Dissertation to obtain the doctoral degree of the faculty of chemistry at the Ruhr-University of Bochum, Germany. by Andreas Kempter

Transcription

1 Coordination chemistry of the low-valent group 13 HC-analogue Ga(DDP) (DDP = bulky bis-imidinate) to transition metal centers: Insertions, substitution reactions and cluster formation. Dissertation to obtain the doctoral degree of the faculty of chemistry at the uhr-university of Bochum, Germany by Andreas Kempter

2

3 eferees: Prof. Dr. oland A. Fischer Prof. Dr. William S. Sheldrick Thesis defense date:

4

5 The present study was performed in between July 2004 and June 2007 at the Chair of Inorganic Chemistry II, Organometallics and Materials Chemistry at the uhr-university Bochum, Germany. Herewith, I would like to authenticate, that I have written this thesis on my own. I have not used other resources, tools or assistance than those reported in this thesis. Andreas Kempter, July 2007

6

7 Sincere thanks goes to Prof. Dr. oland Fischer, for his great confidence, kindness in every respect, constructive talks and ideas, marvellous support in the realisation of the Ende aller Tage and tremendous motivation all over the time.

8

9 I d further like to thank: Dr. Christian Gemel for his helpful suggestions and continous interest throughout this work, his manifold help, his open ear for all kinds of problems and support all over the time. Dr. Arne Baunemann for his help in both science and private life, for interesting talks, a lot of fun and ongoing friendship. Dr. Stephan Hermes for his catching enthusiasm in everything he did, his open ear for all my problems and his motivation and partnership in planning and performing the Ende aller Tage. Especially, I like to thank my office, Dr. Arne Baunemann, Cand. Chem. Timo Bollermann, Dipl. Chem. Beatrice Buchin, Dipl. Chem. Thomas Cadenbach, Dipl. Chem. Malte Hellwig, M. Sc. Todor Hikov, Cand. Chem. Maike Müller, Dr. Marie-Kathrin Schröter and Cand. Chem. Denise Zacher for the convenient and easygoing atmosphere, your support in all areas and a barrel of laughs in all the years. I am very grateful that I had the chance for having worked together with Beatrice Buchin, who regrettably passed away last Christmas, Thomas Cadenbach and Maike Müller, whose ideas, suggestions and great efforts allow certain results of this work. Also, thanks to Dipl. Chem. Felicitas Schröder and Dr. Mirza Cokoja as my direct Lab-neighbours, who helped me with motivating talks and tried to give me new perspectives, when the results are missing. Ms. Sabine Pankau, who always had an open ear for me and any kind of administrative problems and allows to relax in certain talks about all the world and his brother. Dr. Christian Gemel, Dr. Klaus Merz, Priv. Doz. Iris Oppel geb. Müller and Ms. Manuela Winter who helped me at their best by analyzing the crystal structures and the calculation of the structural parameters, as well as Mr. Hans-Jochen Hauswaldt, Mr. Martin Gartmann and Mr. Gregor Barchan for their help in all questions concerning the M spectroscopy and Ms. Karin Bartholomäus for performing the Elemental analysis.

10 Dr. aghunandan Bhakta, Mr. olf Deibert, Jun. Prof. Anjana Devi; Dr. Eliza Gemel, Ms. Heike Gronau-Schmid, Dr. ed J. Hardman, Ms. Ursula Hermann, Dr. Jayaprakash Khanderi, Dr. Emmanuel Lamouroux,. Dr. Eva Maile, M. Sc. Andrian Milanov, Dr. Harish Parala, Dr. Urmila Patil, Dr. amasamy Pothiraja, Dipl. Chem. Daniel ische, Dr. ochus Schmid, Dr. Jelena Sekulic, Dipl.-Ing. Stephan Spoellmann, Dr. Tobias Steinke, Dr. Maxim Tafipolsky and M. Sc. Xiaoning Zhang for all their help, their support to successfully perform the Ende aller Tage, their good advises and the open-mindedness for my concerns. Dr. Mathew C. Kuchta for his help in correcting this thesis. Also, I d like to thank the Harrys: Cand. Chem. Daniel Esken, Cand. Chem. Tobias Thiede, Cand. Chem. Markus Halbherrr, Cand. Chem. Mikhail Meilikov and M. Sc. Saeed Amirjalayer, who brought new life in the chair and established our nice Poker round. Thanks for the good times and your money I was allowed to win ;-). Furthermore, I m very greateful for my girlfriend Ms. Karin Wieczorek, who encourages me in good and bad times, always excuses my bad moods if chemistry doesn t work properly and gives me the support I needed to go on further. Thank you for being at my side. Finally, I d like to thank my beloved parents, Ms. Brigitte Kempter and Mr. Hans Will, who did everything they could to enable my personal development and my scientific studies. I really appreciate your efforts, which means so much to me. I hope I can pay back the confidence you had in me.

11 dedicated to Karin, who means everything to me.

12

13 Table of contents. Table of contents: I. Introduction II. State of knowledge... 4 Synthesis of low-valent group 13 compounds E I Bonding situation in E I -compounds Bonding situation in TM-E I -complexes σ-donor properties π-acceptor properties Coulombic interactions Synthetic methods for the preparation of M-E I compounds Coordination chemistry of E I Cp* and E I C(SiMe 3 ) Substitution of labile ligands Insertion reactions of ECp* into metal-halide bonds Coordination chemistry of neutral four-, six-, and anionic five-membered heterocycles E I eutral four-membered heterocycles eutral six-membered heterocycles Previous results from our group Anionic five-membered heterocycles Bond activation reactions in TM-group 13 metal complexes Summary and goals of this study III. esults and discussion Insertion reaction of Ga(DDP) into M-X bonds Insertion of Ga(DDP) into h-cl bonds Insertion of Ga(DDP) into the h-cl bond of [h(coe) 2 Cl] Insertion reaction of Ga(DDP) into Zn-Me and Zn-Cl bonds Insertion reaction of Ga(DDP) into the Zn-methyl bond of ZnMe Insertion reaction into the Zn-Cl bond of ZnCl Insertion reaction of Ga(DDP) into Gallium-CH 3 bonds eaction of Ga(DDP) with AlMe 3 and GaMe 3 : Insertion of Ga(DDP) into the Ga-methyl bond of GaMe Synthesis of [{(DDP)GaMe} 2 GaMe] I

14 Table of contents Insertion into group 14 element bonds Insertion into Sn-Cl bonds of SnCl 2 Me Insertion reaction of Ga(DDP) into the Si-Cl bond of SiCl Insertion reaction of Ga(DDP) into C- and C-Cl bonds Insertion reaction of Ga(DDP) into C- bonds Insertion reaction of Ga(DDP) into C-Cl bonds Summary and conlusion Cationic Ga(DDP) complexes by halide abstraction Synthesis of the cationic complex [{(DDP)Ga. THF} 2 Au][B(Ar F ) 4 ] (10. 2THF) Cationic Complex with Zn-Ga bonds Substitution reactions of labile olefins with E(DDP) Substitution reactions on i(0) centers eactions of E(DDP) with [i(cod) 2 ] eaction of [i(cod) 2 ] with Ga(DDP) eaction of [i(cod) 2 ] with Al(DDP) Adduct formation in the reaction of [i(cdt)] with Ga(DDP) Substitution reactions of [(cdt)i{ga(ddp)}] with ethylene eaction of [(cdt)i{ga(ddp)}] with other olefins eaction of [(cdt)i{ga(ddp)}] with styrene eaction of [(cdt)i{ga(ddp)}] with dvds monomeric Pd(0)-complexes Substitution reactions on a Pd 0 -center. eaction of [Pd 2 (dvds) 3 ] with E(DDP) Oxidative addition of H 2 and H-SiEt 3 to [Pt(1,3-cod){Ga(DDP)} 2 ] and H/D exchange with C 6 D Oxidative addition of H 2 to [Pt(1,3-cod){Ga(DDP)} 2 ] Oxidative addition of HSiEt 3 to [Pt(1,3-cod){Ga(DDP)} 2 ] Compounds with higher nuclearity Dinuclear compounds Compounds with a i 2 -core eactions of [i(c 2 H 4 ) 3 ] with Ga(DDP) eactivity of [i(cod) 2 ] towards Ga(DDP) in the presents of PhCCPh Compounds with a Pd 2 -core Substitution reactions on [(dvds)pd{ga(ddp)}] II

15 Table of contents eaction of [Pd 2 (dvds) 3 ] with Al(DDP) Ligand substitution in [Pd 2 (dvds) 2 (μ 2 -Al(DDP)] (23) Dimeric compounds with a Pt 2 core eaction of [(1,3-cod)Pt{Ga(DDP)} 2 ] with t BuC Synthesis of the dimeric hydride complex [Pt{μ 2 -Ga(DDP)}(H 2 )] 2 (26) Trinuclear compounds Cluster compounds with a i 3 core : Synthesis of the large cluster [Sn 17 {ClGa(DDP)} 4 ] IV. Conclusion and outlook V. Experimental section General remarks Solvents Analytical methods Elemental analysis M-Spectroscopy Single-crystal x-ray diffraction Computational Details Compounds Syntheses VI. Literature index VII. Appendix Computational details Crystallographic data for the new compounds List of publications Posters and presentations Curriculum vitae Supplementary electronic files III

16

17 I. Introduction. I. Introduction. The surge in interest in the coordination chemistry of aluminium, gallium and indium at transition metal centers originates from two different motivations. One the one hand, organometallic complexes of the type [(L n M)(E I ) b ] (M = transition metal, E = group 13 metal) are regarded as potential precursors for applications in materials science. Beyond applications for metallorganic chemical vapour deposition (MO-CVD), M-E compounds are used in heterogeneous catalysis as well as for the synthesis of intermetallic nanophases. [1-14] The second motivation stems from a resurgence of the chemistry of low-valent group 13 metal (earth metals) compounds E I (E = Al, Ga, In). In contrast to the coordination chemistry of transition metal borylene complexes that has been recently pushed forward namely by H. Braunschweig, [15, 16] low-valent group 13 organyls E I are stable and nowadays quite well accessible. In this respect, the tetrameric [{AlCp*} 4 ], [17-19] the hexameric [{ECp*} 6 ] (E = Ga, In), [20-23] the related alkyl compounds [{E} 4 ] ( = C(SiMe 3 ) 3, CH(SiMe 3 ) 2, Si(SiMe 3 ) 3 ; E = Al, Ga, In), [24-27], as well as the electron-rich neutral monomeric compounds [E I (Tp tbu )] (Tp tbu = tris-(3,5-di-tert-butylpyrazoyl)hydroborat, E = Ga, In) [28, 29], [E I (DDP)] (DDP = 2,6-bis(diisopropylphenylamino)-4-diisopropylphenylimino-2-pentene; E = Al Tl), [30-33] E I Mes* (Mes* = 2,6-bis(2,5,6-triisopropylphenyl)phenyl, E = Al, Ga, In), [34-37] [E(Giso)] (Giso = bis-(2,6-diisopropylphenyl)guanidinato; E = Ga, In, Tl) [38] and the anionic imidazolate [Ga{()C(H)} 2 ] - ( = t Bu, 2,6-diisopropylphenyl) [39, 40] have recently been isolated and ilustrate the structural diversity of E I compounds. Besides their structural chemistry, the potential of these compounds to be suitable ligands for the synthesis and characterisation of novel transition metal-group 13 metal complexes [L n M(E I ) b ] and clusters has been in the focus of research in the last two decades. [41-43] However, the close vicinity of very electrophilic metals E and highly nucleophilic transition metals is an intrinsically interesting feature of the [L n M(E I ) b ] complexes which has stimulated further studies in both preparative, [43-46] as well as theoretical [30, 47-53] chemistry. In particular, the coordination chemistry of the E I Cp* compounds towards π-acceptor-free transition metal compounds was investigated in more detail. Thus, homoleptic complexes of the type [M(ECp*) 4 ] (M = Pd, Pt) can be used as building blocks for dimeric clusters [MPt(GaCp*) 5 ] [54], whereas the complexes [M a (ECp*) b ] (M = Pd, Pt, E = Al, Ga, In; b > a ) are found to undergo exchange reactions and show unusual fluxionality in solution. [55] Moreover, it was found that AlCp* acts as directing ligand in C-H and Si-H activation, which proceed via the reactive, unsaturated intermediate [i(alcp*) 3 ]. [56] Additionally, the insertion 1

18 I. Introduction. of E I Cp* compounds into transition metal halide compounds was also reported in detail [45, 57-59] (Chapter II). In contrast to monovalent organometallic E I species, the coordination chemistry of the electron-rich, low valent group 13 heterocycles is relatively unexplored and thus, a comprehensive study of their coordination chemistry towards transition metal centers is of interest. Whereas a few studies concerning the ability of the four-membered heterocycles [Ga{()C(H)} 2 ] - to act as potential ligands for transition metals has been reproted, [38, 60] the main focus of anionic five-membered heterocycle chemistry is led, besides only some examples for transition metals, [61-65] towards main group metals. [44, 62, 65-69] For the neutral sixmembered heterocycle, a lot of effort, especially for the Al congener, was made concerning the reactivity towards group compounds, [70-79] but also examples of its transition metal [80, 81] chemistry are reported for Ga(DDP) with the late transition metals h and Au. In the present study, the synthesis, structure and reactivity of new π-acceptor-free transition metal compounds of the low-valent group 13 heterocycles E(DDP) (E = Al, Ga) have been investigated. Their increased steric demand (cone angle 147 vs. 112 ) as well as their electronic properties, which differ from ECp* compounds (e.g. stronger σ-donors, see chapter II), make these ligands promising candidates for the synthesis and stabilization of new unsaturated complexes as well as interesting cluster compounds. Furthermore, they may also be suitable for stabilizing intermediates in interesting bond activation reactions (i.e. C-H activations). In this context, classical organometallic methods were used to form metal-group 13 bonds, from insertion of the low-valent group 13 organyl into transition metal-halide or metal methyl bonds and the substitution of labile ligands like olefins by Ga(DDP). Basic questions in this respect are the following: Does the insertion chemistry of Ga(DDP) significantly differs from that reported for the Cp* analogue due to differences in both steric as well as electronic properties of the ligands (Chapter II)? And more importantly, does the use of the more terically demanding Ga(DDP) facilitate the stabilization of unsaturated complexes or intermediates of the type [M(E I ) n ]. Such intermediates are belived to play a significant role in activation reactions [56] and thus, exploration of this chemistry may allow a deeper insight into such reactions. A closer look at the insertion behaviour of Ga(DDP) into metal-halide and metal methyl bonds (chapter III.1) shows that this, however less common pathway of generating M-E I bonds is also suitable for Ga(DDP) and furthermore, not restricted to transition metals. In 2

19 I. Introduction. several cases, the insertion into main group-halide and -methyl bonds occurs, giving the expected {XGa(DDP)} moieties coordinated at the metal centers. Besides the general stability of these transition- and main-group metal compounds, some results suggest a comparably weak bond between the chlorine and Ga centers, which allows the removal of the halide atoms. Thus, the synthesis of some examples of stable cationic complexes coordinated by the bulky bis-imidinate Ga(DDP) are demonstrated and the resulting compounds show, that an increase in the electrophilicity is observed when abstracting the halogen atom (chapter III.2). Unsaturated complexes of the late transition metals i, Pd and Pt, synthesized by the substitution of labile olefins, are presented in chapter III.3. These compounds are, in contrast to the sterically and electronically saturated monomeric complexes [M(E I ) 4 ], not kinetically inert and the coordinated olefins can be replaced by several other ligands such as different olefins and C. Whereas the use of HCs and phosphanes as directing ligands in classical organometallic chemistry is quite common, recent results show that AlCp* may also act in similar fashion as a directing ligand. [56] However, the oxidative addition of small molecules (i.e. H 2 or HSiEt 3 ), together with an observed H/D exchange reaction and the catalytic hydrogenation of olefins, which are discussed in chapter III.3.3, demonstrate, that the sterically demanding ligand Ga(DDP) also exhibits some potential in this respect. According to E I Cp*, the ability of E I (DDP) to effectively bridge two transition metal centers in a μ 2 - coordination mode is used to synthesize di- and trimeric complexes, which, depending on the co-ligands (e.g. olefins), also exhibit interesting reactivity (chapter III.4). Moreover, the reducing capability and the concomitant sterically stabilizing effect of Ga(DDP) allows the synthess of a large Sn 17 -cluster, which is a very promising finding, pointing to the general ability of this type of ligands not only to stabilize small and unsaturated complexes, but also to be used as suitable ligands for the stabilization of naked metal clusters or nanopartikels, as previously reported for E I Cp*. [82] On the basis of these results, a fruitful and promising development of the coordination chemistry of E I to d-metal fragments can be anticipated in the near future. 3

20 II. State of knowledge. II. State of knowledge. In contrast to low valent group 13 halides of the heavier group 13 elements indium and thallium, the analogous Al I X and Ga I X species are generally unstable at room temperature and tend to disproportionate. The pioneering work of Schnöckel, Uhl, oesky, Power, Parkin and Jutzi demonstrated the ability to stabilize such low-valent group 13 compounds by using sterically demanding and electronically stabilizing groups (Cp*, C(SiMe 3 ) 3, t Bu, 2,6- disubstituted aryl groups, bulky bisimidinates and guanidinates, etc.) (Chart 1). Such lowvalent group 13 organyls E I (E = Al, Ga, In) are now well accessible in high yields and purity. Thus, many publications concerning preparative as well as theoretical investigations of [43, 49, 83-88] these compounds were published in recent years. Cy 2 Ga E E E({ArCH} 2 ) - ECp* E = Al, Ga, In, Tl E(Giso) E = Ga, In, Tl 2 1 H B E E 1 E 1 1 E(Tp) E = Ga, In = Me, tbu, Ph,... 2 E(Mes*) E = Ga, In, Tl 1 = i Pr; 2 = i Pr, t Bu E(DDP) E = Al, Ga, In, Tl Chart 1: Examples of stable, low-valent group 13 derivatives E I. The accessibility and stability at room temperature of these compounds opened the door for a detailed investigation, especially for ECp*, of the potential of E I ligands to act as metalloid, yet exotic ligands in classical inorganic and organometallic chemistry. 4

21 II. State of knowledge. Synthesis of low-valent group 13 compounds E I. In 1991, the first example of a room-temperature stable organo-al I -compound, [{AlCp*} 4 ], was described by the group of Schnöckel et al., synthesized by the reaction of metastable AlCl-solution in THF with MgCp* 2 (yield 44%). [17] This achievement shows, that 6π-electron arene systems like substituted cyclopentadienid-anions (Cp*, Cp (SiMe 3 ) 2, Cp (Benzyl) 5) are suitable ligands to stabilize low-valent main group 13 metal centers, due to their steric demand as well as good electron-donating properties. Consequently, the Ga(I) derivative [{GaCp*} 6 ] was synthesized only two years later by substitution of the halide from metastable Ga-Cl solution with the Cp* anion. [20] A more convenient route was reported by oesky et al for [{AlCp*} 4 ] (yield 20%) [18, 22] and Jutzi et al. for [{GaCp*} 6 ] (yield 70%), [22] by simple reductive dehalogenation of [Cp*AlCl 2 ] or [Cp*GaI 2 ] in the presence of potassium, although in comparatively lower yields for Al. A similar method was used to synthesize alkylsubstituted Al(I) and Ga(I) compounds, for example [{AlC(SiMe 3 ) 3 } 4 ]. [26] Today s state of the art synthesis of [{GaCp*} 6 ] (yield 52%) [23] is the salt-elimination reaction between KCp* and Green s metastable GaI, prepared via ultrasonication of a mixture of Ga-metal and I 2 in organic solvents (e.g. C 6 H 6 ). [89] E Chart 2: ECp* derivatives. ECp* E = Al, Ga, In, Tl. In contrast to the generally unstable Al(I) and Ga(I) halide species, the commercially available InCl was used to synthesize In(I) organyls, easily prepared from InCl and lithium organyls, as shown for [{InC(SiMe 3 ) 3 } 4 ] by Uhl et al. [27] and Beachley et al. in the case of [{InCp*} 6 ]. [90] Additionally, the heaviest congener [TlCp*] can be similarly prepared using commercially available TlCl as Tl(I) source. [91] ormally, theses low-valent group 13 derivatives exist as E E bonded tetramers or hexamers in the solid state, whereas monomeric species exist in solution as well as in the gas phase. Thus, [{AlCp*} 4 ] consists of regular tetrahedral arrangement of aluminium atoms in the solid 5

22 II. State of knowledge. state, each of which is capped by a η 5 -Cp* ring. [92] The intermetallic distances in [{AlCp*} 4 ], as well as the related alkyl-compounds [{EC(SiMe 3 ) 3 } 4 ] (E = Al, Ga, In), are shorter or very close to those in the bulk material (Table 1), indicating, that the interactions between the AlCp* units may be stronger than those in Al-metal. These strong E-E interactions are suggested to be the reason for the cluster formation. [41] According to this, an association / dissociation equilibrium was observed for AlCp* in solution (Scheme 1), but monomeric AlCp* units could be identified only above room temperature by 27 Al-M spectroscopy. [93, 94] Cp* Al *Cp *Cp Al Al Al Cp* Δ T 4 AlCp* Cp* Ga Ga Cp* 6 GaCp* *Cp Ga Ga Cp* Ga *Cp Ga Cp* Scheme 1: Solution behaviour of [{AlCp*} 4 ] and [{GaCp*} 6 ]. Octahedral clusters of the type [{ECp*} 6 ] are observed in the solid state for the Ga(I) and In(I) analogues, whereas TlCp* forms a polymeric zig-zag chain of alternating thallium atoms and Cp* rings. [91] In hexameric [{ECp*} 6 ] (E = Ga, In), the E-E bond distances are comparably long (Table 1), suggesting rather weak E I -E I -interactions. Accordingly, the ECp*- polyhedron formation in the solid state (e.g. for Ga) is caused mainly by additionally van-der- Waals-interactions of the organic envelope rather than M-M interactions. [20, 21] This fact is further supported by 71 Ga-M spectroscopy which shows no changes in the resonances for GaCp* when varying the temperature indicating that GaCp* exists in the monomeric GaCp* form in solution and is not involved in association equilibria (Scheme 1). [41] Finally, the polymeric TlCp* also shows no Tl-Tl interactions in solution. 6

23 II. State of knowledge. The search for low-valent E I compounds with a lower degree of aggregation than four in the solid state was in the focus of research in the following years. Besides electronic reasons, the steric properties of the organic ligands are crucial for the stabilization of E I (Table 2). In 1992, Schuhmann, Ghodsi and Esser synthesized the first group 13 compound [η 5 -C 5 Me 4 (PPh 2 )]In, which was monomeric in the solid state. [95] The simple variation of the Cp-ligand prevents the aggregation to clusters. However, with an In-In distance of 5.929(1) Å, a weak interaction between neighbouring In atoms is observed. Table 1: E-E bond distance of ECp* derivatives. Compound E-E distance [{ECp*} n ] E-E distance (bulk) [96] [{AlCp*} 4 ] [17] av Å 2.86 Å [{GaCp*} 6 ] [21] 4.07 and 4.17 Å Å [{InCp*} 6 ] [17] av Å 3.36 and 3.94 Å [17] {TlCp*] n av and 3.43 Å Ligands of the Tp-type (Tp = Tris(pyrazoyl)hydroborato-) exhibit Tolman-angles [97] of up to 183 on coordination to transition metals, depending on the substituents on the pyrazol rings (Table 2). This steric restraint was used in 1994 to stabilize the first In I compound without any In-In contacts in the solid state. [98] Whereas the related Al I compound [Tp tbu2 ]Al could not yet be synthesized, the related Ga I -derivative [Tp tbu2 ]Ga (Tp tbu2 = tris(3,5-ditertbutylpyrazolyl)hydroborato-) eventually was synthesized in 1996 by Parkin et al. [28] Analogously to the preparation of [{GaCp*} 6 ], [Tp tbu2 ]Ga is obtained by the reaction of GaI with the sodium salt [Tp tbu2 ]a, respectively. In both In- and Ga-species, the bulky, tridentate [Tp tbu2 ]-ligand prevents association by coordinatively saturating and sterically stabilizing the group 13 metal center H B E E 1 1 E = Ga, In = Me, t Bu, Ph,... 2 E = Ga, In, Tl 1 = i Pr; 2 = i Pr, t Bu Chart 3: E I Tp- and E I (alkyl) derivatives. 7

24 II. State of knowledge. Further examples for stabile, unassociated low valent E I -compounds in the solid state were synthesized with the extremely bulky terphenyl ligands in the compounds EMes* (E = In or Tl; Mes* = 2,6-Bis(2,4,6-triisopropylphenyl)phenyl) (Tolman cone angle ca. 218 ), which sparked interesting discussions in the late 1990 s. [34-36, ] The group of P. Power was able to isolate dimeric complexes of the type ArGaGaAr (Ar = bulky substituent), which exhibit Ga-Ga bond distances of 2.32 Å to 2.62 Å. [103] This Ga-Ga interaction was found to be rather weak. elated dianionic derivatives of these dimeric compounds were reported by G. obinson in 1997, who described the Ga-Ga interactions as Ga-Ga triple bond on the basis of simple theoretical bonding considerations. [101] Consequently, an intensive debate about the Ga-Ga bond order in such dimers appeared in the literature. However, the related monomeric complex GaMes* was also isolated and characterised by Power et al. showing, that onecoordinated monomers with only two electron pairs in the metal valence shell can be isolated at ambienbt temperature also for Ga. [103] Table 2 Comparison of structural features of the low valent Ga(I) compounds. Compound Tolman angle Ga-X bond distance Ga M calculated according to [104] (X =, C) in Å (if available) GaCp* [21, 22] ca. 112 a) av GaTp tbut[28, 105] ca. 183 a) av n.a. GaMes* [35] ca. 218 a) (3) d) n.a. Ga(DDP) [31, 103] ca. 147 a) n.a. av ca. 180 c) n.a. Ga{(Ar)CH} 2 -[39, 106]a) ca. 173 a) av n.a. Ga(Giso) [38]b) ca. 160 b) av n.a. The Tolman-cone angles are calculated from structural data taken from the compounds a) [(CO) 4 Fe-E); b) [Pt{Ga(Giso)} 3 ]; c) [(1,3-cod)Pt{Ga(DDP)} 2 ]. d) The Ga-C bond length is taken from the dimer ArGaGaAr (Ar = C 6 H 3-2,6-Dipp 2 ) [103] ; The neutral six membered heterocycles E I DDP (E = Al, Ga, In, Tl; DDP = 2-{(2,6- diisopropyl-phenyl)amino}-4-{(2,6-diisopropylphenyl)imino}-2-pentene) represents another class of neutral, low-valent and, due to the increased steric demand compared to Cp* (Tolman cone angle of 147 ), monomeric group 13 compounds (Chart 4 and Table 2). In 2000, Power and oesky synthesized the gallium- and aluminium derivatives by either salt metathesis of Li(DDP) with metastable GaI [31] or by reduction of the respective diiodide [(DDP)AlI 2 ] with potassium metal. [30] Whereas Al(DDP) and Ga(DDP) are stable at room temperature in an inert gas atmosphere for several weeks, the corresponding heavier group 13 analogues 8

25 II. State of knowledge. In(DDP) and Tl(DDP) are photolabile and also decompose in aromatic solvents even in the [32, 52] absence of light, as reported by Hill et al. in 2004 and In 1999 and 2001, Schmidbaur et al. reported the synthesis of an anionic gallium heterocycle [Ga{( t Bu)CH} 2 ] -, which is isoelectronic to the popular -heterocyclic carbenes (HC) and is obtained in an uncoordinated state (i.e. there is no Ga-K coordination) as its potassium salt, [:Ga{( t Bu)CH} 2 }][K(18-crown-6)(THF) 2 ], and as a dimeric potassium complex, [Ga{( t Bu)CH} 2] ][K. tmeda]} 2 (Chart 4). [40, 107] In 2002, a facile and more convenient route for the synthesis of the anionic HC analogue [Ga{(Ar)C(H)} 2 ] - (Ar = 2,6-Pr i 2C 6 H 3 ) was reported by Jones et al. [39] It is carried out via reduction of the related diiodide with potassium metal. otably, the first example of a spectroscopically characterised monovalent boryl compound [B{ArCH} 2 ]Li was prepared in 2006 by Segawa et al. [108], which was found to behave as base or nucleophile on reaction with electrophiles. Ar E Ga: Ar K K Ar :Ga Ar Ga: Ar Ar E(DDP) E = Al, Ga, In, Tl Ar = 2,6- i Pr 2 (C 6 H 3 ) Chart 4: E(DDP) derivatives and the 5-membered heterocycle [Ga{(Ar)C(H)} 2 ] -. Very recently, the first examples of four-membered group 13 heterocycles were reported by Jones et al. in Thus, the reaction of the bulky guanidinate Li(Giso) (Giso = [(Ar)C(Cy 2 )(Ar)]; Ar = C 6 H 3 Pr i 2-2,6; Cy = cyclohexyl) with GaI, InCl or TlBr leads to the formation of the guanidinate complexes Ga(Giso), In(Giso) and Tl(Giso), respectively. [38] Whereas the Ga- and In compound show the expected four-membered heterocycle, the Tlanalogue is monomeric but contains a weak Tl-arene interaction. [38] 9

26 II. State of knowledge. Cy 2 Ga [B(Ar F ) 4 ] - E E(Giso) E = Ga, In, Tl Ga Chart 5: The 4-membered heterocycle E(Giso) and the cationic [Ga 2 Cp*] + [B(Ar F ) 4 ]. Finally, soluble sources of Ga + and In + ions featuring weakly coordinating counterions (e.g. [B(Ar F ) 4 ] (Ar F [109, 110] = C 6 H 3 (CF 3 ) 2-3,5)) have been reported very recently by Cowley, MacDonald [111] and ourselves. [112] Thus, the controlled protolysis of GaCp* with [H(Et 2 O) 2 ][B(Ar F ) 4 ] yields [Ga 2 Cp*] + [B(Ar F ) 4 ], [112] whereas [In][OTf] is formed from InCl and HOtf (Otf = HO 3 SCF 3 ) in high yields. [113] 10

27 II. State of knowledge. Bonding situation in E I -compounds. Following the precedent set with HCs, the coordination chemistry of the E I species was persued. The ligand properties of E I can be understood by closely examining the frontier orbitals of the low-valent E I compounds and their interactions with metal orbitals of suitable symmetry. Whereas the group 13 organyls E I Cp* and E I alkyl have been studied for more than 15 years and several theoretical analyses concerning their bonding situation in TM-E I complexes have been performed, only recently have the related five- and six-membered heterocyclesbeen intensely investigated. The monomeric E I molecules exhibit a free electron pair at the group 13 atom, which is located in a σ-type orbital (HOMO) and thus of directional character. The LUMO, in the case of Al 10.5 kcal/mol higher in energy (Table 4), consists of two unoccupied p-orbitals perpendicular orientated to the E-C bond-axis, so a singulet-ground state is assumed with significant singlet-triplet splitting energies (Table 3). [111] Therefore, the frontier orbitals of the E I compounds are similar to CO, hence E I can be regarded as isolobal to phosphines, CO [83, 111, 114] or HCs. The neutral six-membered group 13 metal(i) heterocycles [E{[( 1 )C( 2 )] 2 CH}] (E =B, Al, Ga, In or Tl; 1 =H, Me or C 6 H 3 Pr i 2-2,6 (Ar); 2 =H or Me) have also been the subject of several theoretical studies. [30, 52, 53] Whereas no systematic study about the bonding situation of these heterocycles in TM-complexes is available so far, the bonding nature in the heterocycles themselves was elucidated using a number of different computational techniques. [53] Thus, the free electron pair at the group 13 metal was described to occupy an sp-type orbital of the metal E, which lies in the plane of the metallacycle, in a pseudo-trigonal planar fashion with respect to the -E linkages. [30, 52] Additionally, the vacant p-π orbitals of the group 13 metal E are not represented by the LUMO, which is entirely ligand-based and of π-symmetry, but by the LUMO+1. [39, 52] Furthermore, the HOMO-LUMO gap for {E{[PhCMe] 2 CH}} (Table 4) is relatively high and increases from B to In, suggesting that these compounds are good σ-donors but weak π-acceptors. Population analysis indicates, that when E = Al In, the group 13 metal carries a partial positive charge and there is a substantial ionic character on the -E bonds. For E = B, there is a partial negative charge on boron and the B- bonds are more covalent in character. [53] 11

