Organometallic manganese complexes as scaffolds

0 downloads 0 Views 332KB Size Report
professor of Inorganic Chemistry. In 1991 ... various fields of organometallic chemistry. ..... 1.088 V, Ep2 = 0.993 V (ΔEp = 0.095 V) when all scan rates applied.
PERSPECTIVE

www.rsc.org/dalton | Dalton Transactions

Organometallic manganese complexes as scaffolds for potential molecular wires† Koushik Venkatesan, Olivier Blacque and Heinz Berke* Received 26th October 2006, Accepted 15th January 2007 First published as an Advance Article on the web 29th January 2007 DOI: 10.1039/b615578b This article reviews recent work in the area of organomanganese chemistry designing organometallic based molecular wires for potential applications in molecular electronics utilising the bottom-up approach. The field of molecular electronics has recently received much attention in the pursuit of continued miniaturization of electronics. Molecular wires that can allow a through-bridge exchange of an electron/electron hole between its remote ends/terminal groups are the basic motifs of single electron devices. Our recent work in this field has been the design and development of transition-metal complexes with a special emphasis on the half sandwich dinuclear manganese complexes and the bis dmpe dinuclear MnII /MnII . In this review, we would like to highlight the importance of the nature of the transition metal and their significant effect on the redox process, which is of paramount importance for the design of systems that could be ultimately wired into circuits for various applications.

Introduction As defined by J. M. Lehn, a molecular wire is a “one dimensional molecule allowing a through-bridge exchange of an electron/hole between its remote ends/terminal groups, themselves able to exchange electrons with the outside world”.1 An organometallic molecular wire would thus consist of a “conducting” organic bridge with redox-active metal endgroups. Depending on the electronic delocalization over both metal centers, a classification of bridged redox couples has been proposed by Robin and Day.2 Our mainstay has been the design of Robin–Day classIII type compounds for which the electron is fully delocalized along the spacer, making the metal centers indistinguishable with the available spectroscopy. The compounds corresponding to class I or II show high barriers for electron transfer and for instance with the appropriate circuit fabrication might function as resistors, diodes, single electron transistors, and photonic switches. p-Conjugated conducting polymers and oligomers containing electrochemically active transition-metal complexes are of great interest as materials for molecular devices. In particular, linear complexes with extended p systems interacting with metal complexes, such as Lm Mn–(C≡C)n –MnLm , are potential candidates for the preparation of functional single-molecule devices.3–14 The metal complexes were assumed to impart some functionality, for example, by acting as charge-trapping sites in single-electron transistors or electron hopping and photoactive sites. Research into the synthesis and electronic properties of metal alkynyl complexes and polymers continues to develop as a key area of organometallic chemistry. Owing to its high stability, ease of functionalisation, and well-defined electrochemistry, transition metals with auxiliary alkylphosphines have been widely used as redox-active centers. These are linked together with a variety of

Anorganisch-Chemisches Institut der Universit¨at Z¨urich, Winterthurerstrasse 190, CH-8057, Z¨urich, Switzerland. E-mail: [email protected] † Dedicated to Prof. Gerhard Erker on the occasion of his 60th birthday.

This journal is © The Royal Society of Chemistry 2007

structural units such as saturated and unsaturated carbon bridges containing aromatic or heteroaromatic spacer groups, delocalised fused rings, and also polymeric and dendritic back bones. Their rigid rod architectures and conjugated backbones make them useful materials in the field of linear and non-linear optics, liquid crystals, and photovoltaic cells. The molecular conduction in these kind of molecules is believed to occur by tunneling and should be proportional to the negative potential of the square root of the energy gap and the chain length. Both parameters can be varied; however, the energy gap is particularly more sensitive to the degree of p delocalization. V ab , the electronic coupling factor for the metal centers, reflects all this.15 This condition for such molecular electronic materials, particularly the molecular wires, was anticipated to be reasonably satisfied by molecular units with electron-rich phosphine substituted manganese endgroups possessing low energy work functions and a Cn -spacer.16–22 Gladysz and co-workers5,6,23–26 have synthesized metal-terminated carbon chains with up to ten alkynyl units (Re(C≡C)n Re, n ≤ 10) for which the Re(C≡C)2 Re species showed two single electron oxidations by cyclic voltammetric studies. For the longer chains Re(C≡C)3 Re and Re(C≡C)4 Re, the oxidation becomes increasingly irreversible. The presence of only a single, presumably two-electron oxidation wave in Re(C≡C)10 Re indicates that once the chain reaches this length, the metals are effectively isolated from one another and do not communicate by the conjugated bridge. Since shorter chains between the two metal centers showed effective communication,5 we focused our efforts to build oligomers and polymers using the building units as Mn(C≡C)Mn and Mn(C≡C)2 Mn. We believe that the buildup of dinuclear complexes possessing labile terminal ends that are capable of extending into oligomers and polymers in a controlled fashion might provide the solution for making longer wires, while keeping intact the high communication between the terminal ends. The concept of using redox-active metal endgroups with low energy work functions is based on a very simple assumption. ¨ Based on Extended Huckel considerations, the hypothetical H2 2− molecule can be derived from the combination of the 1s orbitals Dalton Trans., 2007, 1091–1100 | 1091

Koushik Venkatesan was born in Chennai, India, in 1977. He received his MSc in Chemistry from Sri Sathya Sai Institute of Higher Learning, India in 1999. He graduated in 2003 with a PhD from the University of Zurich, Switzerland under the supervision of Professor Heinz Berke working on the synthesis and characterization of manganese complexes as scaffolds for potential single electron devices. In 2004, he joined Professor Timothy Swager’s group at the Massachusetts Institute of Technology (USA) as a Swiss post-doctoral fellow and is currently developing organometallic luminescent materials for sensors and liquid crystal applications. Olivier Blacque was born in Dijon, France, in 1971. He graduated from the University of Dijon and obtained his PhD under the supervision of Professor Marek Kubicki working on the structures of biscyclopentadienyl transition metal complexes using X-ray crystallography and DFT calculations. In 2000 he joined the group of Professor Heinz Berke at the University of Z¨urich as a post-doctoral fellow and in 2006 he was promoted to a permanent staff member working as an X-ray crystallographer and computational chemist. H. Berke received his Diploma in Chemistry at the University of Erlangen, Germany, in 1971 and his PhD at the University of T¨ubingen, Germany, in 1974. From 1974–1988 he has been at the University of Konstanz, Germany with an intermediate stay in the Laboratory of R. Hoffmann, Cornell University, Ithaca, USA, in 1977. In 1981 he finished his Habilitation and in 1983 he was awarded the Heisenberg fellowship from the “Deutsche Forschungsgemeinschaft” and the Dozentenpreis of the Fonds der Chemischen Industrie, Germany. In 1987 he was promoted to a C2 Professor at the University of Konstanz before he joined the University of Z¨urich, Switzerland, in 1988 as a full professor of Inorganic Chemistry. In 1991 he became director of this institute and stayed at this position until now. H. Berke was a member of the Editorial Board of Dalton Transactions (2004–2006) and is still a member of the Board of Mendeleev Communications and is presently president of the Division of Chemical Research of the Swiss Chemical Society. H. Berke’s fundamental research activities cover various fields of organometallic chemistry. Major efforts are devoted to the area of transition metal hydrides, which is related to homogeneous catalysis, in particular, homogeneous hydrogenation and hydrosilylation. Metal carbon oriented activities concern several catalyses of C–C coupling reactions mediated by transition-metal complexes and in addition metallacumulenes, where carbon chained units are sought to space transition metal centers for potential use as single-electron devices. Another research field deals with the archaeometry of ancient, man-made blue and purple pigments.

