450.

[63] M.M. Zulu, A.J. Lees, Inorg. Chem. 27 (1988) 1139.

[64] P.L. Gaus, J.M. Boncella, K.S. Rosengren, M.O. Funk, Inorg.

Chem. 21 (1982) 2174.

[85] I. Ara, J.R. Berenguer, E. Eguizabal, J. Fornies, E. Lalinde, A.

Martin, F. Martinez, Organometallics 17 (1998) 4578.

[86] S.T. Chow, R.B. Martin, Inorg. Nucl. Chem. Lett. 10 (1974)

1131.

[87] D.M. Adams, J. Chatt, J. Gerratt, A.D. Westland, J. Chem.

Soc. (1964) 734.

[65] K.H. Pannel, M.G. Gonzalez, H. Leano, R. Iglesias, Inorg.

Chem. 17 (1978) 1093.

[66] F.R. Hartley, Organomet. Chem. Rev. A 6 (1970) 119.

[67] J.R. Doyle, P.E. Slade, H.B. Jonassen, in: E.G. Rochow (Ed.),

Inorganic Synthesis, vol. 6, McGraw Hill, New York, 1960, p.

216.

[68] P. Braunstein, R. Bender, J. Jud, Inorg. Synth. 26 (1989) 341.

[69] C.M. Drain, F. Niﬁatis, A. Vasenko, J.D. Batteas, Angew.

Chem., Int. Ed. Engl. 37 (1998) 2344.

[70] C.M. Drain, J.-M. Lehn, J. Chem. Soc., Chem. Commun.

(1994) 2313.

[88] A.J. Bard, L.R. Faulkner, R. Rosset, D. Bauer, Electrochem-

istry. Principle, Methods and Applications, Masson, Paris:

French translation, 1983, p. 249.

[89] P.-C. Kong, F.D. Rochon, Can. J. Chem. 56 (1978) 441.

[90] S. Le Stang, F. Paul, C. Lapinte, work in progress.

[91] P.M. Mailtis, P. Espinet, M.J.H. Russell, Comprehensive

Organometallic Chemistry, vol. 6, 5th Edition, Pergamon, New

York, 1982, p. 235.

[92] T. Weyland, Thesis (1997), University of Rennes I.

[93] J.E. Mac Grady, T. Lovell, R. Stranger, M.G. Humphrey,

Organometallics 16 (1997) 4004.

[71] F.P. Fanizzi, G.J. Sunley, J.A. Wheeler, H. Adams, N.A.

Bailey, P.M. Maitlis, Organometallics 9 (1990) 131.

[72] J.R. Durig, B.R. Mitchell, D.W. Sink, J.N. Willis Jr, Spec-

trochim. Acta Part A 23 (1967) 1121.

[94] F. de Angelis, N. Re, M. Rosi, A. Sgamellotti, C. Floriani, J.

Chem. Soc., Dalton Trans. (1997) 3841.

[95] J. Manna, K.D. John, M.D. Hopkins, Adv. Organomet. Chem.

38 (1995) 80.

[73] C.J. Adams, S.L. James, P.R. Raithby, Chem. Commun. (1997)

2155.

[74] H. Yuan, L. Thomas, L.K. Woo, Inorg. Chem. 35 (1996) 2808.

[75] P.J. Stang, D.H. Cao, S. Saito, A.M. Arif, J. Am. Chem. Soc.

117 (1995) 6773.

[96] N.M. Kostic, R.F. Fenske, Organometallics 1 (1982) 974.

[97] C.G. Atwood, W.E. Geiger, J. Am. Chem. Soc. 116 (1994)

10849.

[98] D.R. Kanis, P.G. Lacroix, M.A. Ratner, T.J. Marks, J. Am.

Chem. Soc. 116 (1994) 1008.

[76] S. Troﬁmenko, Inorg. Chem. 6 (1973) 1215.

[77] G.B. Kauffman, in: J. Kleinberg (Ed.), Inorganic Synthesis, vol.

7, McGraw Hill, New York, 1963, p. 249.

[78] A.R. Siedle, L.H. Pignolet, Inorg. Chem. 21 (1982) 135.

[79] M. Pfeffer, P. Braunstein, J. Dehand, Spectrochim. Acta Part A

30 (1973) 331.

[99] F. Paul, W. Meyer, J.A. Gladysz, C. Lapinte, work in progress.

[100] M. Martin, M.-B. Krogh-Jespersen, M. Hsu, J. Tewksbury, M.

Laurent, K. Viswanath, H. Patterson, Inorg. Chem. 22 (1983)

647.

[101] M.B. Robin, P. Day, Adv. Inorg. Chem. Radiochem. (1967)

247.

[80] G.E. Coates, C. Parkin, J. Chem. Soc. (1963) 421.

[81] J.R. Durig, R. Layton, D.W. Sink, B.R. Mitchell, Spectrochim.

Acta 21 (1967) 1367.

[82] M. Pfeffer, P. Braunstein, J. Dehand, Spectrochim. Acta Part A

30 (1973) 341.

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

120 (1998) 1812.

[103] G.N. Richardson, H. Varenkamp, Remote intervalence transfer

in polynuclear cyanide bridged complexes, Book of Abstracts,

XXXIIIth ICCC Conferences, 1998, p. 400.

[83] H.H. Eysel, E. Guggolz, M. Kopp, M.L. Ziegler, Z. anorg. allg.

Chem. 199 (1983) 31.

[104] H. Vahrenkamp, A. Giess, G.N. Richardson, J. Chem. Soc.,

Dalton Trans. (1997) 3643.

.

Article Doi

Redox active organoiron alkynyl centers with a coordination site: An approach toward molecular plugs. Complexation reactions with (THF)W(CO)5 and (PhCN)2MCl2 (M=Pd, Pt)

DOI: 10.1016/s0020-1693(99)00135-8

Source and publish data:

Authors:

Article abstract of DOI:10.1016/s0020-1693(99)00135-8

Full text of DOI:10.1016/s0020-1693(99)00135-8

Products guided by the article

R&D Labs maybe for 221244-13-9

Relevant to this article

Hot Product