28 II. State of knowledge. Table 3: Singlet-Triplet splitting (kcal/mol) for different E I -compounds. Compound Cp* [111] {(MeCMe) 2 } [115] [53, 116] {(PhCMe) 2 CH} B Al Ga In The analysis of the singlet-triplet splitting energies (Table 3) reveals that the compounds [E I {( 1 )C(Me)] 2 CH}] (E = Al, Ga, In; 1 = Ph, Me) [53, 116] have a singulet ground state, whereas for E = B the triplet state is much closer in energy to the singlet state. These results, together with electron localisation function (ELF) calculations, lead the authors to conclude, that in the case of E = B, a diradical species with a B(II) center is the best description, whereas for E = Al - In the bonding is probably best viewed as a donor-acceptor Lewis structure involving a positively charged monovalent E + -ion coordinated by the bidentate monoanionic [( 1 )C(Me)] 2 CH] - (Chart 6). This finding contrasts the bonding situation in Arduengo-carbenes, where a significant delocalisation over the heterocycle is observed. [ ] evertheless, an analogy to Arduengocarbenes can be maden, because of the presence of a non-bonded electron pair with lone pair character at the metal center. These results imply that the boron system [B(DDP)] is by far too reactive to be isolated under normal conditions, and indeed, no six-membered heterocyle compound of boron has been synthesized thus far. Ga : C : Chart 6: Isolobal analogy between Ga(DDP) and the -heteroleptic Carbenes (HC s). Density functional theory (DFT) and natural bond orbital (BO) calculations were also performed on the five-membered heterocycles [E{(H)C(H)} 2 ], (E =B, Al, Ga and In). [115] As a result, the previously described partial positive charge at the group 13 metal center (E = Al-In) is also found in these compounds suggesting again, that the structure of the heterocycle 12

29 II. State of knowledge. is best formulated as a donor-acceptor interaction between E + (E = Al-In) and a chelating diamido unit, similar to E(DDP) (Chart 6). As a result of the electronegativity difference of and E, only little delocalisation over the E fragment is observed, which is in strong contrast to the significant delocalisation over the C fragment of HC s. [ ] Table 4: HOMO-LUMO Gap (kcal/mol) for different compounds. Compound E I ECp* [20] E[{(Ph)} 2 C(Me 2 )] [38] E{(PhCMe) 2 CH} [116] B n.a. n.a Al Ga n.a In n.a Similar results are also reported for the four-membered heterocycle [E({(Ph)} 2 C(Me 2 ))] (E = Al, Ga, In), [38] showing a lone pair of electrons at the group 13 metal with directional character. The metal s lone pair and p z orbital (orthogonal to the heterocycle plane) are associated with the HOMO and the LUMO, showing significant HOMO-LUMO gaps (Table 4), although being considerably less than those calculated for the six-membered heterocycles. However, a high ionic character of the -E bonds is observed, indicating a bond situation in the four-membered heterocycle comparable to the five- or six-membered heterocycles, respectively. 13

30 II. State of knowledge. Bonding situation in TM-E I -complexes. σ-donor properties. In 2000, Frenking et al. reported on a comprehensive theoretical study of the bonding character of complexes of the type [L n M-E] and [M(E) 4 ], with E being B-Tl and = Cp and related ligands. [49, 85] The use of different organic rests ( = Cp, Ph, (SiH 3 ) 2, Me) and different transition metals (M = Fe, W, i, Pd, Pt) confirms the generally accepted bonding model for the TM-E I ( = Cp, alkyl) bond, which is shown in Chart 7. E s M p a) E M p' E M b) E M q(+) E q(-) M Chart 7: a) schematic draw of the M-E I orbital interaction. b) schematic draw of the electrostatic interactions of the local charge concentration at the donor atom E and the acceptor atom M. E bears a partially positive, M a partially negative charge. The ability of these E I compounds ( = Cp, alkyl or {[(Ar)C(Me)] 2 CH} - ) to act as Lewisbases was examined in more detail in both preparativ as well as theoretical studies. Thus, the Lewis-acid-base adducts [Cp*E EPh F 3] (E = B, Al) were synthesized by the group [120, 121] of Cowley et al., whereas the related GaCp* adducts with coordinatively and electronically unsaturated group 13 fragments B(C 6 F 5 ) 3, Ga( t Bu) 3 and Cp*GaX 2 (X = Cl, I) are prepared and analyzed by Jutzi et al. (Figure 1). [122] 14

31 II. State of knowledge. Figure 1: Molecular structures of [GaCp*B(C 6 F 5 ) 3 ] and [(DDP)AlB(C 6 F 5 ) 3 ]. Whereas for the four- and five-membered heterocycles no such adducts are reported so far, the related Ga-adduct of the six-membered heterocycle [(DDP)Ga B(C 6 F 5 ) 3 ] was synthesized by Power et al. in [123] One approach that has been proposed for the determination of Lewis basicity is to measure the extent of departure from trigonal planarity of the BC 3 skeleton of B(C 6 F 5 ) 3 when the donor acceptor bond is formed, i.e. the sum of the three C B C bond angles. [124] On the basis of the sum of the C B C bond angles ((DDP)Ga (333.5 ); AlCp* (339.8 ) and GaCp* (342,2 )), Cowley et al. determined the order of Lewis basicity to be Ga(DDP) > AlCp* > GaCp*. [120] Additionally, as shown by the group of oesky in 2005, [125] the six-membered aluminiumheterocycle Al(DDP) is different from it s Ga counterpart. Ab-initio calculations with analysis of the Lapacian of the electron density for Al(DDP) revealed that the Al-compound has both a Lewis-acid and Lewis-base character, with the free electron pair accounting for its Lewisbasicity and a charge depletion in the semiplane of the six-membered ring close to the Alatom prooviding its Lewis-acid properties. Accordingly, upon mixing Al(DDP) with B(C 6 F 5 ) 3, not only a single adduct formation between the Al and the B center takes place. Furthermore, a weak, but significant interaction of a F-atom of the C 6 F 5 -group to the Alcenter is observed (Figure 1), which is consequently reflected in the crystal structure by longer C-F bonds for the coordinated F-atom as well as by ab-initio calculations, which quoted the electron density between the Al- and carbon centers with a (Al-F)/F-C) ratio of /

32 II. State of knowledge. Table 5: Changes in the charges of Δq σ (E) in TM-E I complexes. Compound B Al Ga In Tl (CO) 4 Fe-ECp [85] (CO) 4 Fe-EPh [85] [85] i(eme 3 ) [85] Pd(EMe) [85] Pt(EMe) [GaPt(GaCp) 4 ] +[126] a [GaPt(GaCp) 3 ] +[126] a [EPt(PMe) 3 ] +[126] a a a) calculated charge in the complex for the E + -ligand, not Δq A detailed computational investigation (natural bond orbital analysis and charge decomposition analysis) of the charge distribution of the E I fragment in TM-E I complexes (with being Cp, alkyl) was performed by the groups of Frenking [85] and Cowley et. al. [111]. The calculated change in the σ-charges at the element E (Table 5) suggests, that the E I fragments provide a substantial σ-donation to the transition metal in such complexes, thus being strong σ-donors. In accordance to that, the TM-E I bond dissociation energies are very high (Table 6), even if is a poor π-donor, indicating very strong TM-E I bonds, which follow the order B > Al > Ga In > Tl. Table 6: Dissociation energy (D 0 ) of the Fe-E bond (kcal/mol) in [(CO) 4 Fe-E I ] complexes. Compound B Al Ga In Tl (CO) 4 Fe-ECp [85] (CO) 4 Fe-EPh [85] [85] i(eme 3 ) [85] PdEMe) [85] Pt(EMe) What happens if the group 13 metal center does not bear any additional ligands? With the synthesis of [Ga 2 Cp*] + [B(Ar F ) 4 ] and [In][OTf] [112, 126] (vide supra), such naked E+ ions are now readily accessible and thus can be studied in more detail. [126] In this respect, the reaction of one of these naked gallium cations, [Ga 2 Cp*] + [B(Ar F ) 4 ] (Ar F = C 6 H 3 (CF 3 ) 2-3,5), with the electron rich d 10 platinum(0) complex [Pt(GaCp*) 4 ] was performed and yields the novel 16

33 II. State of knowledge. species [GaPt(GaCp*) 4 ] + [B(Ar F ) 4 ]. [126] The results of population analysis on the model compound [GaPt(GaCp) 4 ] +[126, 127] show, that the total occupation of the σ-orbitals (valence s and pz) at the group 13 metal in the model complexes [GaPt(GaCp) 4 ] + and [GaPt(GaCp) 3 ] + is significantly greater than 2 (2.49 and 2.28, respectively). This, taken together with the small calculated positive charge at the Ga ligand (Table 5) and the significantly higher positive charge at the platinum center, suggests that naked Ga + is thus a poor σ-donor in such complexes. This is in well agreement with the lone pair of the Ga + ligand being almost entirely of s-charakter (in contrast to its GaCp* analogue, which shows a directional lone pair, vide supra). In summary, the ability of the low valent group 13 organyls to act as σ-donor ligands depends on the group : The TM-E σ-donation is clearly larger for being a strong π-donor (like Ph), wheras smaller σ-donation is observed for weak π-donors (e.g. Me). Thus, the sixmembered ring Ga(DDP) is suggested to be a stronger σ-donor (on the basis of structural comparisons) compared to its Cp* analogues GaCp* and AlCp*, whereas the naked Ga + is a poor σ-donor. π-acceptor properties. In contrast to the σ-donor capabilities, the TM E π-backbonding and therefore the π- acceptor properties of the group 13 organyls in TM-E I complexes were not clear and have been a controversial issue in the following years. The synthesis of the iron gallium diyl complex, [(Ar*Ga)Fe(CO)4] (Ar* =C 6 H 3 (C 6 H 2 Pr i [35, 101] 3-2,4,6) 2-2,6) (Scheme 2), which shows an unassociated two-coordinated gallium atom and a linear Fe-Ga-C fragment with a very short Ga-Fe bond of Å initiated a discussion regarding wether this compound is a triply-bonded ferrogallyne or simple dative-covalent interaction began to emerge. umerous theoretical studies on this and related complexes have been carried out since then. OC CO (a/k)[fe(co) 4 ] [Mes*GaCl 2 ] Et 2 O OC Fe Ga CO Scheme 2: Synthesis of [(CO) 4 FeGaMes*]. 17

34 II. State of knowledge. Thus, a detailed investigation by the group of Frenking et al. [49, 85] shows, that the covalent contribution of the attractive term of the TM-E bond in [(CO) 4 Fe-E I ] complexes and the ionic contributions have similar strength (Table 7). Similar contributions of 55-57% were found in the complexes bearing a naked Ga + such as [GaPt(GaCp) n ] + (n = 3, 4). [126] In contrast to that, the Fe-EMe bonds in the homoleptic complexes [Fe(EMe) 4 ] have a much larger ionic contribution (app. 65 %). Table 7: covalent contribution (%) of the attractive terms of the TM-E bonds in TM-E I complexes. a) Compound B Al Ga In Tl (CO) 4 Fe-ECp [49] (CO) 4 Fe-EPh [49] (CO) 4 Fe-EMe [49] [49] i(eme 3 ) [(CO) 5 Fe] [49] 48.3 [GaPt(GaCp) 4 ] +[126] 57.2 [GaPt(GaCp) 3 ] +[126] 55.5 [EPt(PMe) 3 ] +[126] a) The values give the percentage contribution to the total attractive interactions reflecting the covalent character of the bond. However, the π-contribution of the covalent part of the TM-E bonds (Table 8) is comparably small (10 20%) for E I in the complexes [(CO) 4 Fe-E I ] ( = Cp, Ph, Me) and decrease upon descending group 13 (i.e. B > Al > Ga > In > Tl). The same is true for the fivemembered heterocycle [Ga{(Ar)C 2 H 2 (Ar)}] -, which shows an increase of the π- contribution (18 to 27%) in the complexes [CpM(CO) n Ga{(Ar)C 2 H 2 (Ar)}] (M = V, Mn, Co, n = 4, 3, 2), when going from V to Co (Table 8). [64] This increase is expected, when moving to more electron rich late transition metal systems wich contain fewer competing π- acidic carbonyl ligands. Higher π-contribution are calculated for the homoleptic complex [i(eme) 4 ], [49] the four-membered heterocycle [Ga(Giso)] in [Pt{Ga(Giso)}], [60] as well as for the complexes exhibiting a naked Ga + (Table 8). The result of the calculated π-back-donation is summarized in Table 9. In the [(CO) 4 Fe-E I ] compounds, the π-back-donation follows the sequence B > Al > Ga > In > Tl, but is also depending on the rests of the E I fragment. Hence, in the case of alkyl-substituted group 13 metals such as [E I Ph], the π-acceptor properties of the compound is higher, due to missing π- 18

35 II. State of knowledge. donor properties of the substituent, compared to the Cp*-analogue, for which the vacant p- orbitals of the E-center are partially populated by the π-donor orbitals of the Cp* groups. [49] Table 8: π-contribution (%) of the covalent part of the M-E bond in TM-E I complexes. a) Compound B Al Ga In Tl (CO) 4 Fe-ECp [49] (CO) 4 Fe-EPh [49] (CO) 4 Fe-EMe [49] [CpV(CO) 4 - Ga{(Ar)C 2 H 2 (Ar)}] [64] 13 [CpMn(CO) 3 - Ga{(Ar)C 2 H 2 (Ar)}] [64] 18 [CpCo(CO) 2 - Ga{(Ar)C 2 H 2 (Ar)}] [64] 27 [49] i(eme 3 ) [60] Pt[GaGiso) [GaPt(GaCp) 4 ] +[126] 39.2 [GaPt(GaCp) 3 ] +[126] 49.0 [EPt(PMe) 3 ] +[126] [(CO) 5 Fe] [49] 47.9 a) The values give the percentage π-contribution to the total orbital interactions. These results show, that the M E I σ-donation is clearly larger than the M E I π-backdonation when is a strong σ-donor and that the degree of metal metal backbonding is minimal due to the relatively high energy of the group 13 metal p-orbitals. However, for the homoleptic complexes [M{EMe} 4 ], a significant π-back-donation is observed, which in some complexes is even larger than the σ-donation. Finally, the fourmembered heterocycle [Ga(Giso)] also shows comparably high π-back-donation (0.29) in the complex [Pt{Ga(Giso)}]. Additionally, the aforementioned population analysis of naked Ga + in the model complexes [GaPt(GaCp) 4 ] + and [GaPt(GaCp) 3 ] + (vide supra) together with the lack of a gallium-bound substituent with π-donor orbitals means, that in contrast to GaCp*, for example, Ga + is a good π-acceptor. 19

36 II. State of knowledge. Table 9: charge σ-donation (d) and π-back-donation (b) in TM-E I complexes. Compound B Al Ga In Tl (CO) 4 Fe-ECp [85] d b (CO) 4 Fe-Eph [85] d b [85] i(eme) 4 [85] PdEMe) 4 [85] Pt(EMe) 4 [60] Pt[GaGiso) 3 d b d b d b d 0.43 b 0.29 In summary, the fundamental σ-donor and π-acceptor properties of these ligands are strongly dependent on the nature of the diyl substituent : Thus, Cp*Ga is known to be a very good σ- donor ligand but to function as a weak π-acceptor towards transition-metal centers owing to strongly competing π-donation from the Cp* fragment. Alkyl substituents offer less competition in this regard, and hence a greater π-component to the M-E bond is found in metal complexes containing, for example, the {MeGa} ligand. Finally, the naked gallium or indium cations featuring no pendant diyl substituent are expected to display very strong π- acceptor properties. [112] Coulombic interactions. Another result of the studies mentioned above is the fact that, besides covalent interactions, Coulomb-interactions must be also considered (Table 7). Therefore, the TM-E bonds are mainly caused by charge attraction between the negatively charged TM atoms and the positively charged group 13 atoms, due to the different electronegativity of the metals M and E. Therefore, the M-E bond can be best described as ionic in nature (M δ+ -E δ-, Chart 7), with covalent contributions being comparably low (covalent bond order < 1) and less important than Coulomb interactions. As such, the aforementioned discussion of the Fe-Ga bond in [(Ar*Ga)Fe(CO) 4 ] being a single or triple bond is in fact meaningless. 20

37 II. State of knowledge. Synthetic methods for the preparation of M-E I compounds. Presently, the stability and facile synthetic accessibility of low-valent group 13 compounds facilitates access to their transition metal complexes. Many synthetic strategies for the preparation of M-E complexes have been established in recent years. [43, ] The formation of M-E I complexes mainly relies on two different reaction types: The substitution of labile ligands by E I is the most general route in this respect. A variety of homoleptic as well as heteroleptic complexes [M a (E I ) b L c ] were synthesized from different transition metal complexes such as carbonyl, olefin or phosphane complexes. However, the possibility of the E I fragments to act as potential reducing agent towards higher oxidation state transition metal complexes [ML n X m ] is another, yet less common synthethic pathway to M-E species. In this method, insertion products can be obtained if the oxidised group 13 moiety remains coordinated at the transition metal center, or low oxidation state transition metal complexes [M a (E I ) b ] are accessible by cleavage of the oxidised [X 2 E] species. Whereas the coordination chemistry of the group 13 organyls E I Cp* and related compounds are relatively well explored, the coordination chemistry of the neutral four-, and sixmembered, as well as those of anionic five-membered heterocycles towards transition metals are relatively unexplored. Therefore, a brief overview over the efforts concerning the coordination chemistry of the E I Cp* and related E I {C(SiMe 3 ) 3 } ligands is presented first followed by the respective results of the various heterocycles. 21

38 II. State of knowledge. Coordination chemistry of E I Cp* and E I C(SiMe 3 ) 3. Substitution of labile ligands. As mentioned previously, the E I fragments are isolobal to CO or phosphines. Accordingly, the substitution of CO in homoleptic transition metal carbonyl complexes has been studied in detail. However, the possibility of substitution of π-acidic CO ligands in such complexes bears an intrinsic problem. When substituting a CO molecule with a stronger σ-donor (and weaker π-acidic) ligand like E I Cp*, the electron density at the transition metal and thus the back-donation of the d-metal center to the remaining carbonyl ligands is increased. Therefore, the M-CO-bonds of the remaining carbonyls become stronger and substitution of further CO is hindered. This was demonstrated by the synthesis of the complexes [M 2 (CO) x (μ 2 -E I ) 2 ] (M = Co, Mn, Fe; x = 6, 8; E I = AlCp*, GaCp*, GaC(SiMe 3 ) 3, InC(SiMe 3 ) 3 ) via treatment of the homoleptic carbonyl compounds with E I or the substitution of chelating olefin ligands (e.g. cyclooctene, norbornadiene, cyclooctatetraene or heptatriene). [22, ] Substitution reactions involving salt metathesis of carbonylmetallates and E III X 2 (Scheme 3), afforded the [22, ] compounds [M(CO) a ECp* b ] (M = Fe, Cr, Mo ;a = 4, 5; b = 1, 2; E = Ga, Al, In). CO OC M 2 ECp* + M 2 (CO) 8 E - 2 CO M E = Al, Ga M = Co, Mn, Fe OC CO CO E CO Scheme 3: Syntesis of [M 2 (CO) 6 (μ 2 -E I Cp*) 2 ] (M = Co, Mn, Fe). Furthermore, some examples of substitution reactions with larger carbonyl-clusters are reported. As seen in the cluster complex [i 4 (μ 2 -GaCp*) 2 (μ 3 -GaCp*) 2 (CO) 6 ] [22] or the series [h 6 (μ 3 -CO) 4-x (μ 3 [138, 139] -GaCp*) x (CO) 12 ] (x = 1-4) synthesized by Jutzi et. al., the E I fragment is μ 3 -bridging in such complexes. This structural feature is known for carbonyls in several CO clusters. 22

39 II. State of knowledge. Another approach to afford transition metal-e I complexes is their addition to coordinatively unsaturated metal centers. Via thermally activated reductive elimination of neopentane from [(dcpe)pt(h)(ch t 2 Bu)] (dcpe = bis(cyclohexyl)phosphinoethane), the 14 VE-complex [(dcpe)pt] is formed in-situ and trapped with E I Cp*. Thus, the [Pt(P 3 )] derivatives [(dcpe)pt(e I ) 2 ] (E = Al, Ga, In, = Cp*; E = Ga. = C(SiMe 3 ) 3 ) can be obtained quantitatively. [87, 140] An unsaturated carbonyl system which was studied in more detail, is the metal-metal triply bonded [CpM(CO) 2 ] 2 (M = Mo, W), which reacts with GaCp* yielding the products [Cp(CO) 2 M(μ-GaCp*)] 2 (Scheme 4). Here, the Ga-Cp* moieties exhibit a η 1 - coordination of the Cp* ring to the bridging Ga center. It is suggested, that this is due to the steric overcrowding in the molecule. [141] OC OC M M CO CO 2 GaCp* OC OC M Ga Ga M CO CO M = Mo, W Scheme 4: Insertion of GaCp* into the M M triple bond of Mo and W. Significant effort has been paid to the synthesis of homoleptic compounds of the type [M a (E I ) b ], with b > a > 1. Whereas the constitution and structural characteristics of the previously discussed complexes and clusters are mainly derived from those of classical carbonyl complexes, the new homoleptic series [M a (E I ) b ] (M = Pd, Pt; b > a >1) has no direct structural analogues in classical phosphine or carbonyl cluster chemistry, although recently the coordination chemistry of P 3 has been extended to include a bridging mode. [142] These complexes exhibiting bridging and terminal E I ligands are still rare in comparison to other M-E complexes. Thus, the ability of E I compounds to effectively bridge two or three transition metal in a η 2 - or η 3 -fashion together with its strong σ-donor capabilities make the coordination chemistry of these ligands a quite novel and interesting field to explore. 23

40 II. State of knowledge. Because of the aforementioned problems associated with the complete substitution of CO ligands in carbonyl-complexes and clusters, several reports concerning the substitution of more-labile ligands in transition metal-olefin or phosphine complexes have been published in the last decade. Thus, the coordinated cod-ligands (cod = 1,5-cyclooctadiene) in [i(cod) 2 ] and [Pt(cod) 2 ] can be readily replaced by ECp* or EC(SiMe 3 ) 3 yielding the monomeric, homoleptic complexes [M(E I [83, 114, 136] ) 4 ] (M = i, Pt; E = Al, Ga, In; = Cp*, C(SiMe 3 ) 3 ). The related Pd compounds can be obtained by the reaction of [Pd(tmeda)(CH 3 ) 2 ] (tmeda =,,, -tetramethylethylenediamine) with ECp*, which proceeds via methyl-migration, liberation of the Cp*EMe 2 moiety and reduction of the Pd-center by ECp*. [54] The aforementioned complexes are found to be rather inert and undergo no ligand substitution by either CO, P 3 or ECp*. The synthesis of the compounds [M a (E I ) b ] with a 2 is accomplished in two different ways, building block synthesis and direct methods. The building block synthesis of [M 2 (GaCp*) 5 ] (M = Pd, Pt) [54] involves the treatment of [M(GaCp*) 4 ] with [Pt(cod) 2 ] and subsequent addition of GaCp*. In [M 2 (GaCp*) 5 ], terminal as well as bridging GaCp* moieties are observed. It was shown by 1 H-M spectroscopy, that the formation of [M 2 (GaCp*) 5 ] proceeds via the reactive, but isolable intermediates [PtM(GaCp*)(μ 2 -GaCp*) 3 (μ 2 -cod)] (Scheme 5). excess GaCp* *CpGa GaCp* M GaCp* GaCp* [Pt(cod) 2 ] *CpGa M Cp* Ga Ga Cp* Pt Ga Cp* GaCp* *CpGa M Ga Cp* Cp* Ga Pt Ga Cp* GaCp* M = Pd, Pt Scheme 5: Building block synthesis of [MPt{GaCp*} 5 ]. 24

41 II. State of knowledge. The direct synthesis of the complexes [M a (E I ) b ] is influenced by kinetic factors such as the reaction temperature, the nature of the transition metal-olefin complex or simply the solubilities of the starting materials. In 2000, [Pt 2 (GaCp*) 5 ] was synthesized by the reaction of [Pt(C 2 H 4 ) 3 ] with an excess of GaCp*. [143] The related Pd compound is obtained, when [Pd 2 (dvds) 3 ] (dvds = 1,3-divinyl-1,1,3,3-tetramethyldisiloxane) is treated with GaCp* in hexane at -30 C. However, by performing this reaction in toluene at room temperature, the trinuclear cluster [Pd 3 (GaCp*) 8 ] is formed in high yields with only small amounts of the dinuclear species present in the reaction mixture. [144] Figure 2: Molecular structure of [Pd 3 {GaCp*} 8 ]. So, on varying the reaction conditions, the series [M a (ECp * ) b ] (M = Pd, Pt, a = 2, 3; E = Ga, In; b = 5, 8) is accessible and the reactivity of these complexes was examined in more detail. Thus, the unsaturated complexes [M a (GaCp*) b ] (M = Pd, Pt, b > a > 1) react with several ligands like AlCp*, CO, phosphanes and isocyanides giving new di- and trimeric substitution products. A detailed review has been given in 2004 by our group. [45] 25

42 II. State of knowledge. Cp* In PPh 3 3 PPh 3 Ph 3 P Pd *CpIn Pd In Cp* Pd PPh 3 Pd 3 (InCp*) 8 2 dppe Ph 2 P P Pd Ph 2 *CpIn Cp* In Pd In Cp* Ph 2 P Pd P Ph 2 Scheme 6: Substitution reactions at [Pd 3 {InCp*} 8 ]. Only one interesting example will be presented herein. The reaction of the linear, trinuclear complex [Pd 3 (InCp*) 8 ] with PPh 3 or dppe (dppe = bis(diphenylphosphanylethane) lead to the rearrangement of the former linear Pd 3 cluster core giving the triangular Pd 3 complexes [Pd 3 (InCp*) 3 (PPh 3 ) 3 ] and [Pd 3 (InCp*) 3 (dppe) 2 ], with two InCp* ligands capping both faces of the Pd 3 triangle resulting in a trigonal bipyramidal geometry (Scheme 6). [55] The third InCp* is bonded in a μ 2 -coordination mode parallel to the Pd 3 plane. Due to an unexpected splitting of the Cp* resonances in the solution 1 H-M spectrum of the trimeric compound [Pd 3 (GaCp*) 8 ], which is not in agreement with the linear-state solid state structure, it has been suggested that in solution, a rearrangement of the GaCp* ligands in [Pd 3 (GaCp*) 8 ] occurs and a triangular arrangement of the type [Pd 3 (μ 3 -GaCp*) 2 (μ 2 -GaCp*) 3 (GaCp*) 3 ] is formed, which is similar to the In-phosphine analogue described above. 26

43 II. State of knowledge. Insertion reactions of ECp* into metal-halide bonds. Insertion reactions of low valent group 13 compounds into metal-metal and metal-halide bonds are known and date back to the 1970 s, where Hsieh published several papers on the insertion of indium(i) halides into metal-metal and metal-halogen bonds of mainly carbonyl containing transition metal centers. Thus, facile insertion of InX (X = Cl, Br) into M-X bonds (M = W, Mo, Mn, Fe, Hg, Co, h) takes place readily and results in microcrystalline compounds featuring a M-InCl 2 moiety. [128, ] However, most of these compounds are found to be poorly soluble and therefore a full structural characterisation was impossible. Also, as seen in recent studies, these reactions are much more complicated than previously thought. Thus, the system [Ph 3 PAuCl]/InCl forms the salt [(Ph 3 P) 2 Au][InCl 4 ]. Only the addition of the chelating phosphine ligand dppe (dppe = bis(diphenylphosphanylethane) allows the isolation of the Au 3 In 3 cluster [(dppe) 2 Au 3 In 3 Cl 6 (THF) 6 ]. [148] This result shows, that insertion reactions of E I X are extremely sensitive to the reaction conditions and many reaction pathways must be considered. The stability and accessibility of the room-temperature stable group 13 organyls E I Cp* and E I C(SiMe 3 ) 3 opens the door for a detailed examination of the structural properties, as well as the driving force of such insertion reactions. In 1994, Schnöckel et al. reported on the reaction of [Cp 2 i] with AlCp*. A formal insertion of a AlCp*-moiety into a icp-bond gave the butterfly-complex [(Cpi) 2 (μ 2 -AlCp*) 2 ], the first example of an all-hydrcarbon M-E I complex. [149] Figure 3: Molecular structure of [(Cpi) 2 (μ 2 -AlCp*) 2 ]. 27

44 II. State of knowledge. Whereas this compound is a structural analogue to the carbonyl-complex [(Cpi) 2 (μ 2 -CO) 2 ], Jutzi et al. reported in 2000 on the insertion of GaCp* into the Fe-Cl bonds of [FeCl 2 (THF) 2 ] giving [Cp*Fe(GaCp*) 2 (GaCl. 2 THF)]. [150] In this reaction, first an insertion into the Fe-Cl bond takes place, followed by a Cp*/Cl exchange and σ/π-rearrangement of the Cp*-ring. Similarly, the solvent-free piano-stool complexes [Cp*Fe(GaCp*)(Ph 3 P)(GaBr 2 )] and [Cp*Fe(GaCp*) 2 (GaBr 2 )] are prepared from [FeBr 2 (L) 2 ] (L = Ph 3 P or CH 3 C) and GaCp*. Additionally, the reaction of AlCp* with [FeBr 2 (PPh 3 ) 2 ] lead to the formation of the orthometallated complex [FeCp*(μ 3 -H)(η 2 -(C 6 H 4 )PPh 2 )(AlCp*)-(AlBr 2 )], [59] perhaps as a resul of the formation of the more acidic AlBr 2 - compared to a GaBr 2 -moiety in the course of the reaction. Insertion reactions into ruthenium- and rhodium-halogen bonds have been extensively studied by our group. [45, 58, 151, 152] Depending on stoichiometry and reaction conditions, the use of the isolable d 6 complexes [{LMCl 2 } 2 ] (M = h, L = Cp*; M = u, L = p-cymene) lead to a variety of insertion products, as seen in Scheme 7. Thus, the monomeric insertion products [LM(ECp*) 2 (ECp*Cl 2 ] (M = h, L = Cp*, E = Ga, In; M = u, L = p-cymene) as well as the Lewis acid-base adducts [LM(GaCp*) 2 (GaCl 3 )] (M = h, L = Cp*; M = u, L = p-cymene) can be synthesized by tuning the reaction conditions and the stoichiometric ratio of the educts. The former shows cage-like motifes with E-Cl-E bridges in the solid state. Generally, one of the chlorine atoms is found in a bridging position, whereas the second one is terminal. otably, the Cp*-ring is either σ-bonded (to a ECl 2 -fragment), η 5 -bonded (ECp* without Clinteraction) or nonclassical bonded as η 2 or η 3 (on the ECl-fragment), suggesting a flexible coordination of the Cp* ring to the group 13 metal. Also, these compounds can be regarded as frozen intermediates of a chlorine-exchange between two ECp* fragments, which is supported by 1 H-M studies. Another interesting result is obtained by treating [{LMCl 2 } 2 ] with only one equivalent GaCp* in boiling toluene. Here, reduction of the metal centers occurs, giving the dimeric complexes [{LM(μ 2 -Cl)} 2 ] (M = h, L = Cp*; M = u, L = p-cymene), respectively. Whereas the hcompound is also accessible by the reaction of the precursor with a/hg or elemental Ga, the u-compound can only be obtained by this route. Aditionally, [{(p-cymene)u(μ 2 -Cl)} 2 ] is suggested to be an intermediate in the formation of the compound [{(p-cymene)u} 2 (GaCp*) 4 Cl 2 ], which exhibits a distorted square-antiprismatic {u 2 Ga 4 Cl 2 }-cage. Finally, the reaction of [{Cp*hCl 2 } 2 ] with InCp* lead to the formation of the zwitterionic complex [Cp* 2 h] + [Cp*h(InCp*){In 2 Cl 4 (μ 2 -Cp*)}] -. The striking feature 28