Koushik Venkatesan

Olivier Blacque

of the two hydrogen atoms producing two molecular orbitals, of which the antibonding orbital exhibiting higher absolute overlap population (P2 = −6.0 for r*1s and P1 = 0.88 for r1s ) because of larger atomic orbital coefficients (c2 = 1.44 for r*1s and c1 = 0.54 for r1s ). Based on this general conclusion, we state that orbitals with low energy work functions (those at higher energies) tend to delocalize stronger through stronger interactions (although antibonding). This even holds for extended orbital systems with electron occupancies of orbitals of higher energies. For instance, in the case of the linear H6 molecule, the HOMO can be described by an orbital with two nodal planes, while the LUMO exhibits more antibonding character with three nodal planes occupied by two electrons for which stronger H–H interactions are expected. In p systems, the same electronic principle applies. A stronger delocalization of the p MO’s of the Cn spacer can thus exist in metal fragment end-capped complexes of the type Lm M–Cn –MLm by aid of additional d metal · · · d metal interactions, where the 1092 | Dalton Trans., 2007, 1091–1100

Heinz Berke

metal d orbitals are occupied by two electrons. In particular, we can suggest that energetically high lying occupied orbitals of any metal fragment can contribute to a shift of the work functions of bridged complexes to lower energies. This is demonstrated in Scheme 1 for a p system of a C4 bridge end-capped by (occupied) metal d orbitals to allow close interactions not only with the HOMO but also with the LUMO of the C4 wire leading to two additional molecular orbitals with substantially more antibonding character and at substantially higher energies with a substantially higher degree of delocalization. As we will apply them in the later context, the various fragments have strongly electron donating properties like the “anonymous” metal in Scheme 1. As a consequence, we could expect the dinuclear manganese complexes equipped with half sandwich and bis-dmpe metal fragments having electronic properties of a class III type. We would like to also bring forth the interplay of the metal and the ancillary ligands and it’s pivotal role to enhance the class III or This journal is © The Royal Society of Chemistry 2007

oxidized mixed-valent form [1]+ was obtained in combination with the trigonally coordinated high spin anion [(g2 -MeC5 H4 )3 Mn]− 2. Treatment of the brown paramagnetic reaction mixture [1]+ [2] with an excess of KPF6 or NaBPh4 , respectively, gave the desired salts [1]+ [A] (A = PF6 − , BPh4 − ) (Scheme 2). Complexes 1, [1]+ [2], [1]+ [PF6 ]− , [1]+ [BPh4 ]− and [1]2+ [PF6 ]2 − have been fully characterized by X-ray diffraction studies.28,44 They exhibit structural variations that are directly related to the oxidation state of the manganese atoms. A similar behaviour was observed in complexes with longer carbon chains between the metal centers, which is discussed in the latter part of the review. According to the resonance forms of eqn (1) the Mn–C bonds shorten with increasing degree of oxidation, while the internal C–C bond elongates. Mn=C=C=Mn ↔ M≡C–C≡Mn Scheme 1

class I behaviour in the systems developed. This indeed has been found21,27,28 and will be described in the following review.

[Mn]–C2 –[Mn] rigid-rod species In our search for such molecular electronic materials, redox-active metal centers combined with C2 -bridges were expected to exhibit very strong through-bridge interactions stronger than those with C4 -bridges. However, such species have been rarely reported in the literature since the transition metal centers bridged by short alkyne chains are unstable.16,18,29–43 Nevertheless, we could synthesize such dinuclear complexes bridged by an ethynyl unit. Halfsandwich complexes of the type [(MeC5 H4 )(dmpe)Mn–C≡C– Mn(dmpe)(MeC5 H4 )]n+ (n = 0, 1, 2) have been prepared starting from a substituted manganocene [(MeC5 H4 )2 Mn],34,35 dmpe (or the dmpe adducts [(MeC5 H4 )2 Mn(dmpe)]) and Me3 Sn–C≡C– SnMe3 . The synthetic potential of manganocenes, in particular with regard to Cp or MeC5 H4 replacement, has not been exploited appropriately.16 Manganocenes turned out to be key starting materials to allow access to quite a variety of complexes. The reaction with Me3 Sn–C≡C–SnMe3, however, did not yield the expected [(MeC5 H4 )(dmpe)MnC2 Mn(dmpe)(C5 H4 Me)] 1, instead its

(1)

This indicates that the cumulenic resonance form is transformed into the bis-carbyne type with a C–C single bond (Table 1). X-Ray diffraction studies prove to be vital in these complexes, since these studies reveal structural features that further confirm the oxidation state of these metal centers, which is quite evident from the bond lengths. The cyclic voltammetric studies offer a quantification of the interaction between the two metal centers. We should stress at this point that the studies are performed in solution. However, we believe if single molecule studies were to be carried out between two electrodes, we might have new insights and further additional perspectives on these molecules, which will be a focus for future studies. The cyclic voltammogram starting from ˚ ) of the Mn–C2 –Mn unit in [1], [1]+ [PF6 ]− , Table 1 Structural features (A [1]+ [BPh4 ]− , [1]+ [2] and [1]2+ [PF6 ]2 −

1 [1]+ [PF6 ]− [1]+ [BPh4 ]− [1]+ [2] [1]2+ [PF6 ]2 −

Mn–C

C–C

C–Mn

1.872(2) 1.792(3) 1.8003(18) 1.817(7) 1.774(6) 1.780(6) 1.733(2)

1.271(4) 1.310(6) 1.291(4) 1.291(8) 1.312(8) 1.322(12) 1.325(5)

1.872(2) 1.792(3) 1.8003(18) 1.760(6) 1.776(6) 1.780(6) 1.733(2)

Scheme 2

This journal is © The Royal Society of Chemistry 2007

Dalton Trans., 2007, 1091–1100 | 1093

[1]2+ [PF6 ]2 − displays three fully reversible waves, one at E 1/2 = −0.847 V corresponding to the MnIII –C2 –MnIII [1]2+ [PF6 ]2 − /MnII – C2 –MnIII [1]2+ [PF6 ]2 − redox couple (DE p = 0.060 V and ipa /ipc ≈ 1 for scan rates of 0.100–0.700 V s−1 ), the second at E 1/2 = −1.835 V due to the MnIII –C2 –MnII [1]+ [PF6 ] − /MnII –C2 –MnII 1 couple and the third one at E 1/2 = −2.824 V due to a MnIII –C2 –MnII 1/MnII – C2 –MnI electron transfer (Fig. 1).