45 II. State of knowledge. of the anion is the Cp*-ring bridging the two InCl 2 -units, which is best described as a μ 2 -η 1 - η 1 -en-allylic coordinated C 5 Me 5 ring. (Cp*h) 2 (GaCp*) 3 (Cp*h) 2 (GaCp*) 4 (GaCl 3 ) (Cp*h) 2 (GaCp*) 2 Cl 2 6 GaCp* M = h L M Cl Cl M L GaCp* M = u Cp* Ga Cl u Ga Cp* Cl GaCp* *CpGa u - ECp*Cl ECp* - h [(Cp*) 2 h] + Cp*In Cl 2 In InCl 2 3 InCp* M = h L Cl M Cl Cl M Cl L 6 E E L M E ECl 2 M = h, L = Cp* M = u, L = p-cymene E = Ga, = Cp*; E = In, = Cp*, C(SiMe 3 ) 3 ) - GaCp*Cl 2 GaCl 3 L 3 GaCp* *CpGa Scheme 7: eactivity of [LMCl 2 ] 2 (M = h, L = Cp*; M = u, L = p-cymene) towards ECp*. M GaCl 3 GaCp* The use of other ruthenium complexes also shows insertion products upon reaction with E I Cp*. Thus, the reaction of [Cp*uCl] 4 with ECp* lead to the related compound [Cp*u(ECp*) 2 (ECp*Cl)] (E = Ga, In), which, in contrast to the above described compounds, readily reacts with the halide-abstraction reagent abph 4 to give [Cp*u(ECp*) 3 ][BPh 4 ]. [151] The compound [u(gacp*) 6 Cl 2 ], easily accessible by the reaction of either [u(dmso) 4 Cl 2 ] (dmso = dimethylsulfoxide) or [u(ph 3 P) 3 Cl 2 ] with GaCp*, exhibits fluxional chloride ligands and coalescence of all GaCp* moieties. 29

46 II. State of knowledge. Coordination chemistry of neutral four-, six-, and anionic five-membered heterocycles E I. eutral four-membered heterocycles. Because the four-membered heterocycles have only very recently been synthesized, the ability of these compounds as ligands in coordination chemistry is poorly explored. Only a handful compounds with the four-membered group 13 heterocycles Ga(Giso) and In(Giso) (Giso = [(Ar)C(Cy 2 )(Ar)]; Ar = C 6 H 3 Pr i 2-2,6; Cy = cyclohexyl) have been reported so far. Cy 2 Cy 2 Ar Ar Ga Ar Ga Pt Ga [Pt(norbornene) 3 ] Cy 2 Ar Ar Ar E : E = Ga, In Ar [i(cod) 2 ] or [(dppe)pt(c 2 H 4 )] L L M Ar E E Cy 2 Ar Ar Cy 2 Ar Cy 2 Ar M = i; E = Ga; L-L = 1,5-cod M = Pt; E = Ga, In; L-L = dppe Scheme 8: Transition metal-complexes of E(Giso). Thus, the complexes [(cod)i{ga(giso) 2 ] and [(dppe)pt{e(giso) 2 ] (dppe = bis(diphenylphosphanylethane; E = Ga, In) are obtained by the reaction of either i(cod) 2 (cod = 1,5-cyclooctadiene) with Ga(Giso) or [(dppe)pt(c 2 H 4 )] with E(Giso) (E = Ga, In). [60] In contrast to the reaction of i(cod) 2 or Pt(cod) 2 with E I Cp*, giving the homoleptic complexes [M(ECp*) 4 ], [(cod)i{ga(giso) 2 ] and [(dppe)pt{e(giso) 2 ] ( E = Ga, In) are not kinetically inert and the latter decomposes on redissolving crystals in toluene. The reaction of [Pt(norbornene) 3 ] with Ga(Giso) finally lead to the homoleptic complex [Pt{Ga(Giso)} 3 ], which exhibits a trigonal-planar Pt center. The Pt-Ga bond lengths in [Pt{Ga(Giso)} 3 ] are Å and are the shortest Pt-Ga bonds reported so far. As suggested from theoretical studies, the covalent component of the Pt-Ga bond possesses significant π-character (Table 8). 30

47 II. State of knowledge. eutral six-membered heterocycles. Whereas the coordination chemistry of the group 13 organyls ECp* is rather well explored, at the beginning of this work, only a hand full compounds with TM-E(DDP) bonds were known (Table 10). This fact could possibly be attributed to the greater interest of different research groups in the reactivity of these compounds (with the group 13 metal E either in the oxidation state +I or +III) in Main-Group chemistry. Especially for Al, a lot of effort was made by the group of Herbert W. oesky. [71, 73, 74, ] [44, 46, 158] Their results are reviewed very recently. Table 10: Transition metal-e(ddp) compounds reported previously to this study. TM-E(DDP) Literature [(DDP)GaFe(CO) 4 ] [103] [(1,3-cod)Pt{(DDP)Ga} 2 ]. [(dvds)pd{ga(ddp)}] [Pd{μ 2 -Ga(DDP)}CO] 2 [Pt{μ 2 -Ga(DDP)}CO] 2 [(Ph 3 P) 2 h{ga(ddp)}(μ-cl] [(Ph 3 P)Au{Ga(DDP)Cl}] [{(DDP)Ga}Au{Ga(DDP)Cl}] [(Ph 3 P)Au{Ga(DDP)Me}] [159] [159] [159] [159] [81] [80] [80] [80] The first example of a TM-E(DDP) compound was synthesized in 2003 by Hardman et al. The complex [(DDP)GaFe(CO) 4 ] [103] was obtained by simply mixing Ga(DDP) and Fe(CO) 5 in toluene at 0 C. [Fe(CO) 5 ] Ga(DDP) OC CO OC Fe Ga CO Scheme 9: Synthesis of [(CO) 4 FeGa(DDP)]. 31

48 II. State of knowledge. A comparison of the resulting CO stretching frequencies of the [(CO) 4 FeE I ] complexes and [Ph 3 PFe(CO) 4 ] was made by the authors (Table 11). As a result, the frequencies of all gallium-iron complexes are lower than the corresponding ones in the phosphine complex, thus implying, that the σ/π donor-acceptor ratio for the gallium ligands in general is greater than that of PPh 3. [103] Table 11: structural and spectroscopic data for [(CO) 4 Fe-Ga I ] compounds. Compound Ga coord. o. Fe-Ga (Å) CO stretch (cm -1 ) [Ar*GaFe(CO) 4 ] [35] (7) 2032, 1959, 1941, 1929 [(DDP)GaFe(CO) 4 ] [103] (4) 2010, 1940, 1915, 1900 [Cp*GaFe(CO) 4 ] [22] (4) 2037, 1966, 1942 [(Tp Me2 )GaFe(CO) 4 ] [105] (3) 2011, 1919, 1890 [Ph 3 PFe(CO) 4 ] [103] , 1979, 1947 Previous results from our group. Following the synthetic approach of substituting labile olefins, the reactions of Ga(DDP) with [Pt(cod) 2 ] and [Pd 2 (dvds) 3 ] were performed. In contrast to the reactions with GaCp*, an incomplete substitution of the bonded olefins was observed, possibly due to the difference in steric demand of Ga(DDP) vs. GaCp* (Tolman cone angle 147 vs. 112, Table 2). In [Pt(cod) 2 ], only one cod-ligand can be replaced by two Ga(DDP) moieties yielding [(1,3- cod)pt{ Ga(DDP)} 2 ]. [159, 160] Interestingly, in the course of the reaction, an isomerisation of the remaining 1,5-cod-ligand to the thermodynamically more stable 1,3-cod takes place. It is suggested, that the mechanism of this isomerisation is similar to the classical catalytic olefin isomerisation processes with a μ 3 -allyl hydride complex as an intermediate. The treatment of [Pd 2 (dvds) 3 ] yields only the substitution of the bridging dvds-ligand and the complex [(dvds)pd{ga(ddp)}] was formed in high yields (Figure 4). The C=C double bonds in the latter are found to be quite similar to the ones in [(HC)Pd(dvds)] (HC=(2,6-diisopropylphenyl-) 2 C 3 H 2 ), [161] but considerably longer than in [(Me 3 P)Pd(η 2 -η 2 -diallylether)], [162] which lead to the assumption, that the σ-donor properties of [Ga(DDP)] are more like those of HCs rather than electron-rich phosphines. However, both compounds [(1,3-cod)Pt{ Ga(DDP)} 2 ] and [(dvds)pd{ga(ddp)}] react with CO to give the dimeric, olefin-free complexes [M{μ 2 -Ga(DDP)}CO] 2 (M = Pd, Pt), which are 32

49 II. State of knowledge. the first examples of a Ga(DDP) fragment being located in a bridging position between two [159, 160] metal centers. Figure 4: Molecular structures of [(1,3-cod)Pt{Ga(DDP)} 2 ] and [(dvds)pd{ga(ddp)}]. As discussed previously for the GaCp* analogue, Ga(DDP) also readily inserts into transition metal-halogen bonds, but only three examples were reported at the beginning of this work. Besides the greater bulk of the Ga(DDP) ligand, the electronic features of Ga(DDP) are also distinctly different from the Cp* congener (vide supra). In summary, the Ga(I) center in Ga(DDP) is isolobal to HCs, the vacant p-orbital is not stabilized by additional π-electrons (as in the case of GaCp*, which is stabilized by the strong π-donor Cp*), which leads to an increase of the electrophilicity of the gallium center on coordination to transition metals. On the other hand, the oxidized species Ga(DDP)X 2 is less acidic than its counterpart GaCp*Cl 2, so a different behaviour of the ligands toward transition metal halogenides and alkyls can be expected. hcl(pph 3 ) 3 Ga(DDP) Ph 3 P Ph 3 P h Ga Cl Scheme 10: Insertion of Ga(DP) into the h-cl bond of [(Ph 3 P) 3 hcl]. 33

50 II. State of knowledge. The reaction of [(Ph 3 P) 3 hcl] with Ga(DDP) lead to the substitution of one Ph 3 P moiety by a Ga(DDP) fragment and the insertion into the h-cl bond to give the first example of a halidebridged transition metal-group 13 metal bond, [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)]. [81, 160] The most interesting feature of this structure is the chlorine-bridged h-ga bond. We suggest, that the steric demand of the coordinated ligands forbids the formation of a stable 16VE hl 4 complex and a more electrophilic 14VE hl 3 center is formed. Thus, two electrophilic metal centers are competing for the electrons of the basic chloride ligand, therefore forcing the chlorine as a 4VE donor into the bridging position. Consequently, [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)] can be considered as a frozen intermediate of an insertion reaction of GaDDP into a h-cl bond. In contrast to isolated GaCp*, which did not give isolable products, a mixture of GaCp* and GaI with [(Ph 3 P)AuCl] gave [Au 3 (μ-gai 2 ) 3 (Cp*Ga) 5 ]. [160, 163] However, molecular compounds featuring Au-Ga bonds were synthesized by the reaction of Ga(DDP) with [(Ph 3 P)AuCl] in high yields. [80] Depending on the Au/Ga ratio, the complexes [(Ph 3 P)Au{Ga(DDP)Cl}] (Au/Ga 1:1) and [{(DDP)Ga}Au{Ga(DDP)Cl}] (Au/Ga 1:2) are obtained (Figure 5). A detailed study of the course of the reaction revealed, that the insertion into the Au-Cl bond is preferred compared to a substitution of the Ph 3 P moiety. Thus, [(Ph 3 P)Au{Ga(DDP)Cl}] can be regarded as an intermediate in the formation of [{(DDP)Ga}Au{Ga(DDP)Cl}]. Finally, the insertion of Ga(DDP) into the Au-Me bond of [(Ph 3 P)AuMe] was also reported, giving only [(Ph 3 P)Au{Ga(DDP)Me}], independent to the Au/Ga ratio, as shown by 1 H-M spectroscopy. [80] Figure 5: Molecular structures of [(Ph 3 P)Au{ClGa(DDP)}] and [{(DDP)Ga}Au- {ClGa(DDP)}]. 34

51 II. State of knowledge. Anionic five-membered heterocycles. The anionic five-membered heterocycle [Ga{(Ar)C(H)} 2 ] - (Ar = C 6 H 3 Pr i 2-2,6) [39] is the group 13 analogue of the -heterocyclic carbene [C{(Ar)C(H)} 2 ] (HC) and also exhibits a free electron pair and thus should act as a strong σ-donor towards transition metals. Therefore, the group of Jones has recently embarked to systematically study the coordination chemistry of this ligand towards both transition metal and main group metal fragments. Some results of this study are summarized in the following section. Ar Ar Ga: Ar Cp'M(CO) n+1 -CO Me M = V, n = 3 M = Mn, n = 2 M = Co, n = 1 (CO) n M Ga Ar Scheme 11: Synthesis of [Cp M(CO) n {[Ga{(Ar)C(H)} 2 ]}] -. Comparable to HCs, the heterocycle [K(tmeda)][Ga{(Ar)C(H)} 2 ] readily reacts with binary metal carbonyl compounds via CO displacement. As seen in the reaction of [Fe(CO) 5 ] with [Ga{(Ar)C(H)} 2 ] -, one CO is replaced giving [(CO) 4 Fe{[Ga{(Ar)C(H)} 2 ]}] -. [106] As expected, the Ga-Fe bond distance in this complex is, due to the three-coordinated Ga-center, significantly longer than in the related diyl compound [(Ar*)GaFe(CO) 4 ] (Ar* = C 6 H 3 (C 6 H 2 Pr i 3-2,4,6) 2-2,6; two-coordinated Ga-center) [35] ( Å vs Å). Similar CO displacement reactions are observed on treatment of the complexes Cp M(CO) n+1 with the heterocycle leading to the anionic half-sandwich complexes [Cp M(CO) n {[Ga{(Ar)C(H)} 2 ]}] - (Cp = C 5 H 4 Me; M = V, n = 3; M = Mn, n = 2; M = Co, n = 1) (Scheme 11), respectively. [64] 35

52 II. State of knowledge. The reaction of [Ga{(Ar)C(H)} 2 ] - with MCp 2 (M = i, Co), proceeds via elimination of KCp and the anionic complex [CpM{[Ga{(Ar)C(H)} 2 ]} 2 ] - (M = i, Co) is obtained in moderate yields (Scheme 12). [164] Similar reactions are observed for HC, which yield via Cp anion displacement, the cationic complexes [CpM(HC) 2 ] +. However, the i-compound [Cpi{[Ga{(Ar)C(H)} 2 ]} 2 ] - readily reacts with the HC [C{(Me)C(Me)} 2 ] giving the neutral, square planar complex [(trans-hc) 2 i{trans-[ga{(ar)c(h)} 2 ]} 2 ]. The Ga-i bond distance herein is increased by ca Å, which was suggested to arise because of the trans-disposition of the electrophilic gallium heterocycles. Another very recent report deals with the synthesis of a lanthanide complex of [Ga{(Ar)C(H)} 2 ] -, [(L)d- [Ga{(Ar)C(H)} 2 ]] (L = Bu t CH 2 CH 2 {C(CHCHBu t )}), which was, again, synthesized via a salt metathesis reaction. [63] Ar Ga Ar MCp 2 -Cp - Ar Ga Ar M Ar Ga Ar : -Cp - Ar Ga Ar M Ar Ga Ar M = i, Co Scheme 12: Synthesis of [Cpi{[Ga{(Ar)C(H)} 2 ]} 2 ] - and its reaction with a HC. The synthesis of several other compounds containing TM-E bonds do not proceed via the direct reaction of [Ga{(Ar)C(H)} 2 ] - with the appropriate TM complexes. Instead, the digallane [Ga{(Ar)C(H)} 2 ] 2, easily prepared by oxidative coupling of [Ga{(Ar)C(H)} 2 ] - with [FeCp* 2 ] + or [Co(CO) 4 ] -, is used as precursor. [165] Thus, the formation of the complexes [Cp 2MGa{(Ar)C(H)} 2 ] (M = V, Cr; Cp = C 5 H 4 Me or C 5 H 5 ) [64] is suggested to proceed via the oxidative insertion of the metallocene into the Ga-Ga bond to give [Cp 2MGa{(Ar)C(H)] 2 } 2 ] as intermediates, which via subsequent conproportionation with excess metallocene give the corresponding compounds (Scheme 13). 36

53 II. State of knowledge. Ar Ar Ar Ga: Ar [Fe(Cp) 2 ] + or [Co 2 (CO) 8 Ga Ar Ga Ar {M(C 5 H 4 Me) 2 ] i) "ZrCp 2 " ii) Bu n Li Me Me M III Ar Ga Ar Ar Ga Ar Zr III {Li(THF) 4 ] + Ar Ga Ar Scheme 13: Synthesis of [Cp Me 2MGa{(Ar)C(H)} 2 ] (M = V, Cr) and [Cp 2 Zr{Ga{(Ar)C(H)} 2 } 2 ]. Several other complexes with main group metals are also reportedand recently reviewed. [44] A very interesting result in this respect is the reduction of [I 2 Ga{(Ar)C(H)] 2 } 2 ] with the group 2 metals Mg and Ca in THF to afford the complexes [(THF) n MGa{(Ar)C(H)] 2 } 2 ] (M = Mg, n = 3; M = Ca, = 4), [66] respectively (Figure 6). Interestingly, salt-elimination reaction of anhydrous MI 2 (M = Mg, Ca, Sr) with Ga{(Ar)C(H)] 2 } - 2 did not give isolable products. The authors suggest the reaction to proceed via reduction of [I 2 Ga{(Ar)C(H)] 2 } 2 ] to the dimeric compound [IGa{(Ar)C(H)] 2 } 2 ] 2, which is reduced further to give the digallane [Ga{(Ar)C(H)] 2 } 2 ] 2 and finally form the isolated complexes. ot surprisingly, the group 2- gallium heterocycle interaction was calculated to have significant ionic character. 37

54 II. State of knowledge. Figure 6: Molecular structures of [(THF) n MGa{(Ar)C(H)] 2 } 2 ] (M = Mg, Ca, = 3, 4). Bond activation reactions in TM-group 13 metal complexes. The major part of the transition metal chemistry of low valent group 13 compounds E I has been focused on structural and theoretical issues. Accordingly, a variety of organometallic compounds and clusters such as [M a (E I ) b ] have been synthesized so far. However, the great resemblance to CO, phosphanes or HC s, which have been widely used as supporting and directing ligands for reactivity and selectivity in organometallic bond activation chemistry in the last decades, creates interest into wether low valent group 13 compounds also exhibit potential as ligands in such activation reactions. Taking into account that the donor-acceptor bond in [L n M E I ] (E = Al, Ga) has been calculated to be rather polar, the electron density of transition metal centers should increase considerably upon coordination of the E I fragments. This increase in electron density at a transition metal center is the key for C-H activation of alkanes, which generally occurs via oxidative addition to a late transition metal center (Fe, u, Ir, h, Pt ). In particular, the high electron density at the metal center increases the π-back donation into the σ*-orbital of the alkane, leading to the C-H bond cleavage and the formation of the oxidative addition product [L n M x+2 (H)()]. [166] ecent results have now revealed that this concept is also relevant for E I compounds. The reaction of [i(cod) 2 ] with four equivalents of AlCp* in benzene did not give the homoleptic complex [i(alcp*) 4 ], as expected from the previous results with GaCp*. Instead, the compound [i(h)(alcp*) 3 (AlCp*Ph)] is formed in almost quantitative yields. [56] This is interesting, because such C-H activation reaction of benzene does not occur in the case of [i(co) n ], [i(p 3 ) n ] or other d 10 i complexes. It is suggested, that the activation takes 38

55 II. State of knowledge. place on a reactive, low-coordinated [i(alcp*) n ] (n<4) fragment, which could not be formed from [i(alcp*) 4 ] by a dissociative process, but is rather an intermediate in its formation. The suggestion of [(AlCp*) n i(h)(c 6 H 5 )] being a suitable intermediate is further supported by the formation of the hydrosilyl complex [(AlCp*) 3 i(h)(siet 3 )], the product of the reaction of HSiEt 3 [i(cod) 2 ] and AlCp*, respectively, which in benzene selectively reacts to [i(h)(alcp*) 3 (AlCp*Ph)] by liberating HSiEt 3 (Scheme 14, left). + 1 / 4 [AlCp*] 4 n-hexane *CpAl AlCp* i AlCp* AlCp* [i(alcp*) 3 ] + 1 / 4 [AlCp*] 4 - HSiEt 3 AlCp* H i AlCp* Et 3 Si AlCp* Me Me h L -L L = dmso, pyridine + GaCp* Me Me h GaCp* C-C bond activation C 6 H 6 - HSiEt 3 partial decomposition h Ga(CH 3 ) / 4 [AlCp*] 4 C 6 H 6 *CpAl H AlCp* i AlCp* AlCp* Scheme 14: C-H, Si-H and C-C bond activation reactions at ECp*-containing TM complexes. The compound with the empirical formula [Fe(AlCp*) 5 ], [167] synthesized by the reaction of [(η 6 -C 6 H 5 CH 3 )Fe(η 4 -C 4 H 8 )] with AlCp*, also shows an activation of a C-H bond in the course of the reaction. Thus, the solid state structure of [Fe(AlCp*) 5 ] reveals a central Fe center in distorted trigonal bipyramidal coordination environment with the AlCp* moieties not coordinating the iron center in an uncommon fashion. Instead, two C-H bonds of two Cp* ring methyl groups are attached to adjacent AlCp* ligands, giving an unusual {Cp*Al-CH 2 (C 5 Me 4 )Al-CH 2 (C 5 Me 4 )Al} chelating system, with the hydrides being in bridging positions between the Al and Fe centers. 39

56 II. State of knowledge. Although activation reactions are observed, it is not absolutely clear whether the transition metal, the group 13 metal or both centers are involved in such activations. However, the electrophilic E δ+ centers in such TM-E I complexes must also be considered to play a crucial part in such activation reactions. In this respect, the reaction of GaCp* with [Cp*h(CH 3 ) 2 (L)] (L = dimethyl sulfoxide, pyridine) giving the zwitterionic species [Cp*h{(η 5 -C 5 Me 4 )Ga(CH 3 ) 3 ] was studied in both experimental as well as theoretical manner in more detail (Scheme 14, right). [57, 168] As a result, the first step of the reaction is the substitution of the ligand L by GaCp* to give the thermally unstable compound [Cp*h(GaCp*)(CH 3 ) 2 ], which is isostructural to the starting complex. In a further step, the C-C bond activation takes place, also slowly at room temperature, and [Cp*h{(η 5 - C 5 Me 4 )Ga(CH 3 ) 3 ] is formed. The driving force includes the migration of the methyl groups to the ECp* fragment, accompanied by the oxidation of the group 13 center and the formation of a strong E-C bond. It is suggested that during the reaction both metal centers effectively cooperate in a redox cascade creating intermediates with exceptionally reactive metal centers in close proximity to each other and thus creating a number of low activation energy pathways. Thus, E I ligands represent strong donor ligands which are not innocent spectators such as phosphanes or HCs but actively take part in such activation reactions. 40

57 II. State of knowledge. Summary and goals of this study Summary and goals of this study. Starting with the aggregated E I Cp* compounds, a variety of monomeric, low valent group 13 E I complexes in solution as well as in the solid state were synthesized and characterized so far, including the Tris(pyrazoyl)hydroborato-derivative (Tp tbu2 )Ga, the bulky aryl-compound Ar*Ga, as well as several compounds with a group 13-heterocycle (4-, 5- and 6-membered rings) like the guanidinate E{(Ar)C(H)} 2, the diazabutadienido [Ga{(Ar)C(H)} 2 ] - and the bis-imidinate E(DDP) (E = Al, Ga). These compounds are formally isolobal with CO, phosphanes and HCs and exhibit a lone electron pair at the group 13 metal center, which is the reason for the high potential of these compounds to act as suitable ligands in the coordination chemistry towards d-metal centers. Besides the strong σ-donor properties, the ability of ECp* (E = Al, Ga, In) to effectively bridge two or three transition metal centers via μ 2 - or μ 3 -coordination modes has led to an intensive study of the coordination chemistry of this ligand-system. In contrast to HCs, the bonds of these ligands to transition metals have been calculated to be very polar, the strongest contribution being the coulombic attraction between the metals. In order to elucidate the general properties of these new class of ligands, early research concentrated on transition metal carbonyl fragments. However, recent research is focusing on more reactive transition metal fragments, without competing strong π-acceptor ligands such as CO. A general pathway to establish M-E bonds is the substitution of labile ligands like olefins and phosphines by E I, which lead to the characterization of a wide variety of mixed metal complexes of different nuclearities and structures like the homoleptic [M a (E I ) b ] (M = i, Pd, Pt, b>a>1, E = Al, Ga, In) and several heteroleptic complexes [M a (E I ) b L c ]. The ease of oxidation of E I provides a less common, yet alternative synthetic pathway. Thus, E I Cp* compounds can act as in situ reducing reagents towards higher oxidation state transition metal complexes [L n MX n ], giving either insertion products if the oxidised group 13 metal remains coordinated to the transition metal center or low oxidation state complexes [M a (E I ) b ] by elimination of X 2 E I. The strong polarity of the metal-metal bond is quite interesting in the context of organometallic chemistry. In particular, for bond activation reactions, the high increase in electron density of transition metal centers by E I is potentially useful for oxidative addition reactions. Thus, the activation of a benzene molecule, which is suggested to proceed via a reactive intermediate [i(alcp*) 3 ], and the C-C bond activation of a Cp* moiety at a h(i) center shows, that these exotic metalloid ligands E I may be very useful as ancillary ligands in classical organometallic chemistry. 41

58 II. State of knowledge. Summary and goals of this study However, the reactivity of the unsaturated fragments [M(E I ) n ] is not yet understood. To gain deeper insight into such activation reactions, the stabilization of the reactive intermediates might help to elucidate the relevant steps. A major problem, however, is the thermodynamically favourable formation of electronically and sterically saturated clusters such as [M(E I Cp*) 4 ], which prevents further reactivity. One strategy to overcome this problem is a tailored increase of the steric bulk of E I, leading to electronically unsaturated clusters which can additionally coordinate the desired small molecules. With a Tolman cone angle of 112, which is comparable to those reported for PMe [104] 3, GaCp* represents the smallest ligand in the series of monomeric, low valent group 13 organyls (Table 2), thus the steric stabilisation of unsaturated transition metal centers is not possible, yielding the formation of saturated complexes [M(ECp*) 4 ]. The sterically more demanding Ga(DDP) exhibits a Tolman cone angle of 147 and thus is comparable in steric bulk to PPh 3. As seen in the synthesis of [(1,3-cod)Pt{Ga(DDP)} 2 ], the formation of saturated complexes is avoided by the steric demand of the ligand. This fact, together with its strong σ- donor capabilities, are suitable properties to allow the formation of unsaturated compounds of the type [M(E I ) n ] (n < 4) or intermediates in, for example, insertion reactions, as already seen in the stabilization of the h complex[(ph 3 P) 2 h{ga(ddp)}(μ-cl)]. [81] Moreover, the ability of this ligand to effectively bridge transition metal centers in a μ 2 -coordination mode holds much promise for the discovery of novel cluster compounds and unusual reactions. However, sterically more demanding ligands, which are available at the time of this work (E I Tp, E I Mes*, Tolman cone angles of 183 and 218, respectively) might also provide appropriate properties in this respect, but due to the more convenient preparation of Ga(DDP), this is the ligand of choice for investigation in more detail. Thus, the aim of this study is to gain a more detailed insight into the coordination chemistry of the six-membered heterocycles E(DDP) (E = Al, Ga), with some emphasis on the Gaanalouge Ga(DDP). In this respect, the main focus is the synthesis of unsaturated complexes [M{Ga(DDP)} n ] (n < 4) and compounds with M-E bonds, which should be accessible via classical ligand substitution of labile bonded ligands like olefins or phosphines and insertion reactions into transition metal halide or -alkyl bonds. Whereas comparably less is known about the coordination behaviour of E(DDP), a detailed analysis of the obtained complexes with single crystal x-ray diffraction as well as solution M spectroscopy will reveal the differences and similarities in the coordination chemistry of the bis-imidinate Ga(DDP) 42

59 II. State of knowledge. Summary and goals of this study compared to that of the group 13 organyls E I Cp* and related species, as well as of CO and phosphines. Also, a comparison between Ga(DDP) and -heterocyclic carbenes (HC) will be drawn, because both ligand classes show several similarities in both steric properties as well as electronic features. The dimension of these similarities will be discussed in this work for the sake of a more specific characterisation and classification of this ligand class from an experimental point of view. As shown in the case of E I Cp*, the compounds with a M-E bond exhibit an amazing reactivity. It remains to verify how the bonding properties of the Ga(DDP) ligands (σ-donor, π-acceptor, polarity) compete with the protective properties of the bulky groups of the DDP-backbone in such complexes and what the consequences are in terms of reactivity and stability. Therefore, a detailed investigation of the reactivity of Ga(DDP) containing complexes, for example, in substitution reactions at the metal center or the use of such compounds a building blocks for the synthesis of larger transition metal clusters, according to the reactivity of E I Cp*, will be performed and discussed herein. 43

60 III. esults and discussion. 1. Insertion of Ga(DDP) into M-X bonds III. esults and discussion. 1. Insertion reaction of Ga(DDP) into M-X bonds. The insertion of low-valent group 13 fragments into the M-X bonds of appropriate precursors is a common and well studied synthetic pathway for the synthesis of stable bonds between d-metal centers and group 13 metals. The first examples of this route appeared more than 50 years ago and involve simple low valent group 13 metal halides E I X (E = Ga, In, X = Cl, Br, I). [128, , ] This chemistry always suffers from the poor solubility of the polymeric E I X species as well as the resulting complexes, thus a detailed structural prediction of the compounds was not possible. One way of overcoming this dilemma was the use of the recently accessible low valent group 13 organyls E I which exhibit better solubility of the carbenoid ligand as well as their insertion products. Me Me Me Ga Me Ga 4 Me Ga Cl h 1 + GaMe 3 + [h(coe) 2 Cl] 2 in C 6 H 6 + SiCl 4 Ga 7 Cl + ZnMe 2 Si Cl Cl Cl Me Ga Zn 2 Me Ga Ga Ga Me 5 Ga Me Me Sn Ga Ga Cl Cl 6 + GaMe 3 + SnMe 2 Cl2 Ga(DDP) + t BuCl Ga + ZnCl 2, THF + t BuC Ga 8 O C Cl O Zn 3 Cl Ga 9 Cl Scheme 15: Insertion reactions of Ga(DDP) into transition- and main group metal halide and methyl bonds, discribed in detail in this chapter. 44

61 III. esults and discussion. 1. Insertion of Ga(DDP) into M-X bonds Moreover, with the possibility of tuning the electronic as well as steric properties of the E I fragment by varying the organic rest (Table 2), a detailed insight into such insertion reactions might be possible by stabilizing reactive intermediates. As a result, a variety of different insertion products of ECp* into rhodium, ruthenium and iron-halide and alkyl bonds was reported in the last decade (Chapter II). [57-59, 81, 151, 167, 168] Furthermore, the sterically more bulky bis-imidinate Ga(DDP) readily inserts into the Au-Cl of [(Ph 3 P)AuCl] to give [{(DDP)ClGa}Au{Ga(DDP)}], which was found to proceed via the isolable intermediate [(Ph 3 P)Au{Ga(DDP)Cl}]. [80] In order to understand the insertion reactions of these new ligands in more detail, a variety of reactions with both main group and transition metal complexes were performed (Scheme 15), which exhibit either a M-Cl or M-CH 3 motif. In all cases, the insertion of the Ga(DDP) moiety into the M-X bond takes place, giving new complexes with a coordinated {(DDP)GaX} motif. 45