Fig. 1 Cyclic voltammogram of [1]2+ [PF6 ]2 − (10−3 M in CH3 CN n-Bu4 PF6 /THF; vs Fc/Fc+ ; Gold electrode 100 mV s−1 ; DE 1 = 0.988 V and DE 2 = 0.988 V).

This species [(MeC5 H4 )(dmpe)Mn–C2 –Mn(dmpe)(MeC5 H4 )]− is new and could not be synthesized on a preparative scale. However, this redox couple may have a too negative potential to reach the anionic state by chemical means. The difference of values between the first two waves (DE 1/2 = 0.988 V) establishes a comproportionation constant of 8.6 × 1016 (K c = exp(FDE 1/2 /RT)).45 This comproportionation constant stresses the fact that the manganese centers have a very strong electronic coupling. We believe that the K c obtained for these dinuclear complexes is by far the largest observed for such kinds of complexes. Incidentally, all potential differences between the three redox processes have the same large value, which emphasizes the existence of strong interactions between the manganese centers transmitted through the orbitals of the C2 bridge. The UV-Vis-NIR spectra of the series of complexes 1, [1]+ [PF6 ]− and [1]2+ [PF6 ]2 − showed well-structured spectra with an intense band along with a shoulder for all these complexes and one additional band of low intensity at shorter wavelength for 1 and [1]+ [PF6 ]− . In comparison with the spectra of 1 and [1]2+ [PF6 ]2 − , [1]+ [PF6 ]− did not reveal an extra broad absorption at the long wavelength end, which could have been assigned to a mixedvalence band6,23,26,45–53 as expected for class II systems according to the Robin–Day classification. Furthermore, it is quite remarkable to see how strongly the systems are affected by charge, since the two major transitions, which are expected to arise from the same types of orbitals,2 shift to shorter wavelengths on going from [1]+ [PF6 ]− to [1]2+ [PF6 ]2 − . It is presumably due to the filled orbitals in the HOMO region of [1]+ [PF6 ]− , which are particularly sensitive to positive charge and are therefore significantly lowered in energy. [Mn]≡C–CR=CR–C≡[Mn] species The preparation of divinylidene-bridged complexes of the general formula [Ln M=C=CR–CR=C=MLn ]m+ (M = Fe, Ru, Mo; R = H, Me, n-Bu, Ph, SiMe3 ; m = 0, 1) were reported earlier.54–61 All 1094 | Dalton Trans., 2007, 1091–1100

of them were obtained via dimerization of odd-electron complexes generated by redox transformations. However, in situ generated or isolable radicals [(C5 R5 )M(dppe)(C=C(H)R )]+ (M = Fe,62 Ru;63 R = H, Me; R = t-Bu, Ph) are stable, demonstrating that the course of such radical dimerizations depends on the bulkiness of the organic substituents, the metal center, and the ancillary ligands. We reported earlier17 such a dimeric compound [(MeC5 H4 )(dmpe)Mn(≡C–C(Ph)=C(Ph)–C≡)Mn(dmpe)(MeC5 H4 )] 4 was obtained from the neutral mononuclear compound [(MeC5 H4 )(dmpe)Mn(–C≡CPh)] 3 by reaction with n-Bu3 SnH as H• source. 4 was obtained only in moderate yields due to the competition with the hydrogen abstraction process. The corresponding dicationic dinuclear compound [4]2+ has also been obtained starting from the cationic species [(MeC5 H4 )(dmpe)Mn(–C≡CPh)]+ [3]+ by a slow dimerization process. It can be separated from the other decomposition products by crystallization in about 20% yield. [4]2+ can also be prepared in quantitative yield by oxidation of 4 with 2 equiv. of the ferrocenium salt (Scheme 3). Re-reduction can be accomplished by reaction of [4]2+ with methylcobaltocene. Both compounds 4 and [4]2+ are thermally stable, and complex [4]2+ is even air-stable. The structure of both the neutral complex and the dicationic complex was unequivocally established by X-ray diffraction studies.17

Scheme 3

Recently, we were able to synthesize a very similar dinuclear species [(MeC5 H4 )(depe)Mn(≡C–C(H)=C(H)–C≡)Mn(depe)(MeC5 H4 )] 5 which was obtained by the reaction between (MeC5 H4 )2 Mn, depe and an excess of Bu3 Sn–C≡C–H along with the mononuclear vinylidene complex (Scheme 4).67 Only a small amount of the title compound crystallized from the above mixture due to the high solubility of 5. Nevertheless, 5 could be separated efficiently by oxidation of the purified mixture with an excess of [Cp2 Fe][PF6 ]. Due to the difference in solubility between the two dicationic complexes, the dinuclear dicationic bis(carbyne) complex (MnII –MnII ) 5 compound could be isolated in a 50% yield. The structures of both the neutral complex (Fig. 2) and the dicationic complex were also unequivocally established by X-ray diffraction studies.44 It is worth to note that the 1 H NMR showed diamagnetic signals instead of paramagnetic signals which are quite unusual for MnII complexes. This behaviour could be attributed to the strong antiferromagnetic coupling between the two MnII centers. This journal is © The Royal Society of Chemistry 2007

Scheme 4

Fig. 2 Ortep representation of the X-ray structure of 5. Thermal ellipsoids are shown with a 30% probability level.

As the redox couple of complexes 4 and [4]2+ and 5 and [5]2+ can be reversibly converted by chemical reagents, we have studied the redox properties of 4 and 5 in more detail with cyclic voltammetry. Starting from compound 4 two fully reversible redox couples have been found at E 1/2 = −0.656 and −0.445 V vs Fc/Fc+ (DE p = 0.060–0.065 V and ipa /ipc ≈ 1 for scan rates of 0.050–0.500 V s−1 ). The separation of these waves of 0.211 V is consistent with two successive oxidation steps, yielding the monocation [4]+ and the dication [4]2+ , respectively. Another pair of waves is observed at a higher potential about 1.5 V, which can be assigned to further oxidation steps producing Mn(IV) derivatives. The first wave (E 1/2 = 0.945 V; DE p = 0.065 V) is reversible at high scan rates, but at values lower than 0.400 V s−1 the intensity of the cathodic peak grows to give ipa /ipc < 1. A similar situation exists for the wave at E p1 = 1.088 V, E p2 = 0.993 V (DE p = 0.095 V) when all scan rates applied. These cathodic stripping peaks show that the Mn(IV) derivatives precipitate on the working electrode. The formal separation of

these waves is less than in the previous case (DE = 0.096 V). From the separation between two redox steps, the comproportionation constant17 has been established at K c1 = 8 × 103 . A similar cyclic voltammetric studies have been performed on the neutral compound 5. The CV of the 5/[5]+ /[5]2+ system displays two reversible waves at E 1/2 = −0.82 V and E 1/2 = −1.38 V corresponding to the Mn(II)–Mn(II) ([5]2+ )/Mn(II)–Mn(I) ([5]+ ) and Mn(II)–Mn(I) ([5]+ )/Mn(I)–Mn(I) (5) redox couples, respectively. The difference of these two values of DE 1/2 = 0.576 V establishes a comproportionation constant K c of 6.6 × 109 . Surprisingly, this comproportionation constant is larger by about 106 than the one obtained for the very similar phenylated compound 4. This huge difference observed between these quite similar compounds can be attributed from the solid state structures of 4 and 5. In contrast to the corresponding dicationic species [4]2+ , the neutral species 5 and the centrosymmetric dimer [Mo2 (C7 H7 )2 (dppe)2 (l-C4 Ph2 )](PF6 )2 ,55 the neutral binuclear complex [(MeC5 H4 )(dmpe)Mn(≡C–C(Ph)=C(Ph)–C≡)Mn(dmpe)(MeC5 H4 )] 4 possess no center of symmetry similar to the other structurally related dimers [(C5 H5 )2 Fe2 (dppe)2 (l-C4 Me2 )](BF4 )2 54 and [(g5 C5 Me5 )(CO)2 Mn=C=CPh]2 .66 The crystallographic inversion symmetry in [4]2+ and 5 gives rise to planar trans conformations at the central C–C bond characterized by C1–C2–C2a–C1a torsional angles of 180◦ , like in the iron and molybdenum dimers. Even when no center of symmetry is present in [5]2+ the same trans geometry is observed with C1–C2–C3–C4 of 179.7◦ . Surprisingly, the geometry of 4 is quite different from [4]2+ , 5 and [5]2+ . Both metal fragments (MeC5 H4 )(dmpe)Mn(≡C–CPh) are oriented almost perpendicularly to each other with the corresponding C1–C2–C32–C31 torsional angle of 86.4◦ . This kind of conformation has also been found in the neutral complex [(g5 C5 Me5 )(CO)2 Mn=C=CPh]2 66 with a torsional angle of 76.4◦ . The orientation of the vinylidene fragments in 4 is almost horizontal: the angle between the plane Cp1–Mn1–C1 and the vinylidene plane C1–C2–C3 amounts to 74◦ , and that between the planes Cp2–Mn31–C31 and C31–C32–C33 amounts to 79◦ . In the iron dimer the corresponding angles are 90 and 117◦ ,54 while in the molybdenum dimer the ligand orientation approximates that of a vertical one with a 19◦ tilt55 and in [4]2+ the corresponding angle exhibits an intermediate value of 46◦ . Other structural parameters of the Mn–C4 –Mn bis(vinylidene) moiety such as M=C and C=C distances and M=C=C angles fall in a typical range (Table 2).

˚ ) of the Mn–C4 R2 –Mn unit in 4, [4]2+ [BF4 ]2 − , 5, [5]2+ [PF6 ]2 − and other similar compounds from the literature Table 2 Structural features (A

4 [4]2+ 5 [5]2+ [(g5 -C5 Me5 )(CO)2 Mn=C=CPh]2 [{Cp(dmpe)Fe=C=CPh}2 ]2+ 2BF4 − [{(g7 -C7 H7 )(dppe)Mo=C=CPh}2 ]2+ 2PF6 − [(POMe)3 (CO)2 Fe=C=C(C(H)O2 C3 H6 )]2 [{(g5 -C5 H4 Me)2 (CO)Nb=C=CMe}2 ]2+ 2BPh4 −

This journal is © The Royal Society of Chemistry 2007

Mn–Ca

Ca –Cb

Cb –Cb 

Ref.

1.742(5) 1.737(4) 1.684(3) 1.759(2) 1.723(13) 1.720(14) 1.760(5) 1.762(5)

1.351(6) 1.353(6) 1.421(5) 1.332(3) 1.332(16) 1.322(17) 1.342(6) 1.332(6) 1.33(1) 1.353(7) 1.34(1)–1.35(2) 1.31(1) 1.320(3)

1.529(6)

17

1.400(7) 1.477(5) 1.270(12)

17 — 44

1.510(6)

66

1.50(1) 1.472(12) 1.49(1) 1.54(1) 1.468(3)

54 55 64 65

Dalton Trans., 2007, 1091–1100 | 1095

Scheme 5

The perpendicular orientation of the metal fragments (MeC5 H4 )(dmpe)Mn(≡C–CPh) have been confirmed by DFT calculations in the case of the neutral and the cationic model complexes [(MeC5 H5 )(dHpe)Mn≡C–C(Ph)=C(Ph)–C≡Mn(dHpe)(MeC5 H5 )]n+ . Indeed, in these two cases the perpendicular conformations are favored over the linear arrangement by 22.2 and 0.4 kJ mol−1 , respectively. In the dicationic species, the planar conformation is preferred (14.2 kcal mol−1 ) as observed for [(MeC5 H5 )(dHpe)Mn≡C–C(H)=C(H)–C≡Mn(dHpe)(MeC5 H5 )]n+ with n = 0, 1, and 2. These conformational observations have been found at the molecular orbital level. Due to the shape of the LUMO in [(MeC5 H5 )(dHpe)Mn≡C–C(R)=C(R)– C≡Mn(dHpe)(MeC5 H5 )] with a central (R)C–C(R) r antibonding interaction and Mn–C p antibonding character, the rotation of one metallic fragment along the central (R)C–C(R) bond does not change drastically the orbital picture (Scheme 5). On the other hand, the HOMO is more affected by the rotation, and the p antibonding interaction observed along the central (H)C–C(H) bond turns into a nonbonding character along (Ph)C–C(Ph) and consequently, the communication between the metal centers may be reduced in the case of the phenylated compound.