62 III. esults and discussion Insertion of Ga(DDP) into h-cl bonds [Lit. 4, Chapter VII.3] 1.1. Insertion of Ga(DDP) into h-cl bonds. The concept of trapping a reactive intermediate by the bulky Ga(DDP) ligand was demonstrated by the reaction of [hcl(pph 3 ) 3 ] with Ga(DDP). Here, the replacement of a Ph 3 P and the insertion of the Ga(DDP) moiety into the h-cl bond takes place giving [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)] in high yields. [81] Obviously, the steric demand of the coordinated ligands (Ph 3 P, Ga(DDP)) forbids the formation of a stable 16VE hl 4 complex. Therefore, the hypothetic complex [(Ph 3 P) 2 h{clga(ddp)}] with a more electrophilic 14VE hl 3 center is formed. In this complex, two electrophilic metal centers (Ga, h) are competing for the electrons of the basic chloride ligand, therefore "forcing" the chlorine as a 4VE donor into the bridging position, giving [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)] as isolable product. Thus, compound [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)] can indeed be regarded as a frozen intermediate of the insertion of Ga(DDP) into a h-chloride bond Insertion of Ga(DDP) into the h-cl bond of [h(coe) 2 Cl] 2. In order to strengthen the suggestion that the steric demand is a limiting factor in this reaction, the reaction of a sterically less hindered h(i) compound, [h(coe) 2 Cl] 2 (COE = cyclooctene), with Ga(DDP) was performed. With this precursor, a replacement of the more labile olefins and, due to the reduced steric demand of the ligands, the formation of a stable 16VE [hl 4 ] fragment should be possible. Indeed, according to Scheme 16, the replacement of one olefin by the Ga(I) ligand with full migration of the chloride from the rhodium to the gallium center was observed. With the coordination of the solvent C 6 H 6, the 18 VE compound [(COE)(η 6 -benzene)h{(ddp)gacl}] (1) was isolated as deep green solid. It should be noted that the reaction in absence of C 6 H 6 or in present of other olefins (coe, hexene) did not give defined products. ecrystallisation of the crude product from hexane gives 1 as deep green crystals in 60% yield. h Cl Cl h + 2 Ga(DDP) benzene, T - 2 coe 2 Ga Cl h 1 Scheme 16: eaction of [h(coe) 2 Cl] 2 with Ga(DDP). 46

63 III. esults and discussion Insertion of Ga(DDP) into h-cl bonds The 1 H M spectrum of 1 in THF-d 8 at room temperature shows the expected signal set for a {(DDP)GaCl}-moiety with reduced C s symmetry, i.e. two distinct septets for the i Pr-CH groups and four distinct doublets for the i Pr-CH 3 groups, respectively. The COE signals are found at 2.67, 1.83, 1.22 and 1.03 ppm, whereas a sharp singlet signal at 5.65 ppm is assignable to a coordinated C 6 H 6 molecule. Interestingly, in C 6 D 6 no signal for coordinated C 6 H 6 was detected, instead, a singlet for free C 6 H 6 at 7.15 ppm integrating to about six protons was found. Obviously, an exchange between coordinated C 6 H 6 and the deuterated solvent C 6 D 6 takes place. In accordance with this, the 13 C M spectrum of 1 in C 6 D 6 shows a triplet signal in the range of a coordinated C 6 D 6 ligand (96.1 ppm) besides the expected signals for the coordinated {Ga(DDP)Cl} and COE moieties. Such arene exchange reactions are also known for other h I compounds, such as in [(h(η-arh)(η-c 8 H 14 ){SnCl( 2 ) 2 }] (ArH = PhMe, C 6 H 3 Me 3-1,3,5 or C 6 Me 6 ) [173] and [(h(η-arh)(p(oph) 3 ) 2 ]ClO 4 (ArH = PhMe, C 6 H 3 Me 3-1,3,5 or C 6 H 6 ). [174] Figure 7: Molecular structure of [(COE)(η 6 -benzene)h{(ddp)gacl}] (1). For structural parameters see Table

64 III. esults and discussion Insertion of Ga(DDP) into h-cl bonds Crystals suitable for X-ray analysis could be obtained by slow evaporation of the solvent of a saturated benzene-solution. Compound 1 crystallizes in the triclinic space group P 1. Selected bond lengths and angles are shown in Table 12. The molecular structure in the solid state (Figure 7) can be described as a half-sandwich complex with a two legged piano-stool coordination sphere consisting of a η 2 -coordinated cyclooctene ligand and a {(DDP)GaCl}- moiety. The covalent Ga-Cl interaction is reflected by the only slightly distorted tetrahedral environment of the gallium center (sum of angles of the 1-2-Ga-h plane: 345 ). Also, the Ga 2 C 3 ring does not adopt planar geometry. Thus, the Ga 2 and the C 3 2 planes of this ring system intersect each other by an angle of This is a rather common geometrical feature for E(III) compounds [E(DDP)X 2 ] (E = Al, Ga, In) [175] and also observed for insertion products of Ga(DDP), for example, in [(Ph 3 P)Au{Ga(DDP)Cl}] (29.3 ). [80] Table 12: Selected bond lengths (Å) and angles ( ) for [(COE)(η 6 -benzene)h{(ddp)gacl}] (1). h-ga (8) h-ga-cl (4) Ga-Cl 2.321(2) 1-Ga (12) Ga (3) 1-Ga-h (9) Ga (3) 2- Ga-h (9) h-c (4) h-c (4) C30-C (5) The h-ga-cl bond angle of 112 is also in accordance with a complete migration of the chlorine atom from the rhodium to the gallium center in the course of the reaction. The same is true for the h-cl distance, which at 3.92 Å is clearly a non-bonding interaction. The Ga-Cl bond distance (2.32 Å) is significantly longer than those in [(DDP)GaCl 2 ] (2.22 Å), [175] but only slightly longer than in the gold compounds [{(DDP)Ga}Au{Ga(DDP)Cl}] and [(Ph 3 P)Au{Ga(DDP)Cl}] (Ga-Cl: Å). [80] The h-ga distance (2.40Å) is longer than in [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)] (2.3870(6)), [81] but in the same range as found in the GaCp* compounds [Cp*h(GaCp*) 2 (GaCl 3 )] (h-ga: Å) or the dimeric complex [(Cp*h) 2 (GaCp*) 3 ] (h-ga: Å), [58] respectively. 48

65 III. esults and discussion Insertion of Ga(DDP) into h-cl bonds Interestingly, the h-c olefin bond distances in 1 (h-c COE : 2.113(4) Å and 2.141(4) Å) are significantly shorter than in the dimeric compound [hcp*(c 2 H 4 )(μ 2 -GaCp*)] 2 (h-c ethylene : Å), [58] suggesting a higher degree of π-backbonding in 1. This is consistent with a relatively long C=C bond distance of the coordinated olefin in 1 (1.424(5), which is slightly longer than the related C=C bond distance in [( i Pr 3 P)h(COE)(C 6 H 6 )]. [176] The h-c benzene bond distances lay within 2.209(4) to 2.347(4), with a h-c 6 H 6-centroid bond distance of Å and thus are in the same range as reported for coordinated benzene molecules at a h center, e.g. in [( i Pr 3 P)h(COE)(C 6 H 6 )]. 49

66 III. esults and discussion Insertion into Zn-Me and Zn-Cl bonds [Lit. 9, Chapter VII.3] 1.2. Insertion reaction of Ga(DDP) into Zn-Me and Zn-Cl bonds. Look of the large number on M-E complexes of these type of ligands reported to date, the elements of the Zn-group, Zn, Cd and Hg, have thus far remained an exception. Two examples with very short Hg-In interactions are known, [177] whereas still no molecular complexes of Hg and Cd with the lighter group 13 ligands [Al] n or [Ga] n have been isolated and characterized so far. Most recently however, the group of C. Jones communicated the salt elimination reaction of [Ga{[(Ar)C(H)] 2 }] - with,-chelated zinc chloride complexes to give the very first examples of compounds featuring a Zn-Ga bond. [61] In contrast to that, the related neutral six-membered heterocycle Ga(DDP) [178] readily inserts into metal-halide bonds to give neutral complexes with a M-Ga-X motif (X = halide), as seen in the case of h(i) (see chapter 1.1) as well as Au(I) [80] complexes. Thus, Ga(DDP) was reacted with appropriate Zn precursors to form Zn-Ga bonds Insertion reaction of Ga(DDP) into the Zn-methyl bond of ZnMe 2. The reaction of Ga(DDP) with Me 2 Zn in a ratio of 2:1 in flourobenzene at room temperature gives the insertion product [{(DDP)GaMe} 2 Zn] (2) as orange-red crystals in yields around 80% (Scheme 17). Compound 2 was found to be air sensitive in solution as well as in the solid state, but can be stored in an inert gas atmosphere for several weeks without decomposition. It dissolves well in common, aprotic organic solvents like hexane, THF or benzene. ZnMe Ga(DDP) Me Ga Zn 2 Me Ga Scheme 17: Synthesis of [{(DDP)GaMe} 2 Zn] (2). The 1 H-M spectrum of 2 in CDCl 3 at room temperature shows only one set of signals for the DDP ligand with a reduced (C S ) symmetry (e.g. two signals for the i Pr-CH protons). This is in agreement with an insertion of a Ga(DDP) moiety in each Zn-Me bond of ZnMe 2 and a 50

67 III. esults and discussion Insertion into Zn-Me bonds full migration of the methyl group from the Zn to the Ga center, giving two {(DDP)GaMe}- moieties. In accordance with this, the i Pr-Me protons give rise to four distinct doublet signals at 1.28, 1.24, 1.22 and 1.09 pm (each 6H), whereas a signal at ppm (6H) can be assigned to the protons of the methyl groups bound to the gallium atoms. The 13 C-M spectrum is in well agreement with these results. Deep orange-red crystals of 2 suitable for single crystal x-ray analysis could be obtained by recrystallizing the raw product from hexane at -30 C overnight. The molecular structure of 2 is shown in Figure 8. Selected bond lengths and angles are shown in Table 13. Figure 8: Molecular structure of [{(DDP)GaMe} 2 Zn] (2). For structural parameters see Table 13. Compound 2 shows a nearly ideal linear coordination of the central Zn by the Ga-atoms (Ga- Zn-Ga: ) with equal Ga-Zn bond distances of 2.46 Å. These bond lengths are significantly longer than in [{ i Pr 2 C[(Ar)] 2 }ZnGa{[(Ar)C(H)] 2 }] (2.32 Å), but comparable to those in [(tmeda)zn{ga{[(ar)c(h)] 2 }} 2 ] (Ar = C 6 H i 3 Pr 2-2,6) (2.45 and 2.43 Å). [61] In the course of the reaction, the Ga(DDP) ligand inserts into the Zn-Me bond, giving a distorted tetrahedral geometry at the Ga center (sum of angles of the --Ga-Zn plane: 315 ). As expected for an insertion product, the Ga metals lie out of the C 3 2 plane showing a dihedral angle between the C 3 2 and Ga 2 arrays of 32.2 (Ga1) and 35.4 (Ga2), 51

68 III. esults and discussion Insertion into Zn-Me bonds respectively. The {(DDP)GaMe}-moieties are oriented perpendicular to each other (torsion angle C100-Ga1-Ga2-C200: 93.4 ), obviously to minimize steric strain of the i Pr-Me groups of the DDP-ligand and the methyl groups at the gallium centers. This is in contrast to the 1 H- M spectrum observed for 2. Expectedly, a fast rotation around the Zn-Ga bond occurs in solution which indicates a chemical equivalence of the {(DDP)GaMe}-moieties in solution, whereas this rotation is frozen in the solid state, due to steric interactions. The average Ga-Me bond distances (1.99 Å) are distinctly longer than in [(DDP)GaMe 2 ] (av Å) [175] or [(DDP)GaMeCl] (1.96 Å), [179] possibly as a result of the lower oxidation state and thus less electrophilic character of the gallium in 2. In agreement with that, the Ga- bond distances are 2.00 Å and 2.03 Å, which are only slightly shorter than in free Ga(DDP) (2.05Å), [31] but similar to those found in [{(DDP)Ga}Au{Ga(DDP)Cl}] (1.99 Å for the {Ga(DDP)Cl}- moiety). [80] Table 13: Selected bond lengths (Å) and angles ( ) for [{(DDP)GaMe} 2 Zn] (2). Zn1-Ga (7) Ga1-Zn1-Ga (2) Zn1-Ga (7) 11-Ga (11) Ga1-C (3) 21-Ga (12) Ga2-C (4) Zn1-Ga1-C (10) Ga (3) Zn1-Ga2-C (11) Ga (3) C100-Ga1-Ga2-C Ga (3) Ga (3) As reported for [{(DDP)Ga}Au{Ga(DDP)Cl}], the first step of the reaction is the insertion of a Ga(DDP) moiety into the Au-Cl bond to give [ (Ph 3 P)Au{Ga(DDP)Cl}], followed by the coordination of the second Ga(DDP) ligand. [80] In order to elucidate, whether the insertion of Ga(DDP) into the Zn-Me bond is a concerted reaction or proceeds stepwise via the anticipated intermediate [MeZn{MeGa(DDP)}], we reacted ZnMe 2 with only one equivalent of Ga(DDP). Unfortunately, the variation of the Zn to Ga ratio only effects the yield of 2, which remains the only isolable, and also traceable product in solution M investigations. 52

69 III. esults and discussion Insertion into Zn-Cl bonds [Lit. 4, Chapter VII.3] Insertion reaction into the Zn-Cl bond of ZnCl 2. In contrast to the reaction of the five-membered heterocycle [Ga{[(Ar)C(H)] 2 }] -, which reacts with [(tmeda)zncl 2 ] under salt elimination to give [(tmeda)zn{ga{[(ar)c(h)] 2 }} 2 ], the reaction of ZnCl 2 with Ga(DDP) in THF does not give the related compound [{(DDP)GaCl} 2 Zn], respectively. Possibly because of the increased steric bulk of the ligand, accompanied by the additional coordination of a chlorine atom and two THF molecules in the in situ formed adduct [Zn(THF) 2 Cl 2 ], only a single insertion can be observed giving the compound [{(DDP)GaCl}ZnCl(THF) 2 ] (3) in good yields (Scheme 18), independent of the ligand to metal ratio used. Pale yellow prismatic crystals of 3 could be grown by storing the solution at -30 C overnight. It should be noted that several attempts to react ZnCl 2 with Ga(DDP) in fluorobenzene as a non-coordinating solvent did not give any isolable products, possibly due to the poor solubility of ZnCl 2 in non-coordinating solvents. Compound 3 is soluble in THF, Et 2 O or CHCl 3 and can be stored in the solid state in an inert gas atmosphere for several days without decomposition. Ga(DDP) + ZnCl 2 THF O O Zn Cl 3 Ga Cl Scheme 18: Synthesis of [{(DDP)GaCl}ZnCl(THF) 2 ] (3). The 1 H-M spectrum of 3 in CDCl 3 shows one signal for the DDP-backbone proton at 5.27 ppm as well as signals indicating a C 2v symmetry around the Ga(DDP) moiety (i.e. one signal for the i Pr-CH protons and two signals for the i Pr-Me protons), which is not in accordance with the solid state structure. Obviously, the chlorine atoms exchange fast in solution leading to coalescence of the DDP signals. However, also at low temperature (-60 C), no significant splitting of the signals could be observed. 53

70 III. esults and discussion Insertion into Zn-Cl bonds Crystals, suitable for single crystal x-ray diffraction were grown by slowly cooling a saturated solution of 3 in THF to -30 C. A mercury plot of 3 is shown in Figure 9, with selected bond lengths and angles summarized in Table 14. Compound 3 crystallizes in the monoclinic space group P2 1 /m. It consists of a Zn atom which is coordinated in a distorted tetrahedral geometry by two THF molecules, a chlorine and a {(DDP)GaCl}-moiety. Figure 9: Molecular structure of [{(DDP)GaCl}ZnCl(THF) 2 ] (3). For structural parameters see Table 14. The Zn-Cl bond (2.2016(14) Å) is slightly longer than in the parent compound [(THF) 2 ZnCl 2 ] (2.18 Å), [180] but well in the range for other Zn-Cl bonds reported in the literature. The same is true for the Zn-O bond distance, which is 2.1 Å (average) in 3 and slightly longer than in [(THF) 2 ZnCl 2 ] (2.00 Å). With 2.39 Å, the Ga-Zn bond distance is slightly shorter than in 2 (2.46 Å) or the tetracoordinated Zn complex [(tmeda)zn{ga{[(ar)c(h)] 2 }} 2 ] (2.45 Å and 2.43 Å), but significantly longer than in [{ i Pr 2 C[(Ar)] 2 }ZnGa{[(Ar)C(H)] 2 }] (2.32 Å). This results indicate, that the difference in the Zn-Ga bond length can be attributed to steric interactions of the coordinated group 13 ligands rather than to the zinc coordination numbers, as previously suggested for the five-membered heterocyle complexes. [61] The two chlorine atoms, the gallium and the zinc are coplanar (torsion angle Cl-Ga-Zn-Cl: 180 ), with the chlorine atoms being in trans-position to each other. 54

71 III. esults and discussion Insertion into Zn-Cl bonds Whereas the Ga-Cl bond distance (2.2820(10) Å) is slightly longer to those in [(DDP)GaCl 2 ] (2.15 Å 2.23 Å), [175] it is comparable to those in other transition metal complexes containing the {(DDP)GaCl}-moiety, such as [(Ph 3 P)Au{Ga(DDP)Cl}] (2.29 Å), [{(DDP)Ga}Au{Ga(DDP)Cl}] (2.28 Å) [80] or [(COE)(η 6 -benzene)h{(ddp)gacl}] (1) (2.32 Å), respectively. This comparably long Ga-Cl distance is in good agreement with the lability of the chloride observed in solution, although such an exchange reaction was not observed in the latter cases. Again, the Ga atom is located out of the C 3 2 ring by 0.53 Å (dihedral angle between the C 3 2 and Ga 2 arrays: ). Table 14: Selected bond lengths (Å) and angles ( ) for [{(DDP)GaCl}ZnCl(THF) 2 ] (3). Ga-Zn (6) 1-Ga-1* 94.18(13) Zn-O (2) Zn-Ga-Cl (3) Zn-Cl (14) Cl1-Zn-Ga (4) Ga-Cl (10) O1-Zn-O1* 87.26(17) Ga (2) Cl1-Ga-Zn-Cl

72 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) 1.3. Insertion reaction of Ga(DDP) into Gallium-CH 3 bonds. Trivalent organometallic compounds of the group 13 elements are known as Lewis acids. Classical examples of such acceptors include the trialkyls or triaryls of aluminium and boron. [181] The acceptor behaviour arises because of the presence of a formally vacant p- orbital and the absence of a lone pair of electrons on the group 13 element. In contrast to that, the alanediyls Al I and their heavier congeners E I (E = Ga, In) were found to exhibit a strong Lewis basicity, with their electron ground state calculated to be singlet in nature and with the singlet-triplet energy gap increasing with increasing atomic number. [111] Similar results were reported for the heterocycles E(DDP) (E = Al, Ga), [30, 31, 116] therefore E(DDP) compounds should be able to form group 13-group 13 donor-acceptor complexes. Indeed, a number of examples of compounds with group 13 group 13 donor acceptor bonds have been reported, including several homonuclear *E-E 3 and heteronuclear complexes *E-E 3 (E, E = B, Al, Ga, In; * = Cp*, DDP, bulky Aryl), mostly containing the Lewis acids B(C 6 F 5 ) 3 or Al(C 6 F 5 ) 3. [79, 103, 120, 122, 123, 125] Interestingly, only few examples of complexes with the related -heteroleptic carbenes (HC s) are reported, including ECl 3 (E = B, Al, In, Tl) or EMe 3 (E = Al, Ga) as the respective Lewis acids. [182] In order to shed light on the similarities and differences between HC s and Ga(DDP), the reactivity of Ga(DDP) towards the strong Lewis acids AlMe 3 and GaMe 3 was investigated in more detail eaction of Ga(DDP) with AlMe 3 and GaMe 3 : Insertion of Ga(DDP) into the Ga-methyl bond of GaMe 3. When mixing an excess of AlMe 3 with Ga(DDP) in C 6 D 6, no colour change of the solution could be observed. However, the 1 H-M spectrum of the reaction mixture shows a shift in the resonances for the Ga(DDP) moiety: The γ-proton of the backbone gives rise to a singlet at 5.14 ppm, whereas the i Pr-Me and i Pr-CH resonances are shifted to 1.06, 1.36 and 2.88 ppm, respectively. This result indicates the formation of a donor-acceptor bond between Ga(DDP) and AlMe 3, wherein the Ga(DDP) remains in a symmetric coordination environment, i.e. no insertion of the Ga(DDP) moiety into the Al-Me bonds takes place. This assumption is further supported by the difference in acidity of the group 13 metal centers (Al > Ga), which might prevent the rupture of the Al-Me bond and / or the insertion of the Gametal into the Al-Me bond. Therefore, the formation of a simple Lewis donor-acceptor complex like [(DDP)Ga AlMe 3 ] is a reasonable suggestion. However, all attempts to grow 56

73 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) crystals from a solution containing AlMe 3 and Ga(DDP) failed so far. Interestingly, when performing the reaction in hexane, drying the resulting yellow precipitate in vacuo and redissolving the solid in C 6 D 6, the M spectra solely shows signals for free, uncoordinated Ga(DDP). This fact indicates, that only in solution, a weak donor-acceptor interaction between the Ga center of Ga(DDP) and AlMe 3 can be observed, which does not remain upon crystallisation. In contrast to that, when adding one equivalent of GaMe 3 (as a 0.25 M solution in hexane) to a yellow solution of Ga(DDP) in hexane under vigorous stirring, the colour of the solution turns to intense yellow. Storing the resulting solution at -30 C overnight, bright yellow crystals of [{(DDP)GaMe}GaMe 2 ] (4) can be obtained in 81 % yield (Scheme 19). Complex 4 is quite air sensitive, but stable in an inert gas atmosphere for several days and can be dissolved in apolar organic solvents such as hexane or toluene. Ga(DDP) + xs. GaMe 3 Me Me Ga Ga Me Scheme 19: Formation of [{(DDP)GaMe}GaMe 2 ] (4). The 1 H-M spectrum of 4 in C 6 D 6 at room temperature (Figure 11) shows the typical set of signals for a Ga(DDP) moiety with reduced (C S ) symmetry, which is a result of the insertion of Ga(DDP) in one Ga-Me bond of GaMe 3 and known for the insertion into transition metal halide bonds (e.g. in [(Ph 3 P)AuCl] and [(COE) 2 hcl] 2 ). [80, 81] Additionally, signals at 0.37 ppm (6H) and ppm (3H) can be assigned to the methyl protons of the GaMe 2 moiety and the GaMe(DDP) ligand, respectively. The 13 C-M spectrum is in good agreement with this result, showing two distinct signals for {MeGa(DDP)} (1.4 ppm) and {GaMe 2 } (-6.1 ppm), respectively. 57

74 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) Single crystals for x-ray diffraction could be obtained from a saturated solution of hexane at -30 C. The molecular structure of 4 is shown in Figure 10. Selected bond lengths and angles are presented in Table 15. Figure 10: Molecular structure of [{(DDP)GaMe}GaMe 2 ] (4). For structural parameters see Table 15. Compound 4 consists of a Ga atom which is coordinated in a trigonal-planar geometry by two methyl and a {(DDP)GaMe} ligand, respectively. The Ga-Ga bond distance (2.45 Å) is normal for single, covalent Ga-Ga interactions and therefore similar to the ones found in the Ga(I)-Ga(III) adduct [Cp*Ga(GaCp*Cl 2 )] (2.43 Å) [122] or the dimer [{(CSiMe 3 )GaBr} 2 ] (2.43 Å). [183] Table 15: Selected bond lengths (Å) and angles ( ) for [{(DDP)GaMe}GaMe 2 ] (4). Ga1-Ga (18) 1-Ga (2) Ga (5) C30-Ga1-Ga (2) Ga (6) C31-Ga2-Ga (3) Ga1-C (6) C32-Ga2-Ga (2) Ga2-C (7) C31-Ga2-C (4) Ga2-C (7) C30-Ga1-Ga2-C

75 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) Interestingly, the Ga-C bond length of the methyl groups cis to each other are quite similar with 1.98 Å (Ga1-C30) and 1.99 Å (Ga2-C32) and thus comparable to those in [(DDP)GaMe 2 ] (av Å) [175] or [(DDP)GaMeCl] (1.96 Å), [179] respectively, whereas the bond length of the methyl group cis to the DDP backbone is surprisingly shorter (Ga2-C31: 1.81 Å). However, similar bond distances can be found e.g. in [(Me 3 Ga) 4 (C 6 4 )] (1.837 Å; C 6 4 = hexamethylenetetramine, 4, adamantane). [184] The insertion of the Ga(DDP) moiety into the Ga-Me bond is reflected in the distorted tetrahedral geometry at the Ga-center, coordinated by the DDP-backbone, the methyl and the GaMe 2 moiety, respectively (angular sum Ga---Me: ), with a dihedral angle between the C 3 2 and the Ga 2 arrays of The Ga- bond distances are 1.98 Å and 2.00 Å, which is consistent with an electron rich Ga-center. Similar Ga- bond distances can be found e.g. in [{(DDP)Ga}Au{Ga(DDP)Cl}] (1.99Å) [80] or [(coe)(η 6 -benzene)h{(ddp)gacl}] (1) (1.97 Å 2.01 Å) Synthesis of [{(DDP)GaMe} 2 GaMe]. Due to the high steric demand, it seems that the number of Ga(DDP) ligands coordinated to a metal center is limited. Up to day, only several examples of Ga(DDP) containing complexes are reported, with either one or a maximum of two Ga(DDP) ligands coordinating the metal center, [(DDP)GaFe(CO) 4 ] [103] or [{(DDP)Ga}Au{Ga(DDP)Cl}] [80] being the most prominent examples. As seen in the latter case, a two step synthesis involving the insertion of Ga(DDP) into the M-Cl bond followed by the substitution of the labile bonded Ph 3 P ligand, gives the twofold coordinated metal center. Another route comprises the concomitant insertion into the M-Me bonds of Me 2 Zn to give [{(DDP)GaMe} 2 Zn] (2), as shown in Chapter III With the synthesis of [{(DDP)GaMe}GaMe 2 ] (4) (see above, chapter III ), the question arises, if this compound is the thermodynamic product in the course of the reaction or if a second insertion, similar to the formation of 2, can be performed. However, the existence of a second remaining methyl group at the metal center must be considered as an obstacle in the formation of such a compound. The effect of steric demanding ligands are known to be crucial in the synthesis of Ga(DDP) containing complexes, as seen in the different behaviour of [(Ph 3 P) 3 hcl] and [(COE)hCl] 2 towards Ga(DDP) (see chapter III.1.1.). 59

76 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) evertheless, on slow addition of a GaMe 3 solution in hexane to predissolved Ga(DDP) (2 equivalents, with respect to GaMe 3 ), the colour of the solution again turns to intense yellow (after the addition of ½ eq. GaMe 3 ) and then changes to orange after the addition was completed. Stirring the reaction mixture for 1h at room temperature and cooling it to -30 C overnight, a mixture of bright yellow crystals of 4 (as proven by single crystal x-ray diffraction analysis) and deep orange crystals of [{(DDP)GaMe} 2 GaMe] (5) deposited (Scheme 20). The overall yield of 5 was low (22-25 %) and could only determined by separating the crystals manually. 5 is stable in the solid state in an inert gas atmosphere for several days without decomposition, but readily dissociates to 4 and Ga(DDP) on redissolving in all common organic solvents at room temperature, as proven by 1 H-M spectroscopy. Therefore, variable-temperature- 1 H-M spectroscopy of a 2:1 mixture of Ga(DDP) and Me 3 Ga in toluene was performed (Figure 11). Ga(DDP) + GaMe 3 Me Me Ga Ga Me + Ga(DDP) - Ga(DDP) Me Ga Ga Me Me Ga high temperature low temperature Scheme 20: Synthesis of [{(DDP)GaMe}GaMe 2 ] (4) and [{(DDP)GaMe} 2 GaMe] (5). Only the signals for the γ-h of the DDP-backbone of the Ga-ligand (in the range of 5.20 ppm to 4.70 ppm), which can be used as a diagnostic tool in this chemistry, as well as signals of the methyl protons coordinated at a Ga-center (in the range of 1 ppm to -0.5 ppm) are discussed. At 50 C, only signals for free Ga(DDP) (5.18 ppm) and 4 (4.83 ppm) are observed. According to this, the GaMe- and GaMe 2 -protons are found at 0.34 and ppm. Besides these two sets of signals (slightly shifted to due to temperature effects: 5.18, 4.81, 0.38 and ppm, respectively), new signals occur on cooling the solution to room temperature. These signals are assigned to [{(DDP)GaMe} 2 GaMe] (5), showing a singlet at 4.74 ppm and 60

77 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) two related signals for the GaMe- (0.42 ppm) and the {(DDP)GaMe}-protons (-0.01 ppm) in the ratio 1:2. At further cooling to -40 C, the signals for free Ga(DDP) and 4 vanish and only the set of signals for 5 remains (4.64 ppm (2H), 0.76 ppm (3H) and ppm (6H)). Figure 11: 1 H-M spectra of a 2:1 mixture of Ga(DDP) and Me 3 Ga at various temperatures. This results show, that a temperature dependent equilibrium between 5, 4 and Ga(DDP) exists, which is shifted towards 5 at lower temperature (Scheme 20). This fact mirrors well the steric interaction of the third methyl group at the Ga center in 5 with the coordinated {(DDP)GaMe} moieties. At higher temperature, the atoms in the molecule are more flexible, thus an increase of the steric repulsion of the DDP-aryl groups and the methyl group takes place. Therefore, along with entropic effects, the second insertion is disfavoured at higher temperatures and the reverse reaction, de-insertion, takes place. 61

78 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) Crystals of 5, suitable for x-ray diffraction were grown by cooling a 2:1 mixture of Ga(DDP) and Me 3 Ga in hexane at -30 C overnight and separating orange crystals of 5 from yellow crystals of 4. Compound 5 crystallizes in the triclinic space group P-1 with half of the molecule in the unit cell and the {(DDP)GaMe}-moieties showing a disorder which could not be refined properly. Selected bond lengths and angles are shown in Table 17. Figure 12: Molecular structure of [{(DDP)GaMe} 2 GaMe] (5). For structural parameters see Table 17. Figure 12 shows the molecular structure of 5, which consists of a central GaMe-moiety which is trigonal planar coordinated by a methyl and two {(DDP)GaMe}-moieties (angular sum Ga-Ga-Ga-Me: ). Thus, the compound contains a chain of three metal atoms connected by two Ga-Ga bonds with Ga-Ga bond distances of 2.505Å, which are slightly longer than in 4 (2.4508(18)), [{(PEt 3 )GaI 2 } 2 Ga(I)(PEt 3 )] (2.451(1) and 2.460(1) Å) [185] and [( tbu acac)ga{c(sime 3 ) 2 }] 2 (2.45 Å). [186] A comparison can be made between compound 5 and other trigallanes, which are summarized in Table 16. In contrast to 5, which shows a central three-coordinated and two terminal four coordinated Ga-centers, [GaH 2 (Ga{C()C(H)} 2 ) 2 ][K(tmeda)] [187] contains a central four-coordinated and terminal 62

79 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) three-coordinated Ga-centers, whereas [Ga{Ga(GePh 3 ) 3 } 2 ][Li(THF) 4 ]. [188] exhibits a central two-coordinated and terminal four-coordinated Ga-metals. Table 16: Overview about known trigallanes (chaines and triangles). Compound Ga-Ga bond length (Å) Structur [GaH 2 (Ga{C()C(H)} 2 ) 2 ][K(tmeda)] [187] (9) Ga 3 -chain [{(PEt 3 )GaI 2 } 2 Ga(I)(PEt 3 )] [185] 2.451(1), 2.460(1) Ga 3 -chain [{(DDP)GaMe} 2 GaMe] (5) (19), (17) Ga 3 -chain [Ga{Ga(GePh 3 ) 3 } 2 ][Li(THF) 4 ] [188] 2.548(7), 2.536(7) Ga 3 -chain [( t Bu 3 Si) 4 Ga 3 ] [189] (7), 2.879(1) Ga 3 -triangle [( t Bu 3 Si) 4 Ga 3 ][(THF) 2 a(18-crown-6)] [189] 2.494(1), 2.569(2), 2.935(2) Ga 3 -triangle The Ga-Ga bond distance in 5 is in between the ones of [GaH 2 (Ga{C()C(H)} 2 ) 2 ][K(tmeda)] (2.4071(9) Å) and [Ga{Ga(GePh 3 ) 3 } 2 ][Li(THF) 4 ] (2.548(7) and 2.536(7) Å), which can be possibly explained by the increasing steric demand of the ligands (Ga{C()C(H)} 2 < {Ga(DDP)Cl} < Ga(GePh 3 )), which is also reflected in increasing Ga-Ga-Ga bond angles (107 in [GaH 2 (Ga{C()C(H)} 2 ) 2 ][K(tmeda)], 112 in 5 and 179 in the Ga(GePh 3 ) 3 compound). However, comparable bond distances can be found in [(Ar)GaI] 2 (2.49 Å ; Ar = bis(2,6-diisopropylphenyl)phenyl) [190] or [(acac)ga{c(sime 3 ) 3 }] 2 (2.51 Å,) [191] respectively. Table 17: Selected bond lengths (Å) and angles ( ) for [{(DDP)GaMe} 2 GaMe] (5). Ga1-Ga (19) Ga1-Ga3-Ga (6) Ga2-Ga (17) C30-Ga1-Ga (3) Ga (4) C31-Ga2-Ga (2) Ga (4) C32-Ga3-Ga (3) Ga (4) C32-Ga3-Ga (3) Ga (4) 1-Ga (15) Ga1-C (8) 1-Ga (16) Ga2-C (8) C30-Ga1-Ga3-C (5) Ga3-C (10) C31-Ga2-Ga3-C C30-Ga1-Ga2-C

80 III. esults and discussion Insertion into E-Me bonds (E = Al, Ga) The gallium atoms and the methyl groups are coplanar (torsion angles 1.1 and 1.8 ) and coordinated trans to each other giving a Y-shaped arrangement (Ga-Ga-Ga angle: ). The Ga-Me bond lengths 1.96 Å to 1.98 Å, are in the same range found for [(DDP)GaMe 2 ] (1.97 Å and 1.98 Å), [175] [(DDP)GaMeCl] (1.96 Å) [179] or other GaMe containing complexes, as a survey of the Cambridge Crystallographic Database revealed. The DDP-backbones are both orientated perpendicular towards the GaMe-plane. Obviously due to steric interactions, only one (DDP)Ga ring is planar (Ga1), whereas the second one is tilted (52 ). This tilting is known for other Ga(DDP) containing complexes (e.g. [{(DDP)Ga}Au{Ga(DDP)Cl}] (22.4 ) [80] or [(COE)(η 6 -benzene)h{(ddp)gacl}] (1) (29.8 )), but due to less steric restraints in these complexes, the tilt angle of 52 (Ga(2) 2 - C 3 2 ) in 5 is significantly larger. The Ga- bond distances in 5 range from Å to Å which are quite long, presumably because of steric interactions. 64

81 III. esults and discussion Insertion into group 14 element bonds 1.4. Insertion into group 14 element bonds. The reaction of Ga(DDP) with metal-halides is not restricted to transition metal centers. As discussed in previous chapters, the insertion into the Ga-Me bond of GaMe 3 occurs, giving the complexes [{(DDP)GaCl}GaMe 2 ] (4) and [{(DDP)GaCl} 2 GaMe] (5), respectively. Also, Ga(DDP) can be inserted into the C- bond of t BuC to give the oxidative insertion product [(DDP)Ga( t Bu)(C)] (8) (chapter III.1.6). In this respect, the reactivity of E(DDP) (E = Al, Ga) towards a variety of different reagents has been investigated in the last years. With some emphasis on [(DDP)AlX 2 ] (X = Cl, I, OH, H), oesky et al. intensively studied the reactivity of these compounds towards group 15 to 17. [46, 70, 74, 76, 179, 192, 193] Thus, the dihydroxide [(DDP)Al(OH) 2 ] [78] and its higher congeners [(DDP)Al(EH) 2 ] (E = S, Se) [77, 194] could be synthesized, using the dihydrid or diiodide, respectively. Also, dimerisation reactions are observed to give [{(DDP)Al(EH)} 2 (μ 2 -E)] (E = O, Se) [77]. The compound [(DDP)Al(C 2 2 )], [74] easily prepared by the reduction of [(DDP)AlI 2 ] with K in the presence of C 2 2 ( = SiMe 3, Ph), was found to exhibit facile insertion reactions into the Al-C bond by CO 2, CS 2, ketones and nitriles. For a brief overview see literature. [46] However, only a few insertion reactions of the low-valent E(DDP) compounds into element-element bonds are reported so far. Thus, Al(DDP) inserts into the O-O π-bond of O 2 to give the dimeric compound [(DDP)Al(µ 2 -O)] 2, [153] accompanied by the oxidation of the Al(I) center. The related Ga-compound can be prepared by the reaction of Ga(DDP) with 2 O, respectively. [195] Whereas the reaction of Ga(DDP) with S 8 yields the analogue [(DDP)Ga(µ 2 -S)] 2, [195] in the case of Al, the homobimetallic derivative of the sulfur crown S 8, [(DDP)Al(μ-S 3 ] 2, was obtained. [196] In contrast to HC s, which were extensively studied by Arduengo [197], Gehrhus [198], Kuhn [199, 200] and Weidenbruch [201, 202] in the late 1990 s, the coordination of the organyls E I towards late group 14 metals (Ge, Sn, Pb) is limited to only few examples reported by Jones et al in [67] Thus, the use of the heavier alkene analogues 2 E=E 2 (E = Ge, Sn; = CH(SiMe 3 ) 2 ), yield the insertion of the heterocycle [:Ga{[(Ar)C(H)] 2 }] - into the E=E double bond and the formation of the ionic complexes [K(tmeda)][ 2 EGa{[(Ar)C(H)] 2 }] and [K(tmeda)][Sn[Ga{[(Ar)C(H)] 2 }] 2 ]), respectively. The monomeric guanidinato element chlorides, [(Giso)ECl] (E = Ge, or Sn, Giso = [Pr i 2C{(Ar)} 2 ] - ) also react with [:Ga{[(Ar)C(H)] 2 }] - to give the neutral [(Giso)EGa{[(Ar)C(H)] 2 }] via salt elimination metathesis. Apparently, no insertion reactions of E I organyls into group 14-halide or -methyl bonds are reported so far. 65

82 III. esults and discussion Insertion into Sn-Cl bonds Insertion into Sn-Cl bonds of SnCl 2 Me 2. Ga(DDP) readily reacts with [SnMe 2 Cl 2 ] in hexane at room temperature. Thus, when mixing [SnMe 2 Cl 2 ] with two equivalents of Ga(DDP), the colour of the solution turns slightly orange after several minutes. Concentrating and cooling the solution to -30 C overnight gave [Me 2 Sn{ClGa(DDP)} 2 ] (6) as colourless crystals in high yields (Scheme 21). It readily dissolves in all common organic solvents like hexane or THF and can be stored in an inert gas atmosphere for several days, whereas upon contact with air immediate decomposition occurs. Interestingly, only the insertion into the Sn-Cl bond occurs, whereas the insertion into the Sn- Me bond or a reduction of the Sn(IV) center to a Sn(II) center, accompanied by the oxidation of Ga(DDP) to [Cl 2 Ga(DDP)], could not be observed, also when reacting an excess of Ga(DDP) with [SnMe 2 Cl 2 ] in boiling THF. Whereas the latter observation might result from the strong tendency of Sn(II) to act as reducing agent and be oxidized to the Sn(IV) species, the former may be explained by a stronger Ga-Cl bond compared to Ga-Me, respectively. However, further studies to shed light into this selectivity are necessary. Additionally, the reaction of less than two equivalents Ga(DDP) with [SnMe 2 Cl 2 ] only reduces the yield in the reaction and does not effect the reaction product. Additionally, it should be noted, that SnMe 4 behaves inert against Ga(DDP) and no reaction was observed also in 119 Sn M spectroscopy. Me Me Sn Cl Cl + 2 Ga(DDP) Me Me Sn Ga Ga Cl Cl Scheme 21: Synthesis of [Me 2 Sn{ClGa(DDP)} 2 ] (6). The 1 H-M spectrum of 6 at room temperature shows the expected signals for a Ga(DDP) moiety, which is inserted into a metal-halogen bond, i.e. a set of signals with reduced C s symmetry. Thus, the backbone-ch proton gives rise to a signal at 4.96 ppm, whereas the septet-signals at 3.84 ppm and 3.10 ppm can be assigned to the i Pr-CH protons. Expectedly, 66

83 III. esults and discussion Insertion into Sn-Cl bonds the i Pr-CMe protons show four resonances at 1.31 ppm, 1.24 ppm, 1.21 ppm and 0.93 ppm, respectively. The signal for the protons of the SnMe 2 moiety can be found at ppm, showing the expected 119 Sn-satellites with a coupling constant of 21.7 Hz. The 119 Sn-M spectrum shows a resonance at considerably low field at 1189 ppm relative to SnMe 4. This downfield-shift compared to the resonances for the Sn-Ga complexes [K(tmeda)][Sn{CH(SiMe 3 ) 2 } 2 Ga{[(Ar)C-(H)] 2 }] - (δ = ppm) and [K(tmeda)][(Ar ) 2 SnGa{[(Ar)C(H)] 2 }] - (Ar = C 6 H 2 Pr i 3-2,4,6; δ = ppm) is not surprising considering the anionic nature of these complexes, which can be compared to trialkyl stannate anions (e.g., LiSnMe 3 ; δ = -189 ppm). [203] Comparable shifts are reported for the neutral complexes [(Giso)SnGa{[(Ar)C(H)] 2 }] (δ = 455 ppm) or the related HC complex [(HC)Sn(Ar ) 2 ] (δ = 710 ppm). [202] The 13 C-M spectrum is in well agreement with the insertion of a {Ga(DDP)}-moiety into a metal-halogen bond, showing the expected signals. Figure 13: Molecular structure of [Me 2 Sn{ClGa(DDP)} 2 ] (6). For structural parameters see Table

84 III. esults and discussion Insertion into Sn-Cl bonds Crystals, suitable for single crystal x-ray analysis could be obtained by cooling a saturated solution of 6 in hexane to -30 C overnight. Compound 6 crystallizes in the monoclinic space group P2 1 /c. The molecular structure is depicted in Figure 13 and consists of a central SnMe 2 moiety which is distorted tetrahedral coordinated by two {ClGa(DDP)} moieties, which result from the insertion of the Ga(DDP) ligands into the Sn-Cl bonds. Selected bond lengths and angles are shown in Table 18. The Sn-Ga bond distances are almost identical with (7) Å and (7) Å, which are significantly shorter than the reported Sn-Ga bond distances in the anionic complexes [ 2 SnGa{[(Ar)C(H)] 2 }] - (2.72 Å) and [Sn[Ga{[(Ar)C(H)] 2 }] 2 ] (2.63 Å and 2.66 Å) ( = CH(SiMe 3 ) 2 ). [67] Also, the neutral complex [(Giso)SnGa{[(Ar)C(H)] 2 }] exhibits longer bond distances (2.68 Å), possibly as a result of the steric demanding ligands or the lower oxidation state of the Sn-metal center in this complex. The distorted tetrahedral coordination environment at the Sn center is reflected in a comparably large Ga-Sn-Ga bond angle of 127 and a rather small angle of 103 for the Me- Sn-Me arrangement. The Ga-centers of the Ga(DDP) ligand are also distorted tetrahedrally coordinated by the nitrogen atoms of the DDP-backbone, the Sn and the Cl atoms, leading to dihedral angles between the C 3 2 and the Ga 2 arrays of (Ga1) and (Ga2), respectively. The difference in the bending of the Ga 2 C 3 rings can be attributed to packing effects in the crystal. The Ga-Cl bond distances are similar with 2.24 Å and thus comparable to those in the dichloride [(DDP)GaCl 2 ] (2.22 Å) [175], but slightly shorter than in 1 (2.32 Å). Similar to the formation of [{(DDP)GaMe} 2 GaMe] (5), the chlorine atoms in 6 are located at the opposite side of the Ga-Sn-Ga chain with respect to the Me groups at the Sn center, therefore avoiding steric stress in the structure. Table 18: Selected bond lengths (Å) and angles ( ) for [Me 2 Sn{ClGa(DDP)} 2 ] (6). Sn-C (5) Ga (4) Sn-C (5) Sn-Ga (7) C60-Sn-C (2) Sn-Ga (7) Ga1-Sn-Ga (2) Ga1-Cl (15) 11-Ga (18) Ga2-Cl (15) 21-Ga (18) Ga (4) Sn-Ga1-Cl (4) Ga (4) Sn-Ga2-Cl (4) Ga (4) 68

85 III. esults and discussion Insertion into Sn-Cl bonds The Sn-Me bond distances are similar with an average of Å and therefore slightly longer than in the complex [(HC)SnMe 2 Cl 2 ] [200] (Sn-Me: av Å, HC = 2,3-dihydro- 1,3-di-3 -methoxypropyl-4,5-dimethylimidazol-2-ylidene), which shows a SnMe 2 Cl 2 moiety simply coordinated by a HC ligand in a σ-donor mode. otably, the complex [(HC)SnMe 2 Cl 2 ] represents the only example of a SnMe 2 Cl 2 moiety coordinated by a HCligand reported so far. This example clarifies well the different coordination chemistry of Ga(DDP) and HCs: Whereas HCs only coordinate in a σ-donor mode to metal halides, Ga(DDP) readily inserts into the metal halogen bonds to form M-Ga-X motifs. 69

86 III. esults and discussion Insertion into Si-Cl bonds 1.5. Insertion reaction of Ga(DDP) into the Si-Cl bond of SiCl 4. As seen in the case of [SnMe 2 Cl 2 ], Ga(DDP) readily inserts into the M-Cl bond of the heavier group 14 analogues. It is interesting to investigate, whether if the reactivity or selectivity in such insertion reactions depends on the M-X compound used in these reactions. In this respect, the tetrahalide compound SiCl 4 was used in order to form a Si-Ga bond following an insertion pathway. Whereas several compounds with Ga-Si bonds are reported, their formation generally proceeds via salt elimination metathesis, for example, the reaction of [ t Bu 3 SiM] and EX 3 (M =a, Li, X = halogene). Upon mixing Ga(DDP) with SiCl 4 in a ratio of 1:1 in hexane or toluene (Scheme 22), the colour of the solution stays yellow. However, after concentrating the hexane solution, a colourless precipitate was formed. Dissolving the precipitate by gently heating the solution to reflux afforded, after cooling to room temperature, colourless crystals of [Cl 3 Si{ClGa(DDP)}] (7) in high yields. It should be noted that 7 is the only isolable product in this reaction, even though using more than one equivalent of Ga(DDP). Similar to the reaction with t BuCl (chapter III.1.6), Ga(DDP) inserts into the Si-Cl bond of SiCl 4 by oxidative addition of the Cl atom and the SiCl 3 moiety at the Ga-center. Compound 7 is soluble in several organic solvents (hexane, toluene, benzene) and stable at room temperature in an inert gas atmosphere for several weeks without decomposition. SiCl 4 + Ga(DDP) Ga Si Cl Cl Cl Cl Scheme 22: Synthesis of [Cl 3 Si{ClGa(DDP)}] (7). The 1 H-M spectrum suggests the insertion of a Ga(DDP) into the Si-Cl bond of SiCl 4 : The heterocycle-ch proton give rise to a singulet at 4.88 ppm, whereas the expected splitting of the signals for the i Pr-CH (3.75 Å and 317 Å) and i Pr Me protons (1.51 ppm, 1.38 ppm, 1.18 ppm and 0.99 ppm) complete the set of signals for an unsymmetrically coordinated Ga(DDP) ligand. Again, the 13 C-M spectrum is in good agreement with this data. 70

87 III. esults and discussion Insertion into Si-Cl bonds Figure 14: Molecular structure of [Cl 3 Si{ClGa(DDP)}] (7). For structural parameters see Table 19. Compound 7 crystallizes in the monoclinic space group P2 1 /m. Its molecular structure (Figure 14) consists of a distorted tetrahedrally coordinated Ga center, similar to the structure of [(Cl)( t Bu)Ga(DDP)] (9) (chapter III ). Interestingly, only the insertion into one Si-Cl bond can be observed, which is in contrast to the double insertion into both Sn-Cl bonds of [SnMe 2 Cl 2 ] to give [Me 2 Sn{ClGa(DDP)} 2 ] (6). As expected, the Ga metal lies significantly above the C 3 2 plane (0.639Å), causing a dihedral angle of Table 19 shows selected bond lengths and angles of 7. The Ga-Cl bond length of (11) Å is significantly shorter than in free [(DDP)GaCl 2 ] (2.22 Å), [175] in the h-compound 1 (2.32 Å) or the Zn-compound 3 (2.28 Å), but also in the range for Ga(III)-Cl bond distances reported throughout the literature. Table 19: Selected bond lengths (Å) and angles ( ) for [Cl 3 Si{ClGa(DDP)}] (7). Ga-Si (13) 1-Ga-1* 98.46(14) Si-Cl (2) Cl1-Ga-Si (5) Si-Cl (15) Cl2-Si-Cl (8) Ga (2) Cl3-Si-Cl (16) Ga-Cl (11) Cl2-Si-Ga (8) 71

88 III. esults and discussion Insertion into Si-Cl bonds The Ga-Si bond length of (13) Å is shorter than that reported in the tetramer [{( t Bu 3 Si)Ga} 4 ] (2.45 Å) [204] or the dihalide [( t Bu 3 Si)GaCl 2 (THF)] (2.42 Å). [205] Similar bond distances are reported e.g. for the dimeric compound [{(Me 3 Si) 3 SiGaMeCl} 2 ]. [206] otably, the Ga- bond lengths in 7 are significantly shortened with respect to free Ga(DDP) (2.05 Å) or those in [(DDP)GaCl 2 ] (2.22 Å). This fact indicates the increased positive polarization of the Ga center and the strong σ-donor character of Ga(DDP) upon coordination. Comparable Ga- bond lengths were found in the acid base adduct [Ph 3 B-Ga(DDP)] (1,94 Å average). [123] 72

89 III. esults and discussion Insertion into C- and C-Cl bonds 1.6. Insertion reaction of Ga(DDP) into C- and C-Cl bonds. [Lit. 7, Chapter VII.3] Insertion reaction of Ga(DDP) into C- bonds. The reaction of [Pt(cod){Ga(DDP)} 2 ] with an excess of t BuC gives deep red crystals of [Pt{μ-Ga(DDP)}( t BuC)] 2 (25) in high yields, which will be discussed in chapter III in more detail. However, yellow crystals could be obtained as a side product in this reaction. Single crystal x-ray analysis revealed the formation of [(DDP)Ga( t Bu)(C)] (8), which is obviously formed by an insertion of Ga(DDP) into the C- bond of t BuC (Scheme 23). It should be noted, that, best of our knowledge, this is the first example of an insertion reaction of a Ga(DDP) moiety into a sp 3 -hybridised C- bond reported so far. Ga(DDP) + t BuC Ga C Scheme 23: eaction of Ga(DDP) with t BuC to give [(DDP)Ga( t Bu)(C)] (8). After heating a mixture of Ga(DDP) and excess t BuC in hexane to reflux and stirring for 1h, the former yellow solution turns into pale yellow and colourless crystals of 8 could be obtained in 89% yield after storing a saturated solution of 8 at -30 C overnight. Compound 8 was found to be quite stable in solution as well as in the solid state and also for a short period of time in contact with air. It dissolves well in aprotic organic solvents like hexane, THF or Benzene. The 1 H-M spectrum of 8 in C 6 D 6 at room temperature shows the expected signals for the DDP-moiety with reduced (C s ) symmetry as well as signals for the t Bu-methyl groups, respectively, The signal of the γ-h proton is upfield shifted to 4.88 ppm, which is in well agreement with an oxidization of the Ga(I) center to a Ga(III) center in the heterocycle. Thus, similar 1 H-M shifts were observed for Ga(III) compounds of the type [(DDP)Ga 2 ], like [(DDP)GaI 2 ] (4.90 ppm) [175] or the unsymmetrically substituted [(DDP)Ga(Me)Cl] (4.90 ppm). [179] The 13 C M spectrum is in well agreement with these results. 73

90 III. esults and discussion Insertion into C- and C-Cl bonds Figure 15: Molecular structure of [(DDP)Ga( t Bu)(C)] (8). For structural parameters see Table 20. Figure 15 shows a mercury plot of the molecular structure of 8 with selected bond lengths and angles summarized in Table 20. The Ga-center is distorted tetrahedrally coordinated by the nitrogen atoms of the DDP backbone, the t Bu- and the C-moieties, again showing a significant bending of the six-membered ring of The Ga- bond distances of 1.943(6) and 1.956(6) Å are longer than the distances found in the dihalides [(DDP)GaX 2 ] (X = F, Cl, Br, I: Å), [179] but shorter than in [(DDP)GaMe 2 ] (1.979(2) Å and 2.001(2) Å) [175] or in free Ga(DDP) (2.05 Å). [31] Comparable Ga- bond lengths were found in the acid base adduct [Ph 3 B-Ga(DDP)] (1.94 Å average). [123] Compared to [(DDP)GaMe 2 ] (1.97 Å 1.98 Å), the Ga-C bond in 8 is significantly elongated with 2.029(8) Å (C30, C) and 1.997(7) Å (C31, t Bu-C). Table 20: Selected bond lengths (Å) and angles ( ) for [(DDP)Ga( t Bu)(C)] (8). Ga (6) 1-Ga (2) Ga (6) 1-Ga1-C (3) Ga1-C (8) 1-Ga1-C (3) Ga1-C (7) 2-Ga1-C (3) 3-C (9) 2-Ga1-C (3) C31-Ga1-C (3) 74

91 III. esults and discussion Insertion into C- and C-Cl bonds Insertion reaction of Ga(DDP) into C-Cl bonds. As shown in the reaction of t BuC with Ga(DDP) to yield 6, Ga(DDP) not only inserts into transition metal halide or methyl bonds, also activation of the sp 3 -C- bond of t BuC takes place. In order to elucidate the reactivity of Ga(DDP) towards organic compounds, we attemted to activate other C- bonds (= halides, organic rests) by inserting Ga(DDP) into the corresponding bonds of F-Ph, Cl-Ph, CH 2 Cl 2 and CHCl 3. Whereas no reaction was observed in the cases of F-Ph and Cl-Ph, decomposition and undefined products could be observed on reaction with CH 2 Cl 2 and CHCl 3. In contrast to that, the reaction of Ga(DDP) with t BuCl yields [{(DDP)Ga}( t Bu)Cl] (9) as colourless crystals, in analogy to t BuC (Scheme 24). Again, a formal insertion of the Ga(DDP) moiety into the C-Cl bond is accompanied with an oxidation of the Ga(I) center, giving a unsymmetric coordinated Ga(III)- center, respectively. Ga(DDP) + t BuCl Ga Cl Scheme 24: Synthesis of [{(DDP)Ga}( t Bu)Cl] (9). Thus, the 1 H-M spectrum of 9 in C 6 D 6 at room temperature is in good agreement with this result, showing the expected signals for the DDP-moiety with reduced (C s ) symmetry, i.e. two distinct signals for the i Pr-CH and four signals for the i Pr-Me groups, respectively. The 13 C- M spectrum ratifies this observation. The molecular structure of 9 (Figure 16) is quite similar to that of 8, apart from a chlorine atom instead of a C moiety coordinated to the Ga-center. Selected bond lengths and angles can be found in Table 21. Whereas the Ga-C tbu bond distance in 9 is similar to that in 8 (1.994(5) Å), it is elongated compared to the Ga-C bond distance in [(DDP)Ga(Me)Cl] (1.95 Å), obviously due to steric effects. The same is true for the Ga-Cl bond distance, which was found to be 2.25Å in 9 and slightly shorter in [(DDP)Ga(Me)Cl] (2.22 Å). Also, the dihedral angle between the C 3 2 and the Ga 2 arrays is quite similar (32.38 ). 75

92 III. esults and discussion Insertion into C- and C-Cl bonds Figure 16: Molecular structure of [(DDP)Ga( t Bu)(Cl)] (9). For structural parameters see Table 21. The Ga- bond distances, 1.953(4) Å and 1.965(3) Å, are only slightly longer than in 8 (1.943(6) Å and 1.956(6) Å), but significantly elongated compared to the dihalide [(DDP)GaCl 2 ] (1.906 Å and Å) or the silicon compound [Cl 3 Si{ClGa(DDP)}] (7) (1.917 Å). Table 21: Selected bond lengths (Å) and angles ( ) for [(DDP)Ga( t Bu)(Cl)] (9). Ga (3) 1-Ga (15) Ga (4) 1-Ga-C (17) Ga-C (4) 2-Ga-C (15) Ga-Cl 2.251(2) 1-Ga-Cl (11) 2-Ga-Cl (13) C30-Ga-Cl (15) 76

93 III. esults and discussion. 1. Summary and conlusion Summary and conlusion. The low valent group 13 compounds are known to act as reducing agents towards a variety of different metal centers. [45, 54, 140] However, several examples of ECp* being inserted into transition metal halide bonds are reported so far. These reactions are very sensitive towards [58, 151, stoichiometry and reaction conditions, giving a variety of different insertion products. 152] In contrast, the insertion of Ga(DDP) into metal-halide and -methyl bonds is comparably simple, generally showing the full migration of the halide or methyl group from the metal to the Ga-center leading to a terminal M-Ga-X motif (X = Me, Cl). Thus, a variety of insertion reactions with both transition metal and main-group metal centers have been performed (Scheme 15), leading to new insertion products of Ga(DDP). Table 22: Summary of the synthesized compounds via insertion reactions. Compound covalent radius a) coordination number [(COE)(η 6 -benzene)h{(ddp)gacl}] (1) 3 [{(DDP)GaMe} 2 Zn] (2) 125 pm 2 [{(DDP)GaCl}ZnCl(THF) 2 ] (3) 125 pm 4 [{(DDP)GaMe}GaMe 2 ] (4) 125 pm 3 [{(DDP)GaMe} 2 GaMe] (5) 125 pm 3 [Me 2 Sn{ClGa(DDP)} 2 ] (6) 140 pm 4 [Cl 3 Si{ClGa(DDP)}] (7) 117 pm 4 [(DDP)Ga( t Bu)(C)] (8) [(DDP)Ga( t Bu)(Cl)] (9) a) only the covalent radii of those metal centers are shown which show both single and double insertions. The values are taken from [207] and Table 22 gives an overview about the synthesized complexes via insertion into M-X bonds. The differences in reactivity, that are the occurance of mono- or double-insertions, seem to depend on the metal center or, more specifically, the size of the metal atom and their coordination number in such complexes. Therefore, when using a comparably small metal center (e.g. Si, covalent radius 117 pm) with a high coordination number of 4, only single insertion is possible due to the tremendous steric demand of the (DDP)Ga ligand, as seen in the formation of 7 (Chapter III.1.5.). Increasing the size of the metal (Ga or Zn, both 125 pm), the number of coordinated ligands becomes the limiting factor. If the coordination number is only 2 (as in Me 2 Zn), a clean double insertion takes place, giving solely 2 as isolable product 77

94 III. esults and discussion. 1. Summary and conlusion (Chapter III.1.2.). If one additional ligand is present at the metal (coordination number: 3, e.g. in GaMe 3 ), the outcome of the insertion reaction depends on the reaction temperature, as seen in the temperature-depending equilibrium of the formation of 4 and 5 (Chapter III.1.3.). Furthermore, if the coordination number is increased to 4 (similar atom size of 125 pm, but starting with the in-situ formed [Zn(THF) 2 Cl 2 ], giving complex 3, Chapter III.1.2.), only a single insertion is observed. Finally, if the metal atom becomes larger (e.g. Sn, covalent radius 140 pm, coordination number: 4), the coordination number is neglible and the double insertion to give 6 (Chapter III.1.4.) is the only identifiable result. Selected crystallographic details of the new compounds 1-9 are summarized in Table 23. The comparison of the Ga-Cl bond distances of the synthesized species 1, 3, 6 and 7 shows, that on insertion into the TM-Cl bond, the Ga-Cl bond distances (apart from 7) is slightly elongated in all complexes compared to the dichloride [(DDP)GaCl 2 ] (2.2 Å) or the mixed chloride/methyl compound [(DDP)GaClMe] (2.2 Å). This is in contrast to the related Ga-Me bond length in the complexes 2, 3 and 5, which show very similar Ga-Me bond length compared to [(DDP)GaMe 2 ] ( Å) or [(DDP)GaClMe] (1.96 Å). Generally, the coordination of Ga(DDP) to a transition metal is accompanied by an electrophilisation of the Ga-metal of Ga(DDP), which is reflected in the shortening of the Ga- bond distances in the complexes (1.91 to 2.0 Å) compared to free Ga(DDP) (2.05 Å). Similarly, the coordinated {(DDP)GaX} moieties exhibit smaller -Ga- angles compared to the respective dihalide (100.1 ), dimethyl (93.9 ) or the mixed methyl/chloride (97.1 ) compounds. Additionally, the insertion of Ga(DDP) into M-X bonds generally gives a distorted tetrahedrally coordinated Ga-center. Thus, the planarity of the DDP-backbone (dihedral angle between the Ga and C 3 2 plane: 0.38 ) is lost on coordination. However, the observed dihedral angles between the Ga and C 3 2 plane in the compounds 1-9 are depending on the steric situation at the metal center, but showing angles in the range reported for compounds of the type [(DDP)GaXY]. 78

95 III. esults and discussion. 1. Summary and conlusion Table 23: Comparison of structural features of the (DDP)GaX moieties. Compound -Ga bond distance in Å -Ga- Bond angle ( ) Ga-X bond distance in Å a) Dihedral angle ( ) C 3 2 -Ga (3) / 2.018(3) 93.12(12) 2.321(2) (Cl) (3) / 2.027(3) (Ga1) 90.86(11) (Ga1) 1.984(4) (Me) 32.2 (Ga1) 1.998(3) / 2.027(3) (Ga2) 91.18(12) (Ga2) 1.995(3) (Me) 35.4 (Ga2) (2) 94.18(13) 2.282(1) (Cl) (5) / 2.001(6) 93.0(2) 1.981(6) (Me) (4) / 2.113(4) (Ga1) 89.63(15) (Ga1) 1.964(8) (Me) 7.17 (Ga) 2.003(4) / 2.126(4) (Ga2) 89.05(16) (Ga2) 1.979(8) (Me) (Ga2) (4) / 1.977(4) (Ga1) 94.14(18) (Ga1) (15) (Cl) (Ga1) 1.962(4) / 1.979(4) (Ga2) 94.78(18) (Ga2) (15) (Cl) (Ga2) (2) 98.46(14) (11) (Cl) (6) / 1.956(6) 95.4(2) 2.029(8) (C) (3) / 1.965(4) 95.43(15) 2.251(2) (Cl) Ga(DDP) [31] (14) / (13) [175] 1.906(3) / Ga(DDP)Cl (3) [175] 1.979(2) (DDP)GaMe 2 /2.001(2) (DDP)GaMeCl [179] 1.935(1) / 1.949(1) 87.53(5) (1) 2.218(1) / 2.228(1) (7) 1.970(2) / 1.979(2) (1) 2.223(1) (Cl) 1.956(2) (Me) a) the Ga-X bond distances referr to the (DDP)GaX moieties in the complexes To conclude, these results show, that the insertion of low-valent group 13 fragments into the M-X bonds of appropriate precursors can be regarded as a general pathway for the formation of metal complexes featuring a {(DDP)Ga}-M bond. However, the concept of insertion into metal-halogen bonds is not limited to transition metal centers. Insertion reactions into main group metal halide bonds can also be used as a suitable way for the synthesis of stable bonds between main-group metals and group 13 metals. 79