Complexes of the type [(MeC5 H4 )(dmpe)Mn–C≡C–X–C≡C– Mn(dmpe)(C5 H4 Me)] (X = 1,3-C6 H4 ; 1,4-C6 H4 ; 4,4-(C6 H4 )2 ) were obtained by treatment of (MeC5 H4 )2 Mn(dmpe) with 0.5 equiv. of the corresponding acetylides 1,3-C6 H4 (C≡C–SnMe3 )2 , 1,4C6 H4 (C≡C–SnMe3 )2 and 4,4-(C6 H4 )2 (C≡C–SnMe3 )2 in THF for 72 h to afford the corresponding neutral MnII –MnII dinuclear complexes in very good yields (Scheme 6).68

[Mn]–X–[Mn] rigid-rod species It is also worth mentioning that as the chain length increases, however, the synthesis becomes more and more difficult and the stability of the compounds decreases, especially with electron rich termini. For this reason, we have tried to synthesize new compounds incorporating cyclic groups (i) to keep the rigid geometry of the carbon chain, (ii) to facilitate the synthetic process, (iii) to increase the chemical stability of the compounds, and (iv) to maintain a high level of electronic coupling. Among the various p-conjugated systems developed during the past decade, phenyl, biphenyl and thiophene derivatives seem to present a blend of these properties. Complexes of this type have been widely studied in conjunction with potential molecular wire applications. 1096 | Dalton Trans., 2007, 1091–1100

Scheme 6

These complexes could be oxidized to the corresponding dicationic species [(MeC5 H4 )(dmpe)Mn–C≡C–X–C≡C–Mn(dmpe)(C5 H4 Me)]2+ [PF6 ]2 with 2 equiv. of [(C5 H5 )2 Fe][PF6 ]. The cyclic voltammetric studies of these dicationic complexes in CH3 CN showed reversible redox processes except for the [(MeC5 H4 )(dmpe)Mn–C≡C–X–C≡C–Mn(dmpe)(C5 H4 Me)] (X = 1,3-C6 H4 ). This journal is © The Royal Society of Chemistry 2007

The comproportionation constant K c for the other two derivatives [(MeC5 H4)(dmpe)Mn–C≡C–X–C≡C–Mn(dmpe)(C5 H4 Me)] (X = 1,4-C6 H4 ; 4,4-(C6 H4 )2 ) were K c = 1.75 × 104 and 1.35 × 109 , respectively. Our attempts to synthesize the dinuclear complex bridged by the butadiyne spacer did not allow isolation of the corresponding [(MeC5 H4 )(dmpe)Mn–C≡C–C≡C–Mn(dmpe)(C5 H4 Me)] complex, however an in situ oxidation with 2 equiv. led to the isolation of the dmpe bridged dinuclear butadiyne complex [(MeC5 H4 )(dmpe)Mn–C≡C-dmpe-C≡C–Mn(dmpe)(C5 H4 Me)]2+ [PF6 ]2 (Scheme 7). The formation of the complex [(MeC5 H4 )(dmpe)Mn–C≡C-dmpe-C≡C–Mn(dmpe)(C5 H4 Me)]2+ implies the presence of the dinuclear butadiyne MnII –MnII complex [(MeC5 H4 )(dmpe)Mn–C≡C–C≡C–Mn(dmpe)(C5 H4 Me)]2+ which is quite electrophilic and in the presence of dmpe, it gets converted to the more stable ylide-type adduct, which was characterized by an X-ray diffraction study.44

Scheme 7

Our attempts to synthesize [(MeC5 H4 )(dmpe)Mn–C≡C–C≡C– Mn(dmpe)(C5 H4 Me)] were unsuccessful and consequently no cyclic voltammetric results were available to estimate the communication between the two metal centers (via the comproportionation constant K c ). However, a comparison between the model complexes [(C5 H5 )(dHpe)Mn–C≡C–C≡C–Mn(dHpe)(C5 H5 )] and [(C5 H5 )(dHpe)Mn=C=CR–RC=C=Mn(dHpe)(C5 H5 )] (R = H, Ph) has been studied through DFT calculations. The parameter DIP, introduced by Floriani et al.43 to give a theoretical estimate of the metal–metal interaction in the case of [{Cp(CO)2 Fe}2 (l-Cn )] (n = 2, 4, 6, 8) and its dependence on the chain length by DFT calculations was utilized. The parameter DIP was obtained via DFT calculations (including simulations of solvation effects) as the difference between the first and second ionization potentials DIP = IP2 − IP1 (IP1 = E tot (model+ ) − E tot (model) IP2 = E tot (model2+ ) − E tot (model+ )) for the three models studied (Table 3). Considering the larger the DIP parameter, the better is the communication between the metal center, the large experimental difference of about 106 between the comproportionation constants of 4 (R = Ph, 8.0 × 103 ) and 5 (R = H, 6.6 × 109 ) may be highlighted by the very

small DIP value of 4-dHpe (1.03 eV) compared to that for 5-dHpe (2.09 eV). The stabilizing perpendicular orientation of the metal fragments observed in the neutral compound of 4-dHpe yields a relatively high first ionization potential of 4.44 eV compared to 3.62 eV for 5-dHpe which is mainly responsible for the small DIP value of 1.03 for the 4-dHpe series. Interestingly, the DIP calculated for the l-C4 model [(C5 H5 )(dHpe)Mn–C≡C–C≡C– Mn(dHpe)(C5 H5 )] is larger than that for [(C5 H5 )(dHpe)Mn]2 (lC4 H2 ) 5-dHpe, we could thus expect an even higher comproportionation constant and consequently a better communication between the metal centers for the l-C4 system. X(dmpe)2 Mn–C4 –Mn(dmpe)2 X rigid-rod species Recently we reported a series of symmetric bis-dmpe MnII and MnIII complexes of the type [Mn(dmpe)2 (X)2 (l-C4 )]n+ (X = C≡CH, 6; C≡CSiMe3 , 7; n = 0, 1).27 Earlier, Lapinte and coworkers reported the carbon–carbon coupling of two molecules [FeCp(dppe)(C≡CH)]+ to yield a dinuclear vinylidene species, which by subsequent deprotonation with KOtBu afforded a dinuclear compound containing a Fe–C4 –Fe unit.58 A related carbon–carbon coupling was observed in our group using the half-sandwich Mn(MeC5 H4 )(dmpe)(C≡CPh) complex.17 We considered utilizing such coupling processes using [Mn(dmpe)2 (X)(C≡CH)][PF6 ] (R = C≡CH; C≡CSiMe3 ) as starting components to generate the mixed-valent complexes [{Mn(dmpe)2 (C≡CH)}2 (l-C4 )][PF6 ] [6]+ and [{Mn(dmpe)2 (C≡CSiMe3 )}2 (lC4 )][PF6 ] [7]+ with moderate to strong bases (DBU, F− or OH− ) (Scheme 8). These mixed-valent complexes [6]+ and [7]+ were isolated as violet solids and the structure of [7]+ has been unequivocally established by a single-crystal X-ray diffraction study.21,27 The intermediate species [Mn(dmpe)2 (C≡CH)(C≡C|)] and [Mn(dmpe)2 (C≡CSiMe3 )(C≡C|)] has been proposed based on low-temperature 1 H NMR studies supported further by DFT calculations.21,27 Another direct synthetic pathway to obtain the d5 low-spin MnII complexes of the type [Mn(dmpe)2 (I)2 (lC4 )] 8 was by treating [MeCpMn(dmpe)I] with half an equivalent of bis(trimethylstannyl)-1,3-butadiyne in the presence of DMPE (Scheme 9). Consequently the mixed valent and the dicationic compounds were obtained by one-electron oxidations. Such intermediates [Mn(dmpe)2 (C≡CH)(C≡C|)] and [Mn(dmpe)2 (C≡CSiMe3 )(C≡C|)] may possess two principal canonical forms: the singlet form A [Mn(dmpe)2 (C≡CR)(C≡C:)] with a formally MnIII center, and the triplet form B [Mn(dmpe)2 (C≡CR)(C≡C• )] with an oxidized C atom and a reduced MnII center. DFT calculations performed on a hydrogen substituted model [Mn(dHpe)2 (C≡CH)(C≡C)] revealed that out of the two possible spin states, the triplet state B is more stable than the singlet state by ca. 90 kJ mol−1 and that the b-alkynyl carbon atom in the triplet state bears a substantial amount of

Table 3 Differences between the first (IP1) and the second (IP2) ionization energies calculated for the considered model systems

[(C5 H5 )(dHpe)Mn]2 (l-C4 ) [(C5 H5 )(dHpe)Mn]2 (l-C4 H2 ) 5-dHpe [(C5 H5 )(dHpe)Mn]2 (l-C4 Ph2 ) 4-dHpe

This journal is © The Royal Society of Chemistry 2007

DE = IP2 –IP1 /eV

IP1 /eV

IP2 /eV

Kc

2.21 2.09 1.03

4.02 3.62 4.44

6.23 5.71 5.47

— 6.6 × 109 8.0 × 103

Dalton Trans., 2007, 1091–1100 | 1097

Scheme 8

Scheme 9

spin density (+0.61a).21,27 The relative triplet stability and the concomitant longevity of the MnII free radical form further enable the selective C–C coupling process to produce initially the neutral dinuclear compounds [{Mn(dmpe)2 (C≡CR}2 (l-C4 )] (R = H, 6; SiMe3 , 7). The Mn center is playing a similar role as the Cu2+ ion in the Eglinton and McCrae coupling of acetylenic compounds69 formally oxidizes the b-carbon atoms of the terminally deprotonated species which is unprecedented. However, the final mixed valent compound results from the oxidation of the dinuclear neutral species by the mononuclear acetylide complex. This unique and peculiar characteristic of the

manganese center makes it a promising candidate for the build up of a terminal functionalized molecular wire. In 1 H NMR studies, most of these complexes show broad resonances since they have paramagnetic character. However, a simple qualitative idea of these complexes could be elucidated. X-Ray diffraction studies are quite vital since they confirm and unequivocally establish the structure of these paramagnetic dinuclear complexes. Additionally, they reveal structural features in accordance with the oxidation states that consequently bring out the cumulenic or the non-cumulenic behaviour in these structurally characterized complexes (Table 4).

˚ ) of the Mn–C4 –Mn unit in [6]+ [PF6 ]− , [7]+ [PF6 ]− , [7]2+ [PF6 ]2 , 8, [8]+ [PF6 ]− , and [8]2+ [BPh4 ]2 − Table 4 Structural features (A

[6]+ [PF6 ]− [7]+ [PF6 ]− [7]2+ [PF6 ]2 − 8 [8]+ [PF6 ]− [8]2+ [BPh4 ]2 −

Mn–C1

C1–C2

C2–C2*

C2–C1*

C1*–Mn*

Ref.

1.818(4) 1.794(13) 1.776(4) 1.798(15) 1.763(2) 1.768(4)

1.285(6) 1.309(17) 1.292(5) 1.263(17) 1.275(3) 1.289(5)

1.307(9) 1.30(2) 1.282(5) 1.33(3) 1.313(5) 1.295(5)

1.285(6) 1.309(17) 1.302(5) 1.263(17) 1.275(3) 1.298(5)

1.818(4) 1.794(13) 1.769(4) 1.795(15) 1.763(2) 1.770(4)

21 27 27 19 19 19

1098 | Dalton Trans., 2007, 1091–1100

This journal is © The Royal Society of Chemistry 2007

The cyclic voltammograms of complexes 7, [7]+ and [7]2+ display two fully reversible waves, one at E 1/2 = −0.816 V corresponding to the MnII –C4 –MnIII ([7]+ )/MnII –C4 –MnII (7) redox couple (DE p = 0.060 V and ipa /ipc ≈ 1 for scan rates of 0.100–0.700 V s−1 ) and the other one at E 1/2 = −0.271 V due to the MnIII –C4 –MnIII ([7]2+ )/MnIII –C4 –MnII ([7]+ ) couple. The difference of these values of DE 1/2 = 0.545 V establishes a comproportionation constant of 2.2 109 (K c = exp(FDE 1/2 /RT))45 (Scheme 10). Additionally, we can observe a small peak attributable to the formation of a MnIV –C4 –MnIII species, which, once produced, it decomposes at the electrode surface remaining throughout the experiment. The CV of the 6/[6]+ /[6]2+ system displays two reversible waves at E 1/2 = −0.451 V and E 1/2 = +0.124 V corresponding to the MnIII –MnII ([6]+ )/MnII –MnII (6) and MnIII –MnIII ([6]2+ )/MnIII – MnII ([6]+ ) redox couples, respectively. The difference of these two values of DE 1/2 = 0.575 V establishes a comproportionation constant K c of 7.5 × 109 .21 The CV of 8 also shows that it an be oxidized in two steps to [8]+ and [8]2+ . The potential difference between the two redox processes of DE 1/2 = 0.63 V establishes a larger comproportionation constant of 1.1 × 1010 (Table 5). The value of K c obtained for [7]+ (2.2 × Table 5 CV data for the complex systems 1, 4, 5, 6, 7 and 8 Complex system

Couple 1

Couple 2

DE

Kc

1/[1]+ /[1]2+ 6/[6]+ /[6]2+ 7/[7]+ /[7]2+ 8/[8]+ /[8]2+ 4/[4]+ /[4]2+ 5/[5]+ /[5]2+