96 III. esults and discussion. 2. Cationic complexes [Lit. 4, Chapter VII.3] 2. Cationic Ga(DDP) complexes by halide abstraction. As seen in chapter III.1., Ga(DDP) readily inserts into metal halide bonds of both transition metals and main-group metals. However, on coordination of Ga(DDP) to a metal center, an increase in electrophilicity of the Ga-center is observed, leading to shortening of the respective Ga- bond distances (1.95 to 2.03 Å) compared to free Ga(DDP) (2.05 Å). Also, in most of the cases, the Ga-Cl bond distances in coordinated {ClGa(DDP)} moieties (2.25 Å 2.29 Å) are significantly elongated compared to those in non-coordinated [(DDP)GaCl 2 ] (2.22 Å) or related compounds. Further supported by the observed fast chlorine exchange in [{(DDP)GaCl}ZnCl(THF) 2 ] (3), these results point to a comparably weak coordination of the chlorine atoms to the Ga center. In this respect, the question arises, if it is possible to remove these weakly coordinated halides and thus synthesize cationic, yet stable complexes coordinated by the bulky bis-imidinate Ga(DDP). A general route for the synthesis of cationic transition metal complexes is the exchange of halides with non-coordinating anions. The compound a[bar F ] with the non-coordinating anion B(Ar F - ) 4 (B(Ar F - ) 4 = tetra-((3,5-bis-triflourmethyl)phenyl)borate) is a prominent example for such halide abstraction reagents. [208] Therefore, [(COE)(η 6 - benzene)h{(ddp)gacl}] (1) and [(Ph 3 P) 2 h(μ-cl){ga(ddp)}] were both reacted with Tl[B(Ar F ) 4 ] or a[b(ar F ) 4 ] in order to generate complexes of the type [h(l) 2 ] + [BF 4 ] - (L = norbornadiene, 1,5-cyclooctadiene), which are known to be suitable catalysts for di- and trimerisation reactions of Bicyclo[2.2.1]heptadiene. [209, 210] In several attempts to abstract the chlorine atoms of both compounds to produce [(COE)h(C 6 H 6 ){Ga(DDP)}] + or [(Ph 3 P) 2 h{ga(ddp)}] +, no products could be isolated unfortunately. However, both compounds [(COE)(η 6 -benzene)h{(ddp)gacl}] (1) and [(Ph 3 P) 2 h(μ-cl){ga(ddp)}] were found to readily react with Tl[B(Ar F ) 4 ] in THF or C 6 H 5 F, as indicated by a significant change in the 1 H M spectra and the formation of insoluble TlCl during the reaction. In both cases, the characteristic splitting of the i Pr-CH-signal of the {(DDP)GaCl}-moiety in the 1 H-M spectrum disappears, indicating the formation of a C 2V symmetric (i.e., halogenfree) Ga(DDP) unit. In addition, the reaction products were found to be soluble only in polar solvents such as THF or fluorobenzene, which is in agreement with an ionic nature of the products. Unfortunately, no crystals suitable for X-ray analysis of either compound could yet be obtained. The resulting species could also not be spectroscopically characterized in more detail since all attempts to purify the obviously very labile products failed and resulted in decomposition. 80

97 III. esults and discussion. 2. Cationic complexes 2.1. Synthesis of the cationic complex [{(DDP)Ga. THF} 2 Au][B(Ar F ) 4 ] (10. 2THF). To further prove the concept, we selected the related Au(I) compound [{(DDP)Ga}Au{ClGa(DDP)}] [80] which was treated in the same manner with a[b(ar F ) 4 ] in C 6 H 5 F as described before. Thus, halide abstraction from the coordinated {GaCl(DDP)} moiety occurred similar to the h-congeners and the cationic compound [{(DDP)Ga} 2 Au][B(Ar F ) 4 ] (10) was obtained (Scheme 25). Additionally, compound 10 cannot be synthesized by a simple one-pot reaction using [(Ph 3 P)AuCl], Ga(DDP) and a[b(ar F ) 4 ], respectively. Instead, the well known cation [(Ph 3 P) 2 Au] +[211] is formed in high yields. The 1 H-M spectrum of 10 shows only one set of signals for a symmetrically (C 2v ) coordinated Ga(DDP) unit (i.e. one signal for the heterocyle-proton at 5.26 ppm, one septet signal for the i Pr-CH and two doublet signals for the i Pr-Me protons, respectively), which is in agreement with a gold atom linearly coordinated by two Ga(DDP) moieties. Crystals, suitable for X-ay analysis of 10 can be obtained by slow diffusion of hexane into a saturated C 6 H 5 F solution of 10 and indeed shows an undistorted, linear [Au{(GaDDP)} 2 ] moiety as the cation. However, the overall structural quality of several crystals measured was poor due to twinning of the crystals and thus a depiction of this structure and discussion of its features is not included herein. Ga Au Ga Cl + a[b(ar F ) 4 ], C 6 H 5 F - acl Ga Au 10 Ga B(Ar F ) 4 + a[b(ar F ) 4 ], THF - acl O Ga Au THF O Ga BAr F + a [ B(Ar F ) 4 ], + 2 Ga(DDP) THF - acl + a[b(ar F ) 4 ], C 6 H 5 F [(Ph 3 P)AuCl] Scheme 25: Synthesis of [{(DDP)Ga} 2 Au][B(Ar F ) 4 ] (10) and 10. 2THF. 81

98 III. esults and discussion. 2. Cationic complexes However, when performing the reaction in a weak coordinating solvent like THF instead of C 6 H 5 F, the solvent adduct [{(DDP)Ga. THF} 2 Au][B(Ar F ) 4 ] (10. 2THF) is formed in high yields (Scheme 25). In contrast to 10, compound 10. 2THF can alternatively be prepared by a simple one-pot reaction using [(Ph 3 P)AuCl], Ga(DDP) and a[b(ar F ) 4 ] in THF, respectively. The 1 H-M spectrum in THF-d 8 shows only one C 2v symmetric signal-set for both coordinated Ga(DDP) ligands, indicating a fast exchange of the coordinated and free solvent molecule THF on the M timescale, comparable to the chlorine atom exchange in 3. However, it seems, that this association / dissociation process generally takes place on dissolving 10. 2THF in organic solvents, because in a 10:1 mixture of C 6 H 5 F/C 6 D 6, the 1 H- M spectrum of 10. 2THF shows only a C 2v symmetric signal-set for the compound, too. This fact indicates that the coordination of the THF molecule to the Ga-center of Ga(DDP) is only a weak interaction. Finally, the resonances at 7.79 ppm and 7.57 ppm in a ratio of 2:1 can be assigned to the [B(Ar F ) 4 ] - anion. The 13 C M spectrum is in good agreement with this result. Figure 17: Molecular structure of [{(DDP)Ga. THF} 2 Au][B(Ar F ) 4 ] (10. 2THF) with hydrogen atoms and the [B(Ar F ) ] 4 -anion omitted for clarity. For structural parameters see Table

99 III. esults and discussion. 2. Cationic complexes The spectroscopic features of the coordinated Ga(DDP) ligand of 10. 2THF match the observation in case of the unstable cationic h complexes, more precisely the disappearance of the characteristic splitting of the i Pr-CH-signal of the {(DDP)GaCl} moiety on reaction with a[b(ar F ) 4 ]. Thus, the weak coordination of a THF molecule to the electrophilic Ga(DDP) unit (vide infra) and a similar association / dissociation process in the cationic h complexes is a reasonable suggestion, even though no detailed spectroscopic analysis was performed so far. Crystals of 10. 2THF, suitable for X-ay analysis, were obtained by slow diffusion of hexane into a saturated THF-solution at room temperature. The molecular structure (Figure 17) of the cation consists of a central gold atom almost linearly coordinated by two Ga(DDP) ligands (Ga1-Au-Ga2 = ). Selected bond length and angles are shown in Table 24. Both Ga- Au distances are equal (Ga-Au: Å) and slightly shorter than the distances found for the parent compound [{(DDP)Ga}Au{ClGa(DDP)}] (2.41 Å), but longer than the distances found for the equatorial GaCp* units in the trimeric cluster [Au 3 (η-gai 2 )(Cp*Ga) 5 ] (Au-Ga: Å and Å). [163] Table 24: Selected bond length (Å) and angles ( ) for [{(DDP)Ga. THF} 2 Au][B(Ar F ) 4 ] (10. 2THF). a) Au1-Ga Ga1-Au1-Ga Au1-Ga Au1-Ga1-O Ga Ga Ga Ga1-Au Ga1-O Ga1-Au Ga Au1-Ga2-O Ga Ga Ga2-O Ga2-Au Ga2-Au a) no esd s available from the cif-file. A very interesting feature of the solid state structure of 10. 2THF is the aforementioned weak coordination of THF to each of the gallium centers giving [(DDP)Ga(THF)] moieties. This is in agreement with an "electrophilization" of the Ga(I) center of Ga(DDP) when coordinated to a transition metal center, since such adducts are unknown for free Ga(DDP). However, the 83

100 III. esults and discussion. 2. Cationic complexes addition of a electron pair donor, like THF, is expected to occur only perpendicular to the plane of the six-membered ring, which is also clarified by theoretical investigations on the model system H 3 and Ga{H(CH) 3 H} by the group of eiher et al., which shows rejection of the H 3 upon in-plane addition and coordination perpendicular to the plane, due to orbital symmetry and occupancy. [53] It should be noted, that a similar observation was recently reported on the adduct [(DDP)Al-B(C 6 F 5 ) 3 ] [125] showing intramolecular Al-F interactions in the solid state. However, the Ga-O distances ( Å) are significantly longer than in the Ga(III)-THF adduct [(THF)GaCl 3 ] (1.91 Å). [212] Similarly elongated Ga-O bond lengths are only known for hypervalent trigonal-pyramidal or octahedral gallium complexes for axially coordinated THF ligands (e.g. GaX 3 (THF) 2 : Ga-O: 2.11 Å (X = Cl); [213] 2.14 Å (X = Br) [214] ). 84

101 III. esults and discussion. 2. Cationic complexes [Lit. 9, Chapter VII.3] 2.2. Cationic Complex with Zn-Ga bonds. The concept of generating cationic complexes by the use of a[b(ar F ) 4 ] as halide abstraction reagent was proven by the synthesis of [{(DDP)Ga. THF} 2 Au][B(Ar F ) 4 ] (10. 2THF), as described in the previous chapter. With the fast exchange of the chlorine atom in solution, the compound [{(DDP)GaCl}ZnCl(THF) 2 ] (3) might be also a suitable precursor for halide abstraction reactions with a[b(ar F ) 4 ]. Thus, when 3 is reacted with a[b(ar F ) 4 ] in THF or fluorobenzene in a ratio of 1:1, no colour change or precipitation was observed. However, on storing the solution overnight at -30 C, colourless needles of [{THF. Ga(DDP)}Zn(THF)(μ-Cl)] 2 [B(Ar F ) 4 ] 2 (11) were obtained in 72 % yield (Scheme 26). It should be noted, that no further reaction was observed on addition of another equivalent of a[b(ar F ) 4 ] to 11. Compound 11 can be stored in an inert gas atmosphere for several days without decomposition and was found to be soluble only in polar solvents such as THF or fluorobenzene, which is in agreement with the ionic nature of the product a[b(ar F ) 4 ] - - acl O Ga Zn O Cl Cl O Zn O Ga 2 [B(Ar F ) 4 ] - Scheme 26: Synthesis of [{THF. Ga(DDP)}Zn(THF)(μ-Cl)] 2 [B(Ar F ) 4 ] 2 (11). As seen in the M spectrum of 10, the 1 H M spectra of 11 in fluorobenzene at room temperature also shows only one set of signals for a C 2V symmetric Ga(DDP) moiety, i.e. one signal at 5.24 ppm for the heterocyclic proton as well as two doublet-signals for the i Pr-Me protons at 1.16 and 1.09 ppm, respectively. This, together with two broad signals for the THF ligand observed at 3.51 and 1.53 ppm, respectively, are in agreement with a fast exchange of the axially coordinated THF molecule or a dissociation / association process of the THF molecule in solution, as discussed for 10. 2THF (see chapter III.2.1). 85

102 III. esults and discussion. 2. Cationic complexes Colourless needles, suitable for single crystal x-ray diffraction analysis were grown by cooling a saturated solution of 11 in fluorobenzene to -30 C overnight. Compound 11 crystallizes in the triclinic space group P-1 with half of the molecule in the asymmetric unit. Figure 18: Molecular structure of [{THF. Ga(DDP)}Zn(THF)(μ-Cl)] 2 [B(Ar F ) 4 ] 2 (11) with the [B(Ar F ) 4 ]-anions omitted for clarity. For structural parameters see Table 25. The molecular structure of 11 is depicted in Figure 18 with selected bond length and angles shown in Table 25. It consists of a planar Zn 2 Cl 2 core with the two Cl atoms bridging the Zn centers. The Zn atoms show a distorted tetrahedral environment. Whereas the Zn-Cl bond distances of 2.35 Å (Zn-Cl) and 2.36 Å (Zn-Cl*) are similar to those found in the halidebridged complexes [{SiMe 3 ) 3 Si}Zn(THF)(μ-Cl)] 2 (2.36 Å and 2.40 Å) or [{SiMe 3 ) 3 Ge}Zn(THF)(μ-Cl)] 2 (2.36 Å and 2.40 Å), [215] the Zn-Ga (2.39 Å) bond distance is similar to those found in 2 (2.38 Å) and therefore in between those found in [(tmeda)zn{ga{[(ar)c(h)] 2 }} 2 ] (2.45 Å and 2.43 Å) and [{ i Pr 2 C[(Ar)] 2 }ZnGa{[(Ar)C(H)] 2 }] (2.32 Å), respectively. [61] Similar to the formation of [{(DDP)Ga. THF} 2 Au][BAr F ] (10. 2THF), an electrophilasation of the Ga center takes place on coordination to the metal center, which is reflected in the weak coordination of a THF molecule to each Ga atom perpendicular to the plane of the heterocycle. 86

103 III. esults and discussion. 2. Cationic complexes Table 25: Selected bond length (Å) and angles ( ) for [{THF. Ga(DDP)}Zn(THF)(μ- Cl)] 2 [B(Ar F ) 4 ] 2 (11). Ga-Zn (11) 1-Ga (2) Ga (5) O1-Ga-Zn (13) Ga (5) Cl1-Zn-Cl1* 92.96(6) Ga-O (4) O2-Zn-Ga (12) Zn-O (4) O1-Ga-Zn-O (18) Zn-Cl (16) Zn-Cl1* (18) This behaviour to act as a Lewis acid if a Lewis base approaches from above the ring plane again corresponds well with the theoretical findings of eiher et. al. [53] Thus, the Ga-O bond distances in 11 (2.10 Å) are slightly shorter than in (10. 2THF) (2.16Å). The same is true for the Zn-O bond distance (2.02 Å), which is slightly shorter than in 3 (2.10 Å), but comparable to [(THF) 2 ZnCl 2 ] (2.00 Å). [180] Additionally, the THF moieties coordinated to Zn and Ga, respectively, are perpendicularly oriented to each other (torsion angle O-Ga-Zn-O: ). 87

104 III. esults and discussion. 2. Sumary and conclusion Summary and conclusion. The abstraction of halide ligands in transition metal complexes coordinated by the bulky bisimidinate Ga(DDP) is achieved by the use of the halide abstracting reagent a[bar F ]. Whereas the abstraction of the halides in the h complexes [(Ph 3 P) 2 h(μ-cl){ga(ddp)}] and [(coe)h(c 6 D 6 ){ClGa(DDP)}] (1) can be observed when monitoring the reaction with a[b(ar F ) 4 ] in the 1 H-M spectrum, the resulting compounds could not yet be stabilized. However, reaction of a[b(ar F ) 4 ] with the Au(I) complex [{(DDP)Ga}Au{ClGa(DDP)}] leads to the linear, symmetric cationic complex [{(DDP)Ga} 2 Au][ B(Ar F ) 4 ] (10). The rather strong electrophilicity of the coordinated gallium center in the cation 10 becomes visible by crystallization of the product from a saturated solution in THF, leading to a molecular structure of 10. 2THF showing THF molecules axially coordinated to each gallium center. Similarly, chloride abstraction from [{(DDP)GaCl}ZnCl(THF) 2 ] (3) by a[b(ar F ) 4 ] yields [{THF. Ga(DDP)}Zn(THF)(μ-Cl)] 2 [B(Ar F ) 4 ] 2 (11). Again, the axial coordination of THF molecules to each gallium center shows the increased electrophilicity of the Ga center on abstraction of the halogen atom. 88

105 III. esults and discussion. 3. Substitution of labile olefins at i(0) centers 3. Substitution reactions of labile olefins with E(DDP). The entry into the chemistry of potentially reactive, coordinatively unsaturated [M{E(DDP)} n ] complexes or intermediate fragments is based on the substitution of labile olefin ligands. [43, 216] More strongly bound CO ligands of carbonylmetal fragments [L(CO) n M] cannot be fully displaced. [22, 131, 132, 217] The presence of alkyl or halide groups at transition metal centers of medium or higher oxidation state may create a problem due to the strongly reducing power of the group-13 element(i) species, for example leading to the insertion of the E(I) center into M-C and M-X bonds (X = halide, etc.), as seen in chapter II and III.1. However, side reactions of the displaced olefin ligand with the group-13 metalloid ligand must be taken into account as well and need to be avoided by choosing suitable reaction conditions. For example AlCp* reacts with dvds to yield the compound [(dvds)(μ 2 -Cp*Al) 2 ], which can be described as a two fold coordinated adduct of Cp*Al to the two C=C double bonds of the dvds ligand, exhibiting an adamantyl-like structure. [218] Even though such reactions are not reported for the six-membered heterocycles so far, Al(DDP) readily reacts with alkynes to give the aluminacyclopropene [(DDP)Al(η 2 -C 2 H 2 )] at -78 C in high yields. [74] However, his results emphasize the carbenoid character of these kind of ligands Substitution reactions on i(0) centers eactions of E(DDP) with [i(cod) 2 ]. The recently reported activation of a benzene molecule at the reactive 16 VE fragment [i(alcp*) 3 ], [56] synthesized in situ by the reaction of [i(cod) 2 ] and AlCp* in benzene poses the question of whether such reactive intermediates can be stabilized and analysed in more detail. With success, a deeper understanding of the relevant steps in such activation reactions could arise and thus, more applicable catalysts might be accessible. Several strategies including matrix isolation techniques and in-situ spectroscopy have been applied to achieve this goal. Another approach could be the kinetic stabilization of such reactive fragments by the use of bulky ligands. ecent results on the coordination chemistry of Ga(DDP) at the h I center in [h(pph 3 )(μ 2 -Cl){Ga(DDP)}] have shown, that these class of bulky ligands is able to freeze the intermediates in e.g. insertion reactions. [81] Therefore, the reaction with [i(cod) 2 ] holds a lot of promise to get a deeper insight into activation reactions similar to those described above. 89

106 III. esults and discussion eaction of E(DDP) with [i(cod) 2 ] eaction of [i(cod) 2 ] with Ga(DDP). When reacting [i(cod) 2 ] with Ga(DDP) in hexane, the yellow solution turns into deep red after several minutes. Obviously, the cod-ligand is replaced by the Ga(DDP) moiety to form a i-ga complex. Unfortunately, all attempts to grow single crystals from the saturated solution only once were successful. [i(cod) 2 ] + 2 Ga(DDP) Ga i Ga Scheme 27: eaction of [i(cod) 2 ] with Ga(DDP). However, the crystal quality and therefore the refined structure were too poor, so no meaningful discussion can be presented herein. evertheless, we could identify the formed compound as [(1,3-cod)i{Ga(DDP)} 2 ] (Scheme 27), the nickel-analogue of [(1,3-cod)Pt{Ga(DDP)} 2 ] (vide supra), [159] in which one cod-ligand is replaced by two Ga(DDP) moieties. The remaining cod-ligand is coordinated in a η 2 -mode, also undergoing a isomerisation process to give the 1,3-cod adduct. The 1 H-M spectrum of this compound shows the same broad and overlapping signals as reported for the Pt-compound. As mentioned before, the crystallisation of this compound is very difficult. In several experiments with different reaction conditions (lower or ambient temperature, different solvents, different reaction times), a complex interplay between formation of the complex and decomposition was identified to take place, also at ambient temperature. Therefore, the use of the more reactive compound [i(cdt)], which easily undergoes substitution reactions with olefins or phosphines, as an alternative i(0) precursor with respect to i(cod) 2 was employed (see chapter III ). 90

107 III. esults and discussion eaction of E(DDP) with [i(cod) 2 ] eaction of [i(cod) 2 ] with Al(DDP). In contrast to the reaction of Ga(DDP), the reaction of [i(cod) 2 ] with Al(DDP) is far more interesting. When reacting two equivalents Al(DDP) with one equivalent [i(cod) 2 ] in C 6 D 6 and monitoring the reaction in the 1 H-M spectrum, signals for free [i(cod) 2 ] can be observed at 4.29 ppm and 2.07 ppm, indicating that no complex formation takes place. However, after heating the reaction mixture several minutes to ca. 60 C, the M spectrum of Al(DDP) changes. It seems that an oxidation of the Al(DDP) moiety occurs, indicated by the shift of the DDP-backbone proton from 5.18 ppm (free Al(DDP)) to 4.93 ppm. Similar high-field shifts are known for compounds of the type [(DDP)AlX 2 ] (X = Cl (4.91 ppm); H 2 (4.88 ppm) and related), which indicates an oxidation of the Al(I) center to Al(III). However, the i Pr-CH signal (3.16 ppm in free Al(DDP)) splits into two distinct septet signals at 3.59 ppm and 3.14 ppm, accompanied by the significant splitting of the two i Pr-Me doublet signals in free Al(DDP) (1.38 ppm and 1.13 ppm) into four doublet signals at 1.42 ppm, 1.17 ppm, 1.08 ppm and 0.59 ppm, respectively This unusual splitting of the signals is known for an unsymmetric coordinated Al center, as seen in e.g. [(DDP)Al(Cl)(I)]. [154] Thus, these results are in accordance with an oxidative addition of an unsymmetric compound 1-2 to the Al(I) center in Al(DDP), giving a [(DDP)Al( 1 )( 2 )] ( 1 2 ) moiety. Al Benzol Al Benzol, i(cod) 2 Al H Ph Scheme 28: eaction of Al(DDP) with benzene in the presence of [i(cod) 2 ]. Taking into account that, besides [i(cod) 2 ], which apparently stays intact during the reaction, only the solvent C 6 D 6 can be considered as reaction partner, it can be speculated that the C-D bond of C 6 D 6 is activated and the compound [(DDP)Al(C 6 D 5 )(D)] is formed (Scheme 28). Unfortunately, all attempts to grow single crystals from a saturated solution failed so far, 91

108 III. esults and discussion eaction of E(DDP) with [i(cod) 2 ] therefore no single crystal x-ray diffraction analysis could be obtained and the prediction of the product in this reaction is mainly based on the M-spectroscopic data provided above. Interestingly, when dissolving Al(DDP) in C 6 D 6 without additionally [i(cod) 2 ], no such reaction occurs, as seen in the reported M-spectra of Al(DDP) by Cui et al. [30] Obviously, the [i(cod) 2 ] catalyzes the oxidative addition of the benzene molecule. To prove this assumption, we dissolved ca. 100 eq. Al(DDP) in C 6 D 6 and added only 1 eq. [i(cod) 2 ]. After heating the reaction mixture for 10 minutes to 60 C, the M spectrum changes: besides signals for free Al(DDP) and free i(cod) 2 (only very small signals), the previously discussed signals for the suggested oxidative addition product [(DDP)Al(C 6 D 5 )(D)] can be detected. After 24h of heating to 60 C, the signals for free Al(DDP) are vanished. This fact indicates, that the oxidative addition of the C 6 D 6 occurs catalytically. A similar activation reaction, which occur on a i(0) fragment, has been previously reported on the reaction of Al(Cp*) with [i(cod) 2 ] in benzene. The activation proceeds via the hypothetical intermediate [i{al(cp*)} 3 ], which activates the C-H bond of a benzene molecule to give [i{al(cp*)} 3 {(Cp*)Al(H)(Ph)}] (Scheme 29) as the final product, which is described as side-on coordination of the electrophilic fragment {(Cp*)Al(H)(Ph)} to the electron-rich [i(alcp*) 3 ] fragment. AlCp* [i(cod) 2 ] 3/4 [(AlCp*) 4 ] -2 cod [i(alcp*) 3 ] 1/4 [AlCp*) 4 ] Benzol H *CpAl i AlCp* AlCp* Scheme 29: Activation of benzene at the reactive fragment [i(alcp*) 3 ]. Thus, the formation of [(DDP)Al(Ph)(D)] (Scheme 30) can be suggested to proceed by a similar reaction pathway: The first step of the catalytic process may be the elimination of a cod ligand of [i(cod) 2 ] by addition of one Al(DDP) fragment which yields the threefold 92

109 III. esults and discussion eaction of E(DDP) with [i(cod) 2 ] coordinated i(0) center [(η 4 -cod)i{al(ddp)}], similar to [( 3 P)i(η 4 -cod)], [219] which was calculated to be an intermediate in the catalytic cyclodimerisation of butadiene. (DDP) Al Ph Al(DDP) H i(cod) 2 - (DDP) Al i (DDP)Al i Ph H (DDP)Al Ph i H Scheme 30: proposed mechanism for the formation of [(DDP)Al(Ph)(H)]. ext, the oxidative addition of a benzene-molecule at the reactive i(0)-center can be assumed, giving the tetrahedrally coordinated i(ii) intermediate [{(DDP)Al}i(H)(Ph)(η 2 -cod)], in which the cod-ligand is bonded only in a η 2 -coordination mode. Migration of the phenyl rest from the i-center to the Al-center and recoordination of the second cod-double bond possibly gives the intermediate [{i(η 4 -cod)(h)}{al(ddp)(ph)}]. Finally, reductive elimination of [(DDP)Al(Ph)(H)] and re-coordination of the second cod-ligand yields [i(cod) 2 ], which is reintegrated into the catalytic cycle. Unfortunately, because of the very acidic Al-center, the intermediates [(η 4 -cod)i{al(ddp)}] and [{(DDP)Al}i(H)(Ph)(η 2 -cod)] are postulated to be very reactive and could not be isolated so far. It should be again mentioned, that the described catalytic cycle is mainly speculative and a more detailed investigatuion in both experimental and theorethical manner is necessary to elucidate the course of the reaction. 93

110 III. esults and discussion eaction of E(DDP) with [i(cod) 2 ] Additionally, a similar Si-H activation, which is reported for the reaction of AlCp*, [i(cod) 2 ] and HSiEt 3 in hexane leading to the formation of [i(alcp*) 3 (H)(SiEt 3 )], can not be observed for the Al(DDP) congener. Thus, decomposition occurs on mixing [i(cod) 2 ], Al(DDP) and HSiEt 3, giving a variety of untraceable species, as shown by 1 H-M spectroscopy. 94

111 III. esults and discussion eaction of E(DDP) with [i(cdt)] [Lit. 8, Chapter VII.3] Adduct formation in the reaction of [i(cdt)] with Ga(DDP). More than 35 years ago, the first i(0) complex i(cdt) was synthesized by Wilke et al. [ ]. Due to an unoccupied coordination site, it exhibits a very reactive 16VE i center. This high reactivity was used in several substitution reactions, in which the cdt-ligand was replaced by several other ligands like olefins or phosphines. The i(0) center in [i(cdt)] is very Lewis-acidic, thus the coordination of donor ligands like CO, C or P 3 are readily observed, giving a distorted tetrahedral coordinated i(l 4 )-center. [221, 223, 224] In this respect, the reaction with the Lewis basic Ga(DDP) seems to be very promising in terms of a simple Lewis acid-base adduct or, due to the high reactivity of [i(cdt)], in cluster formation. The reaction of Ga(DDP) with an equimolar amount of [i(cdt)] yields the 18 VE complex [(cdt)i{ga(ddp)}] (12) as deep red crystals in 88% yield. Analogously to [(cdt)i(pme 2 )] ( = Me, Menthyl), [223, 224] 12 represents a simple coordination compound of a 2e donor ligand to the unsaturated [i(cdt)] fragment, where the cdt ligand is still bound in a η 6 -coordination mode (Scheme 31). Thus, 12 can be regarded as a transition metal-group 13 metal Lewis acid-base adduct [(DDP)Ga i(cdt)], similar to the previously described [(DDP)Ga B(C 5 F 6 ) 3 ]. [123] Compound 12 is air and moisture sensitive in solution as well as in solid state, but can be stored at room temperature for several weeks in an inert gas atmosphere. It readily dissolves in non polar organic solvents like hexane or benzene. i i(cdt) + Ga(DDP) Ga 12 Scheme 31: Synthesis of [(cdt)i{ga(ddp)}] (12). 95

112 III. esults and discussion eaction of E(DDP) with [i(cdt)] The M spectrum in benzene at room temperature shows one set of signals for the Ga(DDP) ligand, which is only slightly different from free Ga(DDP). Interestingly, four signals are found for the cdt ring, indicating the coordination of the Ga(DDP) moiety to the i(cdt)- fragment. This splitting of the cdt-hydrogen atoms can be best explained by two different conformation modes for the ethylene bridges (two CH 2 -CH 2 -bridges are in syn- (8H, 1.99 ppm) and one bridge is in anti-conformation (4H, 2.15 ppm) with respect to the Ga(DDP) ligand, see Chart 8). Similar spectra with splitting of the cdt signals were reported for the phosphine analogue [(cdt)i(pme 3 )]. [224] The 13 C-M spectrum is in accordance with this result, showing one set of signals for the Ga(DDP), one signal for the C=C double bonds and two distinct signals for the CH 2 groups of the bridges, respectively. Anti Syn i Ga (DDP) Syn Chart 8: syn- and anti-conformation in the cdt-ring of 12. Crystals, suitable for single crystal x-ray analysis could be obtained by cooling a saturated solution of 12 in hexane to -30 C overnight. A mercury plot of the molecular structure of 12 is depicted in Figure 19. Compound 12 crystallizes in the triclinic space group P 1. The coordination geometry around the i center is nearly tetrahedral with angles in the range of to (the C=C centroids are taken for the calculations of the angles in the case of cdt). Selected bond lengths and angles are shown in Table

113 III. esults and discussion eaction of E(DDP) with [i(cdt)] Figure 19: Molecular structure of [(cdt)i{ga(ddp)}] (12). For structural parameters see Table 26. The i-c bond distances of 2.01 Å to 2.20 Å are significantly longer than in [i(cdt)] (av Å), [225] which is as expected because of the distorted tetrahedral coordination of the icenter in 12 compared to a trigonal planar arrangement in the parent compound. Similar i-c bond distances can be found in the phosphine-compounds [(cdt)i(pme 2 )] (' = menthyl), [226] also exhibiting a tetrahedral i-center. The C=C bond length in 12 are significantly longer than in [(cdt)i(pme 2 )] (1.28 Å to 1.33 Å), possibly an effect of the stronger electron donating ability of Ga(DDP) compared to phosphines, as already discussed in the case of [(dvds)pd{ga(ddp)}]. [159] Interestingly, they are very similar to those bare [i(cdt)] (1.33 to 1.39 Å in 12 vs to 1.38 in [i(cdt)]), which can be explained by a better overlap of the metal and olefin π-orbitals in the case of a trigonal planar coordinated i-center (in [i(cdt)]) compared to the tetrahedral one in

114 III. esults and discussion eaction of E(DDP) with [i(cdt)] Table 26: Selected bond lengths (Å) and angles ( ) for [(cdt)i{ga(ddp)}] (12). Ga-i (6) C43-C (3) Ga (3) C47-C (3) Ga (3) C51-C (14) i-c (6) i-c (10) 1-Ga (10) i-c (13) C40-C51-C (9) i-c (11) C43-C44-C (16) i-c (12) C47-C48-C49 129(2) i-c (8) The i-ga distance of 2.35 Å is significantly longer than in [i(gacp*) 4 ] (2.218 Å) [136] or [i(gacsime 3 ) 4 ] (2.17 Å) [83] and can be attributed to the high steric demand of the DDPmoiety. Similarly long i-ga bond distances can only be found in the compound [i{ga[(ar)c(h)] 2 }) 2 {C[(Me)C(Me) 2 ] 2 } 2 ] (Ar = C 6 H 3 Pr i 2-2,6) (2.32 Å), [164] where the square planar i(ii) center shows trans-coordination of the two formally anionic Ga-moieties. Interestingly, the Ga- distances in 12 of and Å lie in the same range as observed for free Ga(DDP) (2.053 and 2.056Å), [31] and therefore are significantly longer than in other compounds with coordinated Ga(DDP) (e.g Å in [{(DDP)Ga}Au{Cl-Ga(DDP)}] [80] or 1.99 Å in [{Ga(DDP)}Pd(dvds)] [159] ). This fact indicates, that upon coordination of Ga(DDP) to [i(cdt)], the electrophilicity of the Ga(I) center is much less increased than observed for the cited examples. 98

115 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] [Lit. 8, Chapter VII.3] Substitution reactions of [(cdt)i{ga(ddp)}] with ethylene. Poerschke, Wilke and Krueger reported in the late 1980 s several reactions in which the cdtligand in [i(cdt)] was readily replaced by different ligands like CO, alkyns, olefins or phosphines. [226] The most prominent example might be the substitution of the cdt-ligand by three ethylene molecules to give [i(c 2 H 4 ) 3 ], [227] which exhibits a trigonal-planar arrangement of the olefins surrounding the i-center. Also, the Lewis acid-base adducts [(cdt)il] (L = e.g. phosphines), were used in this respect. Thus, the compound [(cdt)i{ga(ddp)}] (12) also provides an attraction for substitution reactions because the icenter is distorted tetrahedrally coordinated, which decreases the overlap of the relevant orbitals and hence increases the reactivity towards small molecules. In order to elucidate the reactivity of [(cdt)i{ga(ddp)}] (12), we tried to react this compound with ethylene. Therefore, we mixed [i(cdt)] and Ga(DDP) in the ratio 1:1 in hexane, strirred the solution for 30 minutes to prepare [(cdt)i{ga(ddp)}] in-situ, and finally bubbled ethylene through or applied increased ethylene pressure (2-3 bar) to this solution for 20 minutes. After storing the solution at -30 C in ethylene atmosphere overnight, deep red single crystals of [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13) was isolated in yields around 40% (Scheme 32). Compound 13 can alternatively be prepared by substitution of one ethylene ligand in [i(c 2 H 4 ) 3 ] by Ga(DDP). i Ga + -cdt i Ga + Ga(DDP) - i Scheme 32: Synthesis of [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13). Thus, treatment of a deep red solution of [i(cdt)] in hexane at -10 C with ethylene leads to a colour change to yellowish-brown, indicating the formation of [i(c 2 H 4 ) 3 ] in situ. After addition of a cooled solution of Ga(DDP) (one eq. with respect to i(cdt)), insoluble 99

116 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] impurities are filtered off. Cooling the mixture to -30 C affords deep red single crystals of [{(DDP)Ga}i(C 2 H 4 ) 2 ] 13, which can be isolated in similar yields compared to the method described above (Scheme 32). Additionally, the reaction of [i(cod) 2 ] with Ga(DDP) in the presence of ethylene did not give the desired product 13, instead the formation of several, undefined products can be observed. As known for [i(c 2 H 4 ) 3 ] and [Pt(C 2 H 4 ) 3 ], these compounds are very reactive and thermally unstable, possibly due to the facile liberation of the coordinated ethylene. Therefore, compound 13 is also expected to be very reactive. Indeed, compound 13 is very air, moisture and temperature sensitive and decomposes also in an inert gas atmosphere at -30 C after several days. The 1 H-M spectrum of 13 shows the expected signal set of a C 2v symmetric Ga(DDP) ligand (s, 5.20 ppm; sep, 3.11 ppm; s, 1.68 ppm; d, 1.21 ppm; d, 1.09 ppm) as well as one signal for coordinated ethylene (2.29 ppm). Also the 13 C-M spectrum is in good agreement with the solid state structure. Figure 20: Molecular structure of [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13). For structural parameters see Table 27. As expected for a d 10 -metal compound of the type M(C 2 4 ) 3 (M = i 0, Pt 0, Pd 0 ; = H, Alkyl ), which generally exhibit a coplanar arrangement of the ligands (extended VSEPrules), the crystal structure of 13 shows an almost ideally trigonal planar coordinated i center with angles close to 120 (Ga-i-C30/C31 centr , C30/C31 centr. -i- C32/C33 centr. 100

117 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] and C32/C33 centr. -i-ga ). For selected bond lengths and angles see Table 27. The DDP-backbone is slightly twisted with a torsion angle of The i-ga bond distance in 13 of 2.28 Å is again longer than in [i(gacp*) 4 ] (2.22 Å), [136] but shorter than in 10 (2.35 Å). Table 27: Selected bond lengths (Å) and angles ( ) for [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13). Ga1-i (6) 1-Ga (10) Ga (2) C30-i1-Ga (9) i1-c (3) C30-i1-C (12) i1-c (3) C31-i1-C (14) i1-c (3) C32-i1-C (13) i1-c (3) C33-i1-Ga (10) C30-C (4) C32-C (5) The rather strong donor capabilities of Ga(DDP) are reflected in the C=C double bond lengths in 13 (1.42 and 1.40 Å), which are similar to those found in [(Cy 3 P)i(C 2 H 4 ) 2 ] (1.41 and 1.39 Å) [228] or the Fischer carbene complex [(C 2 H 4 ) 2 i{cph( 2 )}] (1.40 and 1.39 Å) [229]. The Ga- bond distances in 13 of 1.99 Å are significantly shorter than in 12 (av Å) or free Ga(DDP) (av Å), which again can be used as an indication for an increase in electrophilicity of the Ga-center. 101

118 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] [Lit. 8, Chapter VII.3] eaction of [(cdt)i{ga(ddp)}] with other olefins. As seen in the previous chapter, the reaction of [i(cdt)] with ethylene affords [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13) in moderate yields. However, compound 13 was found to be quiet unstable at room temperature and also at -30 C in inert gas atmosphere. This possibly results from the liberation of ethylene during decomposition, which is an entropically favored effect. Therefore, the substitution of cdt with other, non gaseous olefins might provide the synthesis of a more stable compound eaction of [(cdt)i{ga(ddp)}] with styrene. According to the synthesis of 13, the reaction of 12 (in situ prepared by mixing i(cdt) and Ga(DDP) in hexane) with an excess styrene at room temperature yields the substitution of the cdt-ligand by two styrene molecules to give [{(DDP)Ga}i(styrene) 2 ] (14) after workup and recrystallisation as orange-yellow crystals in 71% yield. In contrast to the formation of 13, compound 14 can also be prepared by the use of [i(cod) 2 ] as an alternative i(0) precursor (Scheme 33). Thus the reaction of [i(cod) 2 ], Ga(DDP) and an excess of styrene also afforded 14, but the overall yield was lower (41%). Compound 14 is soluble in common organic solvents like hexane and benzene and, in contrast to 13, stable even for short periods in air. i Ga + 2 styrene -cdt Ga i + 2 styrene + Ga(DDP) - 2 cod [i(cod) 2 ] Scheme 33: Synthesis of [{DDP)Ga}i(styrene) 2 ] (14). The 1 H-M spectra of 14 in benzene at room temperature shows a singlet for the γ-ch proton (at 5.18 ppm) and the Me-group (1.67 ppm) of the DDP backbone, as well as the typical septett for the i Pr-group at the aromatic rings of the DDP-ligand. Interestingly, a 102

119 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] splitting of the signals for the i Pr-Me-protons can be observed, showing four doublet signals at 1.30, 1.09, 1.07 and 1.04 ppm, respectively. This indicates an unsymmetric coordination of the styrene molecules in 14. Furthermore, three different signals for the styrene-olefins can be observed: The signal at 3.39 ppm can be assigned to the vicinal hydrogen, whereas the geminal hydrogen atoms give rise to signals at 2.48 and 2.25 ppm with a geminal coupling constant of 2 J(gem) = 2.7 Hz. Whereas the signal at 2.48 ppm can be assigned to the transhydrogens ( 3 J(E) = Hz), the cis-hydrogen (2.25 ppm) show a coupling constant of 3 J(Z) = 9.25 Hz. The 13 C-M spectrum shows the expected signals for 14. Figure 21: Molecular structure of [{(DDP)Ga}i(styrene) 2 ] (14). For structural parameters see Table 28. A Mercury plot of the molecular structure of 14 is shown in Figure 21, with selected bond lengths and angles shown in Table 28. Compound 14 consists of a central i-center which is surrounded by one Ga(DDP) ligand and the H 2 CCH-moieties of two styrene molecules in a nearly ideal trigonal planar fashion (angular sum: 362,3 ). Hereby, the aromatic rings of the styrene molecules are located trans to each other, facing towards opposite sides of the trigonal 103

120 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] planar arrangement. Due to steric interactions of the i Pr-Me groups of the DDP-backbone and the styrene molecules, an increase of the torsion angle of the DDP-backbone and the olefinic carbon atoms by ca. 29 compared to 13 can be observed (torsion angle in 14: 46 ). The C=C double bond distances of 1.44 Å and 1.45 Å are only slightly longer than in the related ethylene compound 13 (1.42 and 1.40 Å) or in [(Et 3 P) 2 i(η 2 -CH 2 CHC 6 F 5 )] (1.42 Å), [230] which is, to best to our knowledge, the only crystallographically analyzed nickel compound with a coordinated styrene molecule reported so far. Table 28: Selected bond lengths (Å) and angles ( ) for [{(DDP)Ga}i(styrene) 2 ] (14). Ga1-i (2) C30-C (11) Ga (7) C40-C (11), Ga (7) i1-ga (3) i1-c (8) Ga1-i1-C30C31 centroid i1-c (9) Ga1-i1-C40C41 centroid i1-c (8) C30C31 centroid -i1- C40C41 centroid i1-c (8) 11-Ga1-i1-C (4) The Ga- bond distances lie within the previously reported range (1.98 and 1.99 Å). The Ga- i bond distance of 2.28 Å is longer than in the homoleptic compound i(gacp*) 4 (2.22 Å), [136] but similar to the ethylene-compound 13 (2.28 Å). 104

121 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] [Lit. 8, Chapter VII.3] eaction of [(cdt)i{ga(ddp)}] with dvds. In accordance to the formation of 13 and 14, the compound [{(DDP)Ga}i(dvds)] (15) is easily accesible by substituting the cdt-ligand in 12 with the bidentate ligand dvds (1,3-divinyl-(1,1,3,3-tetramethyl)disiloxane). Thus, immediately after addition of an excess of dvds to a solution of 12 or i(cdt) and Ga(DDP) (in order to prepare 12 in situ) in hexane at room temperature, the deep red solution turns to bright yellow. After cooling the reaction mixture to -30 C overnight, yellow needles of [{(DDP)Ga}i(dvds)] (15) are obtained in 65% yield. Similar to 14, complex 15 is stable at room temperature in an inert gas atmosphere for several weeks and also with short contact of air, decomposition occurs primal after several minutes. Comparing with the previously reported Pd-analogues [{(DDP)Ga}Pd(dvds)] [159], the 1 H- M spectrum of 15 is quite similar: a typical set of signals for a C s symmetric, coordinated Ga(DDP) moiety (i.e. a singulet signal at 5.12 ppm for the backbone-proton, a septet signal for the i Pr-CH (3.06 ppm) and two doublet signals (1.21 ppm and 1.06 ppm) for the i Pr-CMe protons), two distinct signals for the SiMe 2 groups (0.50 ppm and ppm) as well as signals for the coordinated C=C double bonds (complex spin system at ppm) of the dvds-ligand can be observed. The 13 C-M spectrum also shows the expected signals. i Ga + dvds -cdt Ga i Si Si O Scheme 34: Synthesis of [{(DDP)Ga}i(dvds)] (15). Crystals, suitable for single crystal x-ray analysis, could be obtained by cooling a saturated solution of 15 to -30 C overnight. Complex 15 crystallizes in the triclinic space group P 1 with two molecules in the unit cell. Selected bond lengths and angles can be found in Table 29. Its molecular structure (Figure 22) exhibits a coplanar arrangement of the vinyl groups of 105

122 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] the dvds ligand and the i and Ga atoms (angular sum: C=C centroid, i, Ga: ). According to 13, the steric interaction of the Ga(DDP)-ligand and the dvds ligand forces the DDP- Backbone out of plane. Therefore, the torsion angle between the DDP-backbone and the olefinic carbon atoms can be measured to be 47.4, similar to that in 13 or in [{(DDP)Ga}Pd(dvds)] (torsion angle: 46.5 ). Figure 22: Molecular structure of [{(DDP)Ga}i(dvds)] (15). For structural parameters see Table 29. The i-c distances in 15 ( Å) are considerably longer than the respective distances in [( 3 P)i(dvds)] ( Å) ( = Ccyclohexyl C 6 H 11, Ph). [231] The same is true for the C=C double bond distances (1.40 Å and 1.43 Å vs. av. C=C bond distances of 1.39 Å in [( 3 P)i(dvds)]), which points to a distinctly higher degree of π-backbonding to the olefins in the Ga(DDP) complex than in its phosphane analogue. Again, the Ga- bond distances lay within the reported range (1.97 and 1.99 Å) for coordinated Ga(DDP) moieties. 106

123 III. esults and discussion Substitution reactions on [(cdt)i{ga(ddp)}] As reported for the analogues Pd-complex [(dvds)pd{ga(ddp)}], reaction with strong π-acceptor ligands like CO or t BuC occurs giving the dimeric complexes [Pd{μ 2 -Ga(DDP)}(L)] (L = CO, t BuC), respectively. [159] Due to the great resemblance of 15 with this complex, a similar reaction can be assumed on treatment of [{(DDP)Ga}i(dvds)] (15) with t BuC. However, when monitoring the reaction in the 1 H-M spectrum, a significant change in the spectrum occurs, but giving several new signals of untraceable species. Also, no crystals of a reaction product could be obtained so far. Likewise, the same is true for the reaction of 15 with CO under different reaction conditions (low or ambient temperature, different solvents). Table 29: Selected bond lengths (Å) and angles ( ) for [{(DDP)Ga}i(dvds)] (15). Ga1-i (17) 11-Ga (3) Ga (6) C30-i1-C (3) Ga (6) C30-i1-C (4) i1-c (9) C31-i1-C (3) i1-c (8) C30-i1-C (4) i1-c (8) C31-i1-C (3) i1-c (8) C36-i1-C (3) C30-C (10) 12-Ga1-i1-C (4) C36-C (10) 12-Ga1-i1-C (4) Si11-O (6) Si12-O (6) 107

124 III. esults and discussion Summary and conclusion Summary and conclusion. As mentioned above, the substitution of labile bond olefin ligands is another way to establish TM-E I bonds. In this respect, [i(cod) 2 ] was reacted with both Ga(DDP) and Al(DDP). Whereas the compound [(1,3-cod)i{Ga(DDP)} 2 ] is formed by the reaction of [i(cod) 2 ] with Ga(DDP), Al(DDP) is by far more reactive. Thus, an catalytic activation of benzene occurs on mixing [i(cod) 2 ] and Al(DDP) in benzene, with the proposed oxidative addition product [(DDP)Al(Ph)(H)] being the only traceable species. However, in both cases, the respective i-complexes are unstable at ambient temperature and decompose ([(1,3-cod)i{Ga(DDP)} 2 ]) or, in the case of Al(DDP), intermediates of the activation could not yet be stabilized. i i(cdt) + Ga(DDP) Ga dvds i Ga + Styrene Ga i Si Si O 13 Ga i Scheme 35: Substitution of the cdt ligand in [(cdt)i{ga(ddp)}] (12) by different olefins. In contrast to the reaction with [i(cod) 2 ] no substitution of the olefin is observed in the reaction of [i(cdt)] with Ga(DDP). Instead, the well known addition of a σ-donor ligand to tha i(0) complex takes place giving the 18VE Lewis acid base adduct [(cdt)i{ga(ddp)}] (12). 108

125 III. esults and discussion Summary and conclusion As expected, the cdt-ligand in 12 can easily be replaced by other olefins. However, in contrast to the related chemistry of GaCp*, the high steric demand of the DDP moiety prevents the formation of kinetically inert homoleptic i/ga complexes or clusters of the type [i(gacp*) 4 ] and thus allows the coordination of additional olefin ligands. Complexes of the type [i{ga(ddp]}(olefin) 2 ] (where "olefin" represents two monodentate (C 2 H 4 (13), styrene (14)) or one chelating diolefin ligand (dvds (15))), are thermodynamically rather stable species with the exception of [i{ga(ddp]}(c 2 H 4 ) 2 ] (13) which decomposes at room temperature in the absence of ethylene. Comparable to the well known [L 2 M(alkene)] complexes (L = phosphane or phosphite, M = i, Pd, Pt), the alkene ligands and the Ga atom are trigonal planar coordinated, which can be explained by the more efficient overlap of the HOMO orbital (d xz orbital of the ML 2 fragment) with the π* orbital of the olefin on in-plane coordination. The increase in stability from C 2 H 4 (immediate decomposition in the absence of ethylene) < styrene (decomposition after short contact with air) < dvds (stable for several minutes in air) can be explained by entropic reasons (in the case of ethylene) as well as the chelating effect of the dvds-ligand. However, comparison with related HC and phosphane complexes show, that Ga(DDP) is not only isolobal to this class of well studied ligands, but also offers a great resemblance in electronic properties to them. 109

126 III. esults and discussion monomeric Pd(0)-complexes [Lit. 7, Chapter VII.3] 3.2. monomeric Pd(0)-complexes Substitution reactions on a Pd 0 -center. eaction of [Pd 2 (dvds) 3 ] with E(DDP). The substitution of labile olefins by low valent group 13 organyls is not restricted to i(0) centers. [Pd 2 (dvds) 3 ] (dvds = 1,3-divinyl-1,1,3,3-tetramethyldisiloxane) was also used as a d 10 metal source in this respect. It has been shown recently, that the terminal as well as the bridging dvds ligands are replaced by the group 13 organyl GaCp* and the homoleptic cluster compounds [Pd 2 (GaCp*) 5 ] [55] and [Pd 3 (GaCp*) 8 ] [144] can be synthesized in high yields, depending on the M:Ga ratio and the reaction conditions. This example demonstrates the subtle kinetic control of cluster formation in such reactions. Also, the group 13 metal influences the constitution of the formed clusters: Using the stronger σ-donor ligand AlCp*, the cluster [Pd 3 (AlCp*) 2 (μ 2 -AlCp*) 2 (μ 3 -AlCp*) 2 ] [78] can be obtained, which features a triangular arrangement of the Pd-centers as well as a smaller M:E ratio of 3:6 compared to the linear compound [Pd 3 (GaCp*) 8 ] (3:8). This result, accompanied by the synthesis of the byproduct [(dvds)(alcp*) 2 ], [218] emphasizes the carbenoide character of AlCp*. This difference in the reactivity and cluster formation of GaCp* and AlCp* is also observed in the case of the heterocycles Al(DDP) and Ga(DDP). When reacting Ga(DDP) with [Pd 2 (dvds) 3 ], only the bridging dvds ligand is replaced giving the monomeric complex [(dvds)pd{ga(ddp)}] [159] (Scheme 36), independent on the amount of Ga(DDP) used. It is believed, that the steric demand of the heterocycle is a limiting factor in this reaction. [Pd 2 (dvds) 3 ] xs. Al(DDP) or 1 Ga(DDP) E Pd Me 2 Si Si Me 2 O E = Al (16) E = Ga Scheme 36: Synthesis of [(dvds)pd{e(ddp)}] (E = Al (16), Ga). In contrast to that, the reaction of [Pd 2 (dvds) 3 ] with Al(DDP) also depends on the ligand-tometal ratio M:Al, similar as observed for GaCp*. Thus, the reaction of [Pd 2 (dvds) 3 ] with an excess of Al(DDP) yields in the monomeric complex [(dvds)pd{al(ddp)}] (16), whereas the dimeric compound [{Pd(dvds)} 2 {μ 2* -Al(DDP)}] (23) can be obtained on the reaction of a 1:1 mixture of [Pd 2 (dvds) 3 ] and Al(DDP) (Pd:Al ratio: 2:1). A detailed discussion of the latter 110

127 III. esults and discussion monomeric Pd(0)-complexes complex as well as of the influence of the metal-to-ligand ratio in this reaction can be found in chapter III Similarly to the synthesis of [(dvds)pd{ga(ddp)}], the related Al compound [(dvds)pd{al(ddp)}] (16) is synthesized by treatment of [Pd 2 (dvds) 3 ] with an excess Al(DDP) (2.5 eq.) in hexane at room temperature in yields around 74% (Scheme 36). Compound 16 forms bright yellow, rhombohedral crystals by slow crystallization from hexane at 4 C. It can be stored in an inert gas atmosphere for several weeks at temerature below -30 C. In contrast to its Ga-analogue, it slowly decomposes at ambient temperatures also in an inert gas atmosphere, and rapidly on contact with air or moisture in the solid state as well as in solution. The 1 H-M spectrum of 16 at room temperature is quite similar to those reported for the Ga-analogue or the related i compound [i(dvds){ga(ddp)}] (15). Thus, a typical set of signals for a C s symmetric, coordinated Ga(DDP) moiety is observed, including the singlet signal for the backbone-proton at 5.21 ppm, the septet of the i Pr-CH protons at 3.02 ppm and the two doublet signals at 1.27 ppm and 1.07 ppm for the i Pr-Me protons, respectively. A complex spin system in the range of 2.76 ppm to 2.42 ppm can be assigned to the dvdsolefins, whereas two distinct signals for the SiMe 2 groups can be observed at 0.52 ppm and ppm, respectively. The 13 C-M spectrum also shows the expected signals. Figure 23: Molecular structure of [(dvds)pd{al(ddp)}] (16). For structural parameters see Table

128 III. esults and discussion monomeric Pd(0)-complexes Complex 16 crystallizes in the triclinic space group P 1. The molecular structure is depicted in Figure 23. Selected bond lengths and angles are shown in Table 30. It exhibits a coplanar arrangement of the vinyl-groups of the dvds ligand, the Pd and the Al atoms (angular sum: C=C centroids, Pd, Ga: ), respectively. The Pd-C distances (2.14 to 2.17 Å) are very similar to the respective distances in [(Me 3 P)Pd(η 2, η 2 -diallylether)] (2.15 to 2.16 Å), [162] whereas the C=C bond distances are considerably longer (1.405(4) Å and 1.411(4) Å in 16 vs Å and Å in [(Me 3 P)Pd(η 2, η 2 -diallylether)]), pointing to a distinctly higher degree of π-backbonding to the olefins in the Al(DDP) complex compared to its phosphine analogon. Both C=C distances are also similar to those found in the Ga analogue [Pd(dvds){Ga(DDP)}], hence indicating similarly strong σ-donor capabilities of Al(DDP) and Ga(DDP), respectively. Table 30: Selected bond lengths (Å) and angles ( ) for [(dvds)pd{al(ddp)}] (16). Pd-Al (10) C50-C (4) Pd-C (3) C56-C (4) Pd-C (3) Pd-C (3) 1-Al (11) Pd-C (3) 1-Al-Pd (9) Al (3) 2-Al-Pd (9) Al (3) Interestingly, the monomeric HC complex [(HC)Pd(dvds)] (HC = (2,6- Diisopropylphenyl-) 2 C 3 H 2 ) [161] shows very similar C=C bond lenghts (1.40Å) as compared with 16. Thus, the σ-donor properties of Al(DDP) and Ga(DDP) seem to be indeed more like HC's rather than electron rich phosphines. The Pd-Al distance is with (10) Å significantly longer than in the monomeric compound [Pd(AlCp*) 4 ] (2.2950(9) Å) [54] or the terminal Pd-Al bond distances in [Pd 2 (AlCp*) 5 ] (Pd-Al: 2.32 Å), [55] but comparable to the terminal Pd-Al bond lengths found in the trinuclear cluster [Pd 3 (AlCp*)2(μ²-AlCp*) 2 (μ 3 - AlCp*) 2 ] (terminal Pd-Al: 2.38 and 2.35Å). [55] The Al- distances in 16 are with 1.906(3) and 1.912(3) shortened with respect to the free ligand (1.96 Å) [30], but similar to those found in the Lewis acid-base adduct [{(DDP)Al}B(C 6 F 5 )] ( Å). [125] 112

129 III. esults and discussion Oxidative addition of H 2 to a Pt(0) complex 3.3. Oxidative addition of H 2 and H-SiEt 3 to [Pt(1,3-cod){Ga(DDP)} 2 ] and H/D exchange with C 6 D Oxidative addition of H 2 to [Pt(1,3-cod){Ga(DDP)} 2 ]. In contrast to [(dvds)pd{ga(ddp)}], which proved to be completely inert towards H 2 and HSiEt 3 under various conditions, [(1,3-cod)Pt{Ga(DDP)} 2 ] was found to readily react with both reagents under ambient conditions. [159] Thus, treatment of [(1,3-cod)Pt{Ga(DDP)} 2 ] with excess of H 2 in hexane (simply by bubbling H 2 through a solution of [(1,3-cod)Pt(Ga(DDP)} 2 ] in hexane or applying increased pressure (3 bar)), leads to the formation of the dihydride complex trans-[pt{ga(ddp)} 2 (H) 2 ] (17) (Scheme 37) as yellow crystals. Analogously, the reaction of [Pt(cod) 2 ] with two eq. Ga(DDP) and subsequent treating with hydrogen also gives 17 in similar yield and purity. As proven by 1 H M spectroscopy, the diolefin ligand 1,3-cod is completely hydrogenated in the course of the reaction giving cyclooctane as a side product. Even addition of an excess of 1,5-cod (more than 10 eq.) leads to a complete, therefore obviously catalytic, hydrogenation of the cod ligand. This observation corresponds well with the known catalytic activity of different Ptcatalyst in hydrogenation reactions. [232] Compound 17 has been characterized by means of M spectroscopy, elemental analysis and single crystal x-ray analysis. It readily dissolves in all common organic solvents, but exhibits only short-term stability in the solid state and decomposes after several hours also in an inert gas atmosphere at -30 C. + H 2 [(1,3-cod)Pt{Ga(DDP)} 2 ] Ga Pt - C 8 H 16 H H Ga 17 Scheme 37: Oxidative addition of H 2 to give trans-[pt{ga(ddp)} 2 (H) 2 ] (17). The 1 H-M spectrum of pure 17 shows only one set of signals for the Ga(DDP) moieties with one singlet at 5.07 ppm (2H), one septet at 3.06 ppm (8H) and two doublets at 1.17 and 1.13 ppm for the i Pr-methyl groups (each 24H). The Pt-H give rise to one characteristic signal at ppm, showing the expected Pt-satellites (2H, J Pt-H : 334 Hz). The existence of only one set of signals for the coordinated Ga(DDP) moieties as well as the hydride to ligand ratio of 113

130 III. esults and discussion Oxidative addition of H 2 to a Pt(0) complex 1:1 suggests the formation of a C 2v symmetric Pt complex, i.e. a trans-ptl 2 H 2 structure. Monitoring the reaction by in situ 1 H-M spectroscopy reveals the formation of an intermediate. Thus, besides 17, a second Pt-hydride species can be detected (-3.87 ppm, 2H, J Pt-H : 377 Hz), showing the same hydrogen to ligand ratio 1:1, but having reduced idealised symmetry (C s ) with respect to 17. A reasonable suggestion for the structure of this compound is a cis-dihydride being the direct product of the oxidative addition of H 2 to the unsaturated [Pt{(Ga(DDP)} 2 ] fragment (Scheme 38). The general stability of [Pt{(Ga(DDP)} 2 X 2 ] complexes with a cis arrangement of the Ga(DDP) ligands is shown in the formation of the cis-hydride-silane complex 18 (Chapter III.3.3.2), however, neither pure samples nor crystals suitable for x-ray analysis of this cis-dihydride complex could be obtained. + H 2 [(1,3-cod)Pt{Ga(DDP)} 2 ] Ga Pt - C 8 H 16 H 17 H Ga Ga H Pt H Ga Scheme 38: proposed mechanism of the oxidative addition of H 2 to form 17. On recrystallisation of the crude product 17 from hexane, yellow, rhombohedral crystals suitable for x-ray analysis were collected. Compound 17 crystallizes in the orthorhombic space group C2/c. The molecular structure (Figure 24) consists of a central Pt-atom linearly coordinated by two Ga(DDP) moieties (Ga-Pt-Ga: 180 ). Table 31 shows a selection of bond lengths and angles of 17. The Pt-Ga distances are with Å to the best of our knowledge the shortest bond distances for a terminally coordinated Ga(I) ligand reported so far, being considerably shorter than in [(1,3-cod)Pt{Ga(DDP)} 2 ] (2.34 Å). [159] Similar short bond distances of av Å are reported for the complex [Pt[Ga({(Ar)} 2 CCy 2 })] 3 ] [60] or Å in [(dcpe)pt{gac(sime 3 ) 3 } 2 ] [dcpe = bis(dicyclohexylphosphino)ethane]. [87] 114

131 III. esults and discussion Oxidative addition of H 2 to a Pt(0) complex Figure 24: Molecular structure of trans-[pt{ga(ddp)} 2 (H) 2 ] (17). For structural parameters see Table 31. Table 31: Selected bond lengths (Å) and angles ( ) for trans-[pt{ga(ddp)} 2 (H) 2 ] (17). Pt-Ga (15) Ga-Pt-Ga* (5) Ga (6) 1-Ga (3) Ga (6) 1-Ga-Pt (19) 2-Ga-Pt 135.3(2) The presumably strong Pt-Ga interaction is also reflected by Ga- distances of 1.96 and 1.94 Å, which are significantly shorter than in free Ga(DDP) (2.05Å) [31] but comparable to [Pt{Ga(DDP)}(CO)] 2 ( Å) [159] and the cationic complex [{(DDP)Ga. THF} 2 Au][BAr F ] (10. 2THF) (1.94 to 1.97 Å). Additionally, the hydrides could not be located in the solid state structure. 115

132 III. esults and discussion Oxidative addition of H 2 to a Pt(0) complex The most striking structural feature of this complex is the parallel arrangement of the Ga 2 planes. Thus, a twist of 90 would be sterically favorable, as seen e.g. in the case of 10, and the fact that the ligands adopt a coplanar arrangement must explained either electronically or by packing effects. Hence, an overlap of the empty p z (Ga) orbitals and a filled Pt-orbital of suitable symmetry (e.g. p z ) in the square planar d 8 complex explains this geometry. However, other results (e.g. the geometrical features of complexes [(1,3-cod)Pt{Ga(DDP)} 2 ] and [(dvds)pd{e(gaddp)}] (E = Ga, Al (16)) suggest this classical π-backbonding, as recently calculated to be relevant for the naked cation Ga +, to be rather weak in these complexes. 116

133 III. esults and discussion Oxidative addition of HSiEt 3 to a Pt(0) complex Oxidative addition of HSiEt 3 to [Pt(1,3-cod){Ga(DDP)} 2 ]. As mentioned before, in the course of the reaction of [(1,3-cod)Pt{Ga(DDP)} 2 ] with H 2, an intermediate can be observed, which is suggested to be cis-[pt(h) 2 {Ga(DDP)} 2 ]. Obviously, a fast rearrangement of the ligands occurs and only the trans-analogue trans-[pt{ga(ddp)} 2 (H) 2 ] (17) can be obtained. As seen in the reaction of HSiEt 3 with the reactive intermediate [i(alcp*) 3 ], HSiEt 3 can undergo an oxidative addition to form [(H)(SiEt 3 )i(alcp*) 3 ]. [56] From this result ist was suggested, that the oxidative addition of the benzene molecule in the formation of [i(alcp*) 3 (AlCp*{Ph})(H)] proceeds via the intermediate [i(alcp*) n (H)(Ph)] (n<4). In this respect, the oxidative addition of HSiEt 3 as sterically more demanding ligand with respect to H 2 should lead to a similar complex. Therefore, [(1,3-cod)Pt{Ga(DDP)} 2 ] was reacted with HSiEt 3. [(1,3-cod)Pt{Ga(DDP)} 2 ] + HSiEt 3 - C 8 H 12 Ga H Pt 18 Ga SiEt 3 Scheme 39: Synthesis of [Pt{Ga(DDP)} 2 (H)(SiEt 3 )] (18). When treating a red solution of [(1,3-cod)Pt{Ga(DDP)} 2 ] in hexane with an excess of HSiEt 3 at room temperature, the color of the solution immediately brightens to orange. After stirring the solution for 2h at room temperature, cis-[pt{ga(ddp)} 2 (H)(SiEt 3 )] (18) can be obtained as an orange solid by precipitation at -30 C overnight (Scheme 39). Compound 18 readily dissolves in all common organic solvents like hexane, benzene or THF and can be stored in the absence of air or moisture for several months without decomposition. The 1 H-M spectrum at room temperature in C 6 D 6 consists of a complex set of partially overlapping and broad signals. Whereas both γ-c of the heterocycles give rise to only one sharp singlet at 4.95 ppm, all other DDP signals are separated, indicating two chemically non equivalent, locally C s symmetric Ga(DDP) ligands. However, most of the signals are 117

134 III. esults and discussion Oxidative addition of HSiEt 3 to a Pt(0) complex distinctly broadened, indicating some kind of fluxional process in solution. The protons of the coordinated SiEt 3 -moiety can be found at 1.19 ppm (q, 6H) and 0.93 ppm (t, 9H). The hydride gives rise to a sharp triplet at ppm (J Pt-H : 530 Hz). At 70 C, the 1 H M spectrum noticeably simplifies, showing only one set of signals for the DDP-ligands (C 2v symmetry). Most likely, either reductive elimination/oxidative addition of HSiEt 3 or dissociation/association of Ga(DDP) is the origin of this fluxional process, however, the exact mechanism being unclear. The catalytic deuteration of HSiEt 3 by C 6 D 6 (vide infra), however, points to the reversibility of the silane addition similar to [(H)(SiEt 3 )i(alcp*) 3 ]. Unfortunately, neither 29 Si nor 195 Pt M spectroscopy gave spectra of reasonable intensities. Also the 13 C M spectrum of 18 is consistent with this fluxional process. Interestingly, the hydride signal disappears on heating the M sample for a long period of time (> 2h), whereas all other signals remain unchanged. Obviously an H/D exchange with the solvent C 6 D 6 is taking place, which is also indicated by a concomitant increase of the solvent signal in the 1 H M spectrum. Indeed, the H/D exchange reaction is catalytic: Thus, heating a solution of 18 in C 6 D 6 in the presence of a large excess of HSiEt 3 (about 10 eq.) leads to complete disappearance of the Si-H signal in the 1 H M spectrum after two days at 70 C, while all other signals (including those of complex 18 exept the Pt-Hydride signal) again remain unchanged. Again the exact mechanism is unclear. However, taking into account that C-H activation reactions of C 6 H 6 on Pt(II) complexes are well known, [233] a C-D activation of C 6 D 6 by 18 forming a Pt(IV) "hydride-deuteride" intermediate followed by reductive elimination of C 6 D 5 H and DSiEt 3 is a reasonable pathway. Crystals of 18 suitable for x-ray analysis were formed by cooling a saturated solution in hexane to -30 C overnight. The molecular structure is shown in Figure 25 and consists of a Pt center surrounded by two Ga(DDP) ligands situated cis to each other as well as a SiEt 3 moiety. Selected bond lengths and angles are shown in Table 32. The Pt, Ga and Si atoms adopt a coplanar arrangement (angular sum: ). The hydride ligand could not be located in the solid state structure. However, the arrangement of the ligands around the platinum center can be best described as a slightly distorted T-shape. Thus, two small angles (Si-Pt-Ga2 105,7 and Ga2-Pt-Ga1 104,1 ) and one considerably larger angle (Si-Pt-Ga1 150,1 ) create an open coordination site, suggesting the hydride to be located trans to Ga2. Considering the high steric demand of the Ga(DDP) ligand, the cis-coordination of the two Ga(DDP) moieties is somehow unusual but consistent with the coordination of the space-consuming SiEt 3 ligand. 118

135 III. esults and discussion Oxidative addition of HSiEt 3 to a Pt(0) complex Figure 25: Molecular structure of [Pt{Ga(DDP)} 2 (H)(SiEt 3 )] (18). For structural parameters see Table 32. The Pt-Ga distances are very similar (2.36 Å and 2.38 Å) and somewhat longer than in [(1,3-cod)Pt{Ga(DDP)} 2 ] (2.34 Å) [159] or [Pt(GaCp*) 4 ] (2.335 Å), [54] but comparable to those found in [(cod)pt[ga({(ar)} 2 CCy 2 })] 2 ] (2.357). [60] The Ga- distance (1.993 Å to Å) are in the same range as found for free Ga(DDP) (2.05Å) or the parent compound [(1,3-cod)Pt{Ga(DDP)} 2 ] ( Å). The Pt-Si bond distance is with Å quite similar to the Pt-Si bond distances of trans-[pt(h)(sih 3 )(PCy 3 ) 2 ] (2.38 Å). [234] Table 32: Selected bond lengths (Å) and angles ( ) for [Pt{Ga(DDP)} 2 (H)(SiEt 3 )] (18). Pt-Ga (9) Si-Pt-Ga (3) Pt-Ga (6) Si-Pt-Ga (3) Pt-Si (12) Ga1-Pt-Ga (3) Ga (3) 11-Ga (11) Ga (3) 21-Ga (11) Ga (3) Ga (3) 119

136 III. esults and discussion Oxidative addition of HSiEt 3 to a Pt(0) complex Summary. The observed oxidative addition of H 2 yield the Pt(II) dihydride species trans-[pt{ga(ddp)} 2 (H) 2 ] (17). Accordingly, the first step of the formation of 17 is suggested to be the direct oxidative addition of H 2 to the unsaturated [Pt{(Ga(DDP)} 2 ] fragment which gives the cis dihydride cis-[pt{ga(ddp)} 2 (H) 2 ], based on the spectroscopic characterization of this intermediate. However, the catalytic hydrogenation of 1,5-cod in this reactions is a promising finding which may stimulate further studies. The same is true for the apparently nicely reversible oxidative addition of HSiEt 3 to [(1,3-cod)Pt{Ga(DDP)} 2 ] to give cis-[pt{ga(ddp)} 2 (H)(SiEt 3 )] (18). Similar results are reported for the chemistry of the [i(alcp*) 3 ] fragment, which is suggested to be the reactive intermediate in such Si-H or C- H activation reactions. The observation of H/D exchange with the solvent C 6 D 6 again shows some potential of the [E(DDP)] ligands to act as ancillary directing ligands similar to HCs and phosphanes in classical organometallic chemistry. The experimental findings described in this chapter are well consistent with previous calculations on these class of compounds stating their strong σ-donor but rather weak π-acceptor properites. The sterically demanding ligands E(DDP) are obviously not only isolobal, but also electronically comparable to the widely used HC ligands, based on geometrical features of solid state structures, spectroscopic data as well as the reactivities of their transition metal complexes. 120

137 III. esults and discussion. 4. Compounds with higher nuclearity 4. Compounds with higher nuclearity. The ability of low valent group 13 organyls ECp* to effectively bridge two or three transition metal centers via μ 2 or μ 3 coordination modes, respectively, is a specific feature of this novel ligand class and has led to the characterization of a wide variety of mixed metal clusters of different nuclearities and structures. [43, 45, 235] The preparation of these mixed metal clusters most often involves replacement of olefins from a transition metal-olefin precursor, e.g. [Pt(cod) 2 ], [i(cod) 2 ], [Pd 2 (dvds) 3 ] etc., by the more strongly binding ligands E I. The direct formation and isolation of larger clusters, e.g. [Pd 2 (GaCp*) 5 ] or [Pd 3 (GaCp*) 8 ] is possible by precisely controlling the molar ratios of the employed transition metal precursor to the E I ligand. [55, 236] The detailed understanding of this chemistry is important for the use of GaCp* and AlCp* together with all-hydrocarbon transition metal complexes as precursors for soft chemical synthesis of the respective intermetallic nanophases, i.e. ial, in non aqueous media. [237] However, the close vicinity of very electrophilic metals E and highly nucleophilic transition metals is an intrinsically interesting feature of the [L n M(E I )] complexes, apart from potential applications in materials science. [5-14] Classical organometallic reactions, such as the activation of small molecules with rather inert chemical bonds (e.g. C-H, Si-H and C-C activation reactions) have been observed in certain mononuclear complexes of the type [M(E I ) n ]. [167, 238] From this point of view it seems quite interesting to introduce the rather exotic low valent group 13 species E I as novel ancillary ligands for mixed metal cluster cores. The question arises, whether it might be possible to extend the classical organometallic activation chemistry from mononuclear, homoleptic complexes [M(E I ) n ] to clusters [M a (E I ) b ]. Conceptually these clusters are regarded as molecular models of transition metal main group metal alloys M/E which are known for tuneable properties in heterogeneous or [239, 240] nano catalysis. A major problem however is the thermodynamically favourable formation of electronically and sterically saturated clusters such as those stated above with ECp* as the typical ligand, which prevents further reactivity. One strategy to overcome this problem is a tailored increase of the steric bulk of E I, leading to electronically unsaturated clusters which can additionally coordinate the desired small molecules. This strategy has already been successfully applied in the preparation of the Pd/Ga clusters [Pd 3 (GaCp* Ph ) 4 (dvds)] [236] and the clusters [M{Ga(DDP)}(CO)] 2 (M = Pd, Pt) [241] using GaCp* Ph (Cp* Ph = C 5 H 4 Ph) and Ga(DDP) as more bulky ligands compared to ECp*. The very high steric bulk of E(DDP) also allows the isolation of unsaturated mono nuclear transition metal complexes such as [(1,3-cod)Pt{Ga(DDP)} 2 ] or [Pd{E(DDP)}(dvds)] (E = 121

138 III. esults and discussion Compounds with a i 2 -core Al, Ga) (16). However, the complexes [M{μ 2 -Ga(DDP)}(CO)] 2 are the only examples with bridging Ga(DDP) units so far. Thus, a more detailed study of the ability of Ga(DDP) and Al(DDP) to stabilize compounds [M a (E I ) b ] of higher nuclearity (a = 2, 3, ) seems to be appropriate Dinuclear compounds. [Lit. 8, Chapter VII.3] Compounds with a i 2 -core eactions of [i(c 2 H 4 ) 3 ] with Ga(DDP). The reaction of [i(c 2 H 4 ) 3 ] with Ga(DDP) in a i/ga ratio of 1:1 leads to the substitution of one ethylene ligand by a Ga(DDP) moiety to afford [i(c 2 H 4 ) 2 {Ga(DDP)}] (13), as described in chapter III However, in order to synthesize cluster compounds with higher nuclearities, the i/ga ratio in this reaction is varied to obtain cluster compounds with bridging Ga(DDP) ligands, as observed e.g. in the case of [M{Ga(DDP)}(CO)] 2. Thus, reactions were performed with a i/ga ratio of 2:1 and 3:2 (see chapter III ). As suggested, the reaction of Ga(DDP) with [i(c 2 H 4 ) 3 ] leads to i clusters with bridging Ga(DDP) ligands. Thus, a ratio of i:ga of 2:1 yields in the formation of the complexes [{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19) and [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 -C 2 H 4 )] (20), respectively (Scheme 27). However, when heating the reaction mixture above room temperature, liberation of ethylene occurs and the cluster [{{μ 2 -Ga(DDP)}i(C 2 H 4 )} 2 i(μ 2 -CH=CH 2 )(H)] (27) was formed quantitatively (with respect to Ga(DDP)), independent on the i/ga ratio used in the reaction. Obviously, with a i/ga ratio of 3:2, compound 27 seems to be the thermodynamic product in these cluster formation processes (see chapter III ). H i + Ga(DDP) 2 i( ) 3 Ga H i H H + Ga i H H i H H Scheme 40: eaction of [i(c 2 H 4 ) 3 ] with Ga(DDP) in a ratio of 2:1. 122

139 III. esults and discussion Compounds with a i 2 -core The two clusters 19 and 20 with a i:ga ratio of 2:1 could be isolated in single crystalline form (see experimental section) coinstantanously, when storing the reaction mixture at ambient or low (-30 C) overnight without heating the solution above 30 C. However, both complexes are not stable in solution in the absence of ethylene and decompose on redissolving in organic, yet ethylene saturated solvents. In both cases, besides substantial amounts of 27, also other signals of untraceable species are found in the 1 H-M spectra of 19 and 20, leading to not interpretable M spectra of both compounds. However, in the solid state compounds 19 and 20 show adequate stability to be analyzed by single crystal x- ray diffraction. Figure 26: Molecular structure of [{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19). For structural parameters see Table 33. The molecular structures of 19 (Figure 26) and 20 (Figure 27) both consist of two i centers bridged by a Ga(DDP) moiety (Ga-i bond distances: 2.33 Å in both clusters). Both compounds contain an ethylene molecule in a "bridging" position between the two nickel 123

140 III. esults and discussion Compounds with a i 2 -core centers. A comparable bridging mode of ethylene has recently been described by.. Schrock in a W-W dimeric complex. [242] In the case of 19, each i center is stabilized by two ethylene molecules, with the two (C 2 H 4 ) 2 i-planes perpendicular to each other (torsion angle: ~93 ). Selected bond lengths and angles for 19 are shown in Table 33. The i-i distance (2.50 Å) is only slightly longer than the i I -i I bond in [(Cpi) 2 {GaC(SiMe 3 ) 3 } 2 ] (i-i: 2.45 Å), [83] and in the range of i-i interactions in the i-ga carbonyl cluster [i 4 (μ 2 -GaCp*) 4 (CO) 5 (μ 2 -CO)]. [22] The i-c bond distances to the terminal ethylenes of 1.88 to 2.06 Å (average 2.01 Å) are slightly longer than in 13 (1.96 Å Å) but comparable to those found in the phosphine-compound [(Cy 3 P)i(C 2 H 4 ) 2 ] ( Å). [228] Table 33: Selected bond lengths (Å) and angles ( ) for [{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19). Ga-i (14) i2-c (10) Ga-i (11) i2-c (8) i1-i (18) i2-c (10) i1-c (9) C30-C (11) i1-c (8) C40-C (11) i1-c (9) C50-C (12) i1-c (7) C60-C (9) i2-c (8) Ga (5) i2-c (9) Ga (6) i2-c (10) i2-c (8) 1-Ga (19) i2-c (10) i1-ga-i (5) i2-c (9) C51-C50-i (7). The C=C bond distance of the terminal ethylenes are 1.34 Å, 1.41Å and 1.42 Å, indicating the strong donor properties of the Ga(DDP) moiety, as already discussed in the case of 13. Thus, similar bond distances can be found in in [(Cy 3 P)i(C 2 H 4 ) 2 ] (1.41 and 1.39 Å) or the Fischer carbene complex [(C 2 H 4 ) 2 i{cph( 2 )}] (1.40 and 1.39 Å). [229] The orientation of the bridging ethylene moiety in 19 suggests either an agostic interaction between the ethylene- C-H bond and the i-center or an activation of a C-H bond of ethylene to give a {(H 2 C=CH)i(H)} moiety (i-c bond distance 2.454(10) Å, i-c-c angle: 128 ). However, 124

141 III. esults and discussion Compounds with a i 2 -core the respective proton could not be located in the refinement and thus no other structural parameters of this interaction can be given. Figure 27: Molecular structure of [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 -C 2 H 4 )] (20). For structural parameters see Table 34. The molecular structure of 20 is very similar to this of 19, with the coordination of three instead of four ethylene ligands. For selected bond lengths and angles see Table 34. The i-i distance (2.48 Å) is slightly shorter than in 19 (2.50 Å) and can be explained by less steric strain due to the loss of a C 2 H 4 moiety in 20. Each i-center is coordinated by one terminal ethylene with i-c bond distances of 1.91 Å to 1.99 Å. The third ethylene is located in a bridging position between the two i-centers, again showing a π-bond/agostic coordination mode or a CH-activation during the reaction. However, the i-c bond distance of the bridging ethylene moiety is with 1.980(5) Å significant shorter than in 19 (2.454(10) Å). To shed light on the nature of the coordination of the bridging ethylene molecule (π-bond/agostic coordination mode vs. C-H-activation), preliminary DFT calculations were 125

142 III. esults and discussion Compounds with a i 2 -core performed on the B3LYP/LAL2DZ level of theory. [Ga{(MeCMe) 2 CH}] was used as a model ligand in these calculations. Depictions and xyz-coordinates of the calculated model structures and a table of basis set comparison for the calculation of ΔE (Sum of electronic and Zero-Point-Energy) can be found in the appendix. The corresponding zero point energy differences between activated (19a, 20a) and non-activated (i.e π-bond / agostic coordinated) species (19b, 20b) are positive and relatively high for 19 (19.8 kcal/mol) and only slightly lower for 20 (13.8 kcal/mol). Thus, also DFT calculations reveal an agostic C-H interactions to the bridging ethylene in both i 2 Ga clusters 19 and 20 rather than a C-H activation. Table 34: Selected bond lengths (Å) and angles ( ) for [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 -C 2 H 4 )] (20). Ga1-i (11) i2-c (5) Ga1-i (10) i2-c (4) i1-i (14) i2-c (4) Ga (3) C30-C (5) Ga (3) C32-C (6) i1-c (5) C34-C (6) i1-c (4) i1-c (5) 1-Ga (13) i1-c (5) i1-ga1-i (4) Taking into account, that the monomeric complex [i(c 2 H 4 ) 2 {Ga(DDP)}] (13) shows poor stability and decomposes by liberating ethylene, it is a reasonable suggestion that the formation of 20 proceeds via complex 19 simply by losing one ethylene moiety and rearrangement of the residual ligands. 126

143 III. esults and discussion Compounds with a i 2 -core eactivity of [i(cod) 2 ] towards Ga(DDP) in the presents of PhCCPh. As discussed in chapter III , the reaction of [i(cod) 2 ] with Ga(DDP) leads to the instable, yet not crystallisable compound [(1,3-cod)i{Ga(DDP)} 2 ], which is similar to its well examined Pt analogue. However, the steric demand of the Ga(DDP) seems to be a limiting factor in such reactions, because similar reactions of [i(cod) 2 ] with HC s (HC = {CH} 2 ; = Me, n Pr, i Pr, Mes) has been reported in 2006 by Schaub et al. [243] The results of this study reflect well the influence of the steric demand of the ligands: The reaction with a methylsubstituted carbene, large enough to prevent tetrasubstitution, affords a trigonal planar complex [i(hc) 3 ], whereas the usage of bulky aryl-substituted carbenes ( = Mes) leads to homoleptic, two-coordinated complexes [i(hc) 2 ]. Dinuclear, cod-bridged complexes [i 2 (HC) 4 (cod)] are obtained from carbenes with intermediate steric bulk ( = n Pr, i Pr), large enough to prevent threefold substitution but too small to stabilize a two-coordinated structure. In the case of Ga(DDP), the steric demand seems to be large enough to prevent a threefold coordination or a cod-bridging motif, but too small to stabilize a [i{ga(ddp)} 2 ] complex. However, the aforementioned cod-bridged complex [{i(hc) 2 } 2 (μ 2 -cod)] smoothly reacts with diphenyl acetylene to give the complexes [i(hc) 2 (PhCCPh)] ( = n Pr, i Pr) by substitution of the cod ligand. As seen in the formation of 14, the mixture of [i(cod) 2 ] and Ga(DDP) also react with additional ligands like olefins to give stable products. Thus, the analogues reaction using PhCCPh as co-ligand was performed. Treatment of [i(cod) 2 ] with Ga(DDP) and PhC CPh in a ratio of 1:2:1 did not give a monomeric complex similar to [i(hc) 2 (PhCCPh)]. Instead, the formation of a dark red solid was observed, which could be analysed to be the i 2 Ga cluster compound [(cod)i 2 {μ 2 -Ga(DDP)}{μ 2 -(PhC CPh)}(PhC CPh)] (21) (Scheme 41). However, the overall yield was poor due to the wrong ratio of the ligands. Accordingly, high yields up to 90% could be obtained when mixing the reactants in a ratio of 2:1:2 (i:ga:phccph). Compound 21 is stable in an inert gas atmosphere for several weeks without decomposition and can be redissolved in common organic solvents like hexane, benzene or toluene. 127

144 III. esults and discussion Compounds with a i 2 -core Ph Ph Ph 2 i(cod) 2 + Ga(DDP) + 2 Ph Ph - 3 cod Ph i Ga i 21 Scheme 41: Synthesis of [(cod)i 2 {μ 2 -Ga(DDP)}{μ 2 -(PhC CPh)}(PhC CPh)] (21). The 1 H-M spectrum of 21 shows the expected signals for a C s symmetric Ga(DDP) ligand, i.e. a singlet at 5.26 ppm (heterocycle-ch), two septet signals at 3.81 ppm and 3.18 ppm for the i Pr-CH protons and the respective four doublet signals for the i Pr-Me groups at 1.50 ppm, 1.24 ppm, 0.82 ppm and 0.52 ppm. The cod-ligand gives rise to two broad signals at 5.57 and 5.49 ppm (2H each) for the olefinic protons as well as a multiplet at 1.85 ppm (8H) for the aliphatic protons. Besides the expected signals for the Ga(DDP) ligand, signals for the acetylenic carbons can be found at 96.6 and 91.4 ppm in the 13 C-M spectrum. Table 35: Selected bond lengths (Å) and angles ( ) for [(cod)i 2 {μ 2 -Ga(DDP)}{μ 2 - (PhC CPh)}(PhC CPh)] (21). Ga-i (4) i2-c (2) Ga-i (4) i2-c (2) i1-i (4) C33-C (3) i1-c (2) C30-C (3) i1-c (2) C40-C (3) i1-c (2) C60-C (3) i1-c (2) Ga (18) i1-c (2) Ga (18) i1-c (2) i2-c (2) 1-Ga (8) i2-c (3) i1-ga-i (12) 128

145 III. esults and discussion Compounds with a i 2 -core Crystals suitable for single x-ray diffraction analysis were obtained from a saturated solution of 21 in hexane by cooling to -30 C overnight. Compound 21 crystallizes in the orthorhombic space group P Similar to 19 and 20, the molecular structure of 21 (Figure 28) consists of two i centers being bridged by a Ga(DDP) moiety. Selected bond lengths and angles are shown in Table 35. The 1,5-cod ligand is coordinated to i1 in a η 4 -coordinaten mode, whereas the second i center (i2) bears a terminally coordinated PhC CPh ligand. In addition, the two i centers are bridged by a PhC CPh moiety. Figure 28: Molecular structure of [(cod)i 2 {μ 2 -Ga(DDP)}{μ 2 -(PhC CPh)}(PhC CPh)] (21). For structural parameters see Table 35. The i-ga bond distances in 21 are with 2.41 Å and 2.46 Å longer than the observed bond length for terminal coordinated Ga(DDP) moieties (e.g. in 13 and 14 (2.28 Å)), as expected for a Ga(DDP) in a bridging position, but also significantly longer than the i-ga bond length in 19 and 20 (2.33 Å average). Similar i-ga bond length could only be found for the 129

146 III. esults and discussion Compounds with a i 2 -core μ 2 -bridging GaCp*-moiety in the cluster compound [(μ 2 -CO)(CO) 5 {(Cp*Ga)i} 4 ] (2.44 Å). [22] The C=C double bond distance of the coordinated cod-ligand is with 1.38 Å similar to the parent compound [i(cod) 2 ]. [244] The C C distance of the bridging PhC CPh (1.31 Å) is somehow shorter than in [(μ 2 -PhC CPh)(μ 2 -tetraphenylbutadienyl)(cp*) 2 i 3 ] (1.33 Å) [245] or in [(μ 2 -PhC CPh){i(cod)} 2 ] (1.388 Å) [246], whereas the terminal PhC CPh ligand (1.279 Å) shows a similar bond distance as reported for other terminal diphenylacetylen compounds (e.g. [( t BuC) 2 i(phc CPh)] (1.276 Å) [247] or [(cod)i(phc CPh)] (1.28 Å). [248] Additionally, the i-c acetylene bond distances (1.86 to 2.04 Å) are in the same range than those reported for other i-(phc CPh) compounds in the literature. 130

147 III. esults and discussion Compounds with a Pd 2 -core [Lit. 7, Chapter VII.3] Compounds with a Pd 2 -core Substitution reactions on [(dvds)pd{ga(ddp)}]. The use of monomeric compounds as building blocks for the synthesis of clusters with higher nuclearity was investigated by us in more detail in [54] Thus, the monomeric compounds [M(GaCp*) 4 ] (M = Pd, Pt), which were found to be kinetically inert towards substitution reactions, react with a suitable M(0) source and additional GaCp* to form the dimeric homoleptic complexes [MPt(GaCp*) 5 ] (M = Pt, Pd). However, the formation is strongly dependent on the order of mixing the three reagents and proceeds via reactive, yet isolable dinuclear intermediates of the type [MPt(η 2 -cod)(gacp*) 4 ] on reaction with [Pt(cod) 2 ], as proven by 1 H-M spectroscopy. These results show that also the choice of the M(0) precursor is very important for the success of the cluster synthesis. Additionally, the coordinated ligands in such cluster compounds were found to readily undergo ligand exchange reactions with a variety of different ligands. [45, 55] Thus, on treatment of [Pt 2 (GaCp*) 5 ] with AlCp*, only the bridging GaCp* ligands were substituted by AlCp* to give [Pt 2 (μ 2 -AlCp*) 3 (GaCp*) 2 ], whereas a complete substitution of the GaCp* ligands is observed in the Pd-analogue to give [Pd 2 (AlCp*) 5 ]. Obviously, AlCp* favors the bridging position, possibly due to the stronger σ-donor properties of AlCp* compared to GaCp*. Interestingly, the two latter dinuclear clusters do not undergo further substitution reactions with phosphines, indicating a higher bond strength of M-Al in the first place. In contrast to that, the terminal coordinated GaCp* ligands in [Pd 2 (GaCp*) 5 ] can be replaced by PPh 3 to give [(Ph 3 P) 2 Pd 2 (μ 2 -GaCp*) 3 ]. Also, on reaction with the strond π-acceptor ligands CO or t BuC, the related compounds [(L) 2 Pt 2 (μ 2 -GaCp*) 3 ] (L = CO, t BuC) can be obtained. As seen in chapter III.3.1. and III.3.3., the monomeric compounds with a coordinated Ga(DDP) ligand are not kinetically inert and thus, the olefinic ligand in e.g. [(1,3- cod)pt{ga(ddp)} 2 ] can be replaced by H 2 or HSiEt 3 to give the oxidative addition products [Pt(H) 2 {Ga(DDP)} 2 ] (17) and [Pt(H)(SiEt 3 ){Ga(DDP)} 2 ] (18), respectively. Interestingly, when treating [(1,3-cod)Pt{Ga(DDP)} 2 ] with strong π-acceptor ligands like CO, the dimeric cluster compound [Pt{μ 2 -Ga(DDP)}(CO)] 2 is formed, which exhibits two Ga(DDP) ligands in a bridging mode between the two Pt-centers. [159] With the synthesis of the monomeric complexes [(dvds)pd{e(ddp)}] (E = Al (16), Ga) and the observed poor stability of the Al 131

148 III. esults and discussion Compounds with a Pd 2 -core analogue, these compounds are potential candidates for such replacement reactions. Thus, substitution reactions of the monomer [(dvds)pd{ga(ddp)}] were examined in more detail. In contrast to [(1,3-cod)Pt{Ga(DDP)} 2 ], which readily reacts with H 2 also at room temperature, no reaction can be observed when treating [(dvds)pd{ga(ddp)}] with H 2 also at increased pressure (3 bar). Likewise, no reaction occurs when mixing HSiEt 3 with [(dvds)pd{ga(ddp)}] in hexane or toluene, also when heating the reaction mixture for 48 h to reflux. Obviously, the Pt compound is more reactive towards oxidative addition reactions than its Pd congener. A comparably high reactivity of [(dvds)pd{ga(ddp)}] towards the strong π-acceptor CO was observed. Similar to the formation of [Pt{μ 2 -Ga(DDP)}(CO)] 2, the analogues Pd complex [Pd{μ 2 -Ga(DDP)}(CO)] 2 can easily be prepared by bubbling CO through a solution of [(dvds)pd{ga(ddp)}] in hexane. [159] However, the use of gaseous reactants always bears the problem of dosage during the reaction, caused by leakages in the used reaction vessel or the pressure line. Therefore, the reaction with another class of strong, yet not gaseous π-acceptor ligands, was performed. 2 [(dvds)pd{ga(ddp)}] xs. t BuC - 2 dvds Ga C t Bu Pd Ga Pd Bu t C 22 Scheme 42: eaction of [(dvds)pd{ga(ddp)}] with t BuC. As expected, the olefins in [(dvds)pd{ga(ddp)}] can also be substituted by isocyanide (C) ligands. The dimeric compound [Pd{μ 2 -Ga(DDP)}( t BuC)] 2 (22) is formed on reaction of [(dvds)pd{ga(ddp)}] with t BuC and could be isolated as a deep red solid in high yields according to Scheme 42. Compound 22 is insoluble in all common organic solvents and thus no solution M spectroscopic data are available. The molecular structure of 22 was confirmed by single crystal x-ray diffraction and is discussed below, deep red/violett crystals being obtained by slow diffusion of t BuC into a solution of [(dvds)pd{ga(ddp)}] in hexane or THF. 132

149 III. esults and discussion Compounds with a Pd 2 -core Figure 29: Molecular structure of [Pd{μ 2 -Ga(DDP)}( t BuC)] 2 (22). For structural parameters see Table 36. Compound 22 crystallizes in the monoclinic space group C2/m. The molecular structure of 22 (Figure 29) is similar to the the carbonyl compound [Pd{μ 2 -Ga(DDP)}(CO)] 2, [159] showing a Pd 2 Ga 2 core with two Pd centers bridged by two Ga(DDP) ligands with an almost ideal square planar arrangement (angular sum: ). Selected bond lengths and angles can be found in Table 36. The Pd-Ga distances are almost equal (Pd-Ga: 2.469(2); Pd*-Ga: 2.484(2) Å) and expectedly longer than in the monomer [(dvds)pd{ga(ddp)}] (2.40 A) and [Pd(GaCp*) 4 ] (2.37 Å) [54]. Slightly longer bond distances are reported for the CO analogue [Pd{μ 2 -Ga(DDP)}(CO)] 2 (2.49 Å and 2.51 Å) as well as for the dimeric complex [Pd 2 (GaCp*) 5 ] (2.487 Å Å). [55] The Ga- bond length (2.022(3) Å) are slightly shorter than in free Ga(DDP) (2.05 Å), [31] but somehow elongated compared to the CO analogue (1.97 Å and 1.98 Å). Also, the Pd-C distance is elongated (1.953(5) Å) with respect to [Pd{μ 2 -Ga(DDP)}(CO)] 2 (1.882 Å), but shorter than in Pd(II)-isocyanide complexes (e.g. 133

150 III. esults and discussion Compounds with a Pd 2 -core in [Pd(trans- t BuC) 2 (trans-pmeph 2 )][SC] 2 : Å [249] ). The C bond distance of the isocyanide (1.161 Å) is slightly longer than in [Pd(trans- t BuC) 2 (trans-pmeph 2 )][SC] 2, pointing to a higher degree of π-backbonding in 22 caused by the 2e-donor Ga(DDP)-ligands. Table 36: Selected bond lengths (Å) and angles ( ) for [Pd{μ 2 -Ga(DDP)}( t BuC)] 2 (22). Pd-Ga 2.469(2) 1-Ga-1* 91.02(16) Pd-Ga* 2.484(2) 1-Ga-Pd (9) Pd-Pd* (16) 1-Ga-Pd* (8) Ga (3) Ga-Pd-Ga* (7) Pd-C (5) Pd-Ga-Pd 65.03(7) C (6) C16-Pd-Pd* (16) Pd-C (5) 134

151 III. esults and discussion [{Pd(dvds)} 2 {μ 2 -Al(DDP)} [Lit. 5 and 7, Chapter VII.3] eaction of [Pd 2 (dvds) 3 ] with Al(DDP). As discussed in chapter III.3.2., the metal-to-ligand ratio in the reaction of ECp* with [Pd 2 (dvds) 3 ] is a crucial factor for the formation of the homoleptic complexes [Pd 2 (GaCp*) 5 ] and [Pd 3 (GaCp*) 8 ], respectively [55, 144]. Whereas the formation of the monomeric complex [(dvds)pd{ga(ddp)}] seems to be the only isolable product and thus the thermodynamic sink in the reaction with Ga(DDP), the reaction of Al(DDP) with [Pd 2 (dvds) 3 ] is more influenced by the metal-to-ligand ratio used in the reaction. Thus, the reaction of [Pd 2 (dvds) 3 ] with an excess of Al(DDP) yield in the formation of the monomeric complex [(dvds)pd{al(ddp)}] (16) (see chapter III.2.1.), whereas on addition of only one equivalent of Al(DDP) in hexane at room temperature (Pd:Al ratio: 2:1), the solution immediately turns yellow. On removal of the solvent, yellow needles of the dimeric compound [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23) were isolated (Scheme 43) in moderate yields. Obviously, comparable to the formation of [(dvds)pd{ga(ddp)}], only the bridging dvds ligand is replaced by Al(DDP). In contrast to the latter, Al(DDP) is located in a bridging position between the two Pd-centers. This preference of the bridging position has been reported previously for the AlCp* analogue [55] and has been attributed to the stronger σ-donor properties of AlCp*. However, the previously reported dual Lewis acid and Lewis base character of Al(DDP), best exemplified by the short Al-F interaction in the adduct [(DDP)Al B(C 6 F 5 )], [125] may also play a role in the case of Al(DDP). Compound 23 was found to be stable at -30 C in an inert gas atmosphere for several weeks, but decomposition occurs at room temperature after several hours or immediately upon contact with air or moisture. [Pd 2 (dvds) 3 ] + 1 Al(DDP) Me 2 Si O Si Me 2 Pd Al 23 Pd SiMe 2 O SiMe 2 Scheme 