−0.847 +0.124 −0.271 −0.021 −0.445 −0.820

−1.835 −0.451 −0.816 −0.651 −0.656 −1.386

0.988 0.575 0.545 0.630 0.211 0.566

8.6 × 1016 7.5 × 109 2.2 × 109 1.1 × 1010 8 × 103 6.6 × 109

109 ) is somewhat lower than that obtained for complexes [6]+ and [8]+ and even marks a value at the lower end of the series of [{Mn(dmpe)2 (X)}2 (l-C4 )]n+ complexes (X = I, 1.1 1010 ; C≡CH, 7.5 109 ; C≡CSiMe3 , 2.2 109 ; C4 –SiMe3 , 1.8 108 ).18,19,21 UV-Vis-NIR spectra of the series of complexes 6, [6]+ and [6]2+ , and 7, [7]+ and [7]2+ have been recorded, which showed wellstructured spectra with two intense bands for all these complexes. For [6]+ and [6]2+ , and [7]+ and [7]2+ one additional band of low intensity at shorter wavelengths was observed. In comparison with the spectra of 6 and [6]2+ , and 7 and [7]2+ , respectively, [6]+ and [7]+ did not reveal an extra broad absorption at the long wavelength end, which could be assigned to a mixed-valent band6,23,26,45–53 as expected for class II systems according to the Robin–Day classification. Furthermore, it is quite remarkable to see how strongly the systems are affected by charge, since the two major transitions, which are expected to arise from the same types of orbitals, shift to shorter wavelengths on going from 6 to [6]+ , or from 7 to [7]+ . It is presumably due to the filled orbitals in the HOMO region of 6 or 7, which are particularly sensitive to positive charge and are therefore significantly lowered in energy. Furthermore the oxidation of 7 with the more strongly coupled Mn centers is at a still more negative potential (−0.816 V) than that of 6 (−0.451 V). DFT calculations including simulations of solvation effects on the model systems {[Mn(dHpe)2 R]2 (l-C4 )}n+ (n = 0, 1, 2; R = C≡CH, C≡C–SiMe3 , I) confirmed energetically high-lying HOMOs for the neutral compounds at −3.40 eV (R = C≡CH), at −3.47 eV (R = C≡C–SiMe3 ) and at −3.77 eV (R = I).27 The triplet state of the neutral compounds is preferred over the singlet state by about 30 kJ mol−1 for all models, while the ground states of mono- and di-cationic species have been found as low spin configurations with one and zero unpaired electrons, respectively. Unfortunately, a theoretical estimation of the metal– metal interaction for 6, 7 and 8 based on their HOMO energies, computed gas-phase ionization potentials and/or the parameter DIP introduced by Floriani et al.43 could not be obtained. The latter seems to be inappropriate to study the ligand influences at least in the given dinuclear manganese systems. However, some trends have been highlighted elsewhere.27

Conclusion We believe that the above in-depth study of the dinuclear manganese complexes highlights the significant role played by the metal centers and ligands in determining the electronic nature of these complexes. These results could provide an impetus for the development of new systems from which one would hope such molecular wires are ultimately wired into circuits given the rapid progression of the present day technology. It is expected that we will see a rapid development in this specific area and applications in electronic devices in the next decade.

References

Scheme 10

This journal is © The Royal Society of Chemistry 2007

1 J. M. Lehn, Supramolecular Chemistry—Concepts and Perspectives, Wiley, Weinheim, Germany, 1995. 2 M. B. Robin and P. Day, Adv. Inorg. Chem. Radiochem., 1967, 10, 247. 3 K. Sonogashira, S. Takahashi and N. Hagihara, Macromolecules, 1977, 10, 879. 4 K. Sonogashira, Y. Fujikura, T. Yatake, N. Toyoshima, S. Takahashi and N. Hagihara, J. Organomet. Chem., 1978, 145, 101.

Dalton Trans., 2007, 1091–1100 | 1099

5 W. Q. Weng, T. Bartik, M. Brady, B. Bartik, J. A. Ramsden, A. M. Arif and J. A. Gladysz, J. Am. Chem. Soc., 1995, 117, 11922. 6 M. Brady, W. Q. Weng, Y. L. Zhou, J. W. Seyler, A. J. Amoroso, A. M. Arif, M. Bohme, G. Frenking and J. A. Gladysz, J. Am. Chem. Soc., 1997, 119, 775. 7 M. Akita, M. C. Chung, A. Sakurai, S. Sugimoto, M. Terada, M. Tanaka and Y. Morooka, Organometallics, 1997, 16, 4882. 8 R. Denis, L. Toupet, F. Paul and C. Lapinte, Organometallics, 2000, 19, 4240. 9 J. Gil-Rubio, M. Laubender and H. Werner, Organometallics, 2000, 19, 1365. 10 M. I. Bruce, B. G. Ellis, B. W. Skelton and A. H. White, J. Organomet. Chem., 2000, 607, 137. 11 T. Ren, G. Zou and J. C. Alvarez, Chem. Commun., 2000, 1197. 12 T. Rappert, O. Nurnberg and H. Werner, Organometallics, 1993, 12, 1359. 13 V. W. W. Yam, V. C. Y. Lau and K. K. Cheung, Organometallics, 1996, 15, 1740. 14 (a) F. Coat, M. A. Guillevic, L. Toupet, F. Paul and C. Lapinte, Organometallics, 1997, 16, 5988; (b) M. Guillemot, L. Toupet and C. Lapinte, Organometallics, 1998, 17, 1928. 15 C. Joachim, J.-P. Launay and S. Woitellier, Chem. Phys., 1990, 147, 131. ¨ 16 D. Unseld, PhD Thesis, University of Zurich, Switzerland, 1999. 17 D. Unseld, V. V. Krivykh, K. Heinze, F. Wild, G. Artus, H. Schmalle and H. Berke, Organometallics, 1999, 18, 1525. ¨ 18 S. Kheradmandan, PhD Thesis, University of Zurich, Switzerland, 2001. 19 S. Kheradmandan, K. Heinze, H. W. Schmalle and H. Berke, Angew. Chem., 1999, 111, 2412; S. Kheradmandan, K. Heinze, H. W. Schmalle and H. Berke, Angew. Chem., Int. Ed., 1999, 38, 2270. 20 F. J. Fernandez, M. Alfonso, H. W. Schmalle and H. Berke, Organometallics, 2001, 20, 3122. 21 F. J. Fernandez, O. Blacque, M. Alfonso and H. Berke, Chem. Commun., 2001, 1266. 22 V. V. Krivykh, I. L. Eremenko, D. Veghini, I. A. Petrunenko, D. L. Pountney, D. Unseld and H. Berke, J. Organomet. Chem., 1996, 511, 111. 23 T. Bartik, B. Bartik, M. Brady, R. Dembinski and J. A. Gladysz, Angew. Chem., Int. Ed. Engl., 1996, 35, 414; T. Bartik, B. Bartik, M. Brady, R. Dembinski and J. A. Gladysz, Angew. Chem., Int. Ed. Engl., 1996, 35, 414. 24 R. Dembinski, T. Lis, S. Szafert, C. L. Mayne, T. Bartik and J. A. Gladysz, J. Organomet. Chem., 1999, 578, 229. 25 R. Dembinski, T. Bartik, B. Bartik, M. Jaeger and J. A. Gladysz, J. Am. Chem. Soc., 2000, 122, 810. 26 F. Paul, W. E. Meyer, L. Toupet, H. J. Jiao, J. A. Gladysz and C. Lapinte, J. Am. Chem. Soc., 2000, 122, 9405. 27 F. J. Fernandez, K. Venkatesan, O. Blacque, M. Alfonso, H. W. Schmalle and H. Berke, Chem.–Eur. J., 2003, 9, 6192. 28 S. Kheradmandan, K. Venkatesan, O. Blacque, H. W. Schmalle and H. Berke, Chem.–Eur. J., 2004, 10, 4872. 29 H. Ogawa, T. Joh, S. Takahashi and K. Sonogashira, J. Chem. Soc., Chem. Commun., 1985, 1220. 30 H. Ogawa, K. Onitsuka, T. Joh, S. Takahashi, Y. Yamamoto and H. Yamazaki, Organometallics, 1988, 7, 2257. 31 R. M. Bullock, F. R. Lemke and D. J. Szalda, J. Am. Chem. Soc., 1990, 112, 3244. 32 F. R. Lemke, D. J. Szalda and R. M. Bullock, J. Am. Chem. Soc., 1991, 113, 8466. 33 J. A. Ramsden, W. Q. Weng, A. M. Arif and J. A. Gladysz, J. Am. Chem. Soc., 1992, 114, 5890. 34 E. O. Fischer and H. Z. Leipfinger, Z. Naturforsch., B, 1955, 10, 353. 35 G. Wilkinson, F. A. Cotton and J. M. Birmingham, J. Inorg. Nucl. Chem., 1956, 2, 95.

1100 | Dalton Trans., 2007, 1091–1100

36 C. G. Howard, G. S. Girolami, G. Wilkinson, M. Thornton-Pett and M. B. Hursthouse, J. Am. Chem. Soc., 1984, 106, 2033. 37 J. Heck, W. Massa and P. Weinig, Angew. Chem., 1984, 96, 699; J. Heck, W. Massa and P. Weinig, Angew. Chem., Int. Ed. Engl., 1984, 23, 722. 38 S. Kheradmandan, H. W. Schmalle, H. Jacobsen, O. Blacque, T. Fox, H. Berke, M. Gross and S. Decurtins, Chem.–Eur. J., 2002, 8, 2526. 39 N. Hebendanz, F. H. Kohler, G. Muller and J. Riede, J. Am. Chem. Soc., 1986, 108, 3281. 40 F. H. Kohler and B. Schlesinger, Inorg. Chem., 1992, 31, 2853. 41 H. Jacobsen and H. Berke, Chem.–Eur. J., 1997, 3, 881. 42 D. C. Young, Computational Chemistry: A Practical Guide for Applying Techniques to Real-World Problems, Wiley, New York, 2001. 43 P. Belanzoni, N. Re, A. Sgamellotti and C. Floriani, J. Chem. Soc., Dalton Trans., 1998, 1825. ¨ 44 K. Venkatesan, PhD Thesis, University of Zurich, Switzerland, 2003; K. Venkatesan, H. W. Schmalle and H. Berke, unpublished results. 45 C. Creutz, Prog. Inorg. Chem., 1983, 30, 1. 46 N. Chanda, R. H. Laye, S. Chakraborty, R. L. Paul, J. C. Jeffrey, M. D. Ward and G. K. Lahiri, J. Chem. Soc., Dalton Trans., 2002, 3496. 47 N. C. Harden, E. R. Humphrey, J. C. Jeffrey, S. M. Lee, M. Marcaccio, J. A. McCleverty, L. H. Rees and M. D. Ward, J. Chem. Soc., Dalton Trans., 1999, 2417. 48 R. H. Laye, S. M. Couchman and M. D. Ward, Inorg. Chem., 2001, 40, 4089. 49 N. S. Hush, Coord. Chem. Rev., 1985, 64, 135. 50 N. S. Hush, Prog. Inorg. Chem., 1967, 8, 391. 51 M. V. Russo, C. Lo Sterzo, P. Franceschini, G. Biagini and A. Furlani, J. Organomet. Chem., 2001, 619, 49. 52 T. Weyland, K. Costuas, L. Toupet, J. F. Halet and C. Lapinte, Organometallics, 2000, 19, 4228. 53 M. Younus, N. J. Long, P. R. Raithby and J. Lewis, J. Organomet. Chem., 1998, 570, 55. 54 R. S. Iyer and J. P. Selegue, J. Am. Chem. Soc., 1987, 109, 910. 55 R. L. Beddoes, C. Bitcon, A. Ricalton and M. W. Whiteley, J. Organomet. Chem., 1989, 367, C21. 56 M. I. Bruce, M. P. Cifuentes, M. R. Snow and E. R. T. Tiekink, J. Organomet. Chem., 1989, 359, 379. 57 N. Le Narvor and C. Lapinte, J. Chem. Soc., Chem. Commun., 1993, 357. 58 N. Le Narvor, L. Toupet and C. Lapinte, J. Am. Chem. Soc., 1995, 117, 7129. 59 R. L. Beddoes, C. Bitcon, R. W. Grime, A. Ricalton and M. W. Whiteley, J. Chem. Soc., Dalton Trans., 1995, 2873. 60 C. Lowe, H. U. Hund and H. Berke, J. Organomet. Chem., 1989, 372, 295. 61 M. I. Bruce, Chem. Rev., 1991, 91, 197. 62 N. G. Connelly, M. P. Gamasa, J. Gimeno, C. Lapinte, E. Lastra, J. P. Maher, N. Lenarvor, A. L. Rieger and P. H. Rieger, J. Chem. Soc., Dalton Trans., 1993, 2575. 63 C. Bitcon and M. W. Whiteley, J. Organomet. Chem., 1987, 336, 385. ¨ 64 C. LoweHans, U. Hund and H. Berke, J. Organomet. Chem., 1989, 372, 295. 65 A. Antinolo, A. Otero, M. Fajardo, C. Garcia-Yebra, R. Gil-Sanz, C. Lopez-Mardomingo, A. Martin and P. Gomez-Sal, Organometallics, 1994, 13, 4679. 66 D. A. Valyaev, M. G. Peterleitner, L. I. Leont’eva, L. N. Novikova, O. V. Semeikin, V. N. Khrustalev, M. Y. Antipin, N. A. Ustynyuk, B. W. Skelton and A. H. White, Organometallics, 2003, 22, 5491. 67 K. Venkatesan, O. Blacque, T. Fox, M. Alfonso, H. W. Schmalle, S. Kheradmandan and H. Berke, Organometallics, 2005, 24, 920. 68 K. Venkatesan, T. Fox, H. W. Schmalle and H. Berke, Eur. J. Inorg. Chem., 2005, 901. 69 G. Eglinton and W. McRae, J. Chem. Soc., 1963, 2295.

This journal is © The Royal Society of Chemistry 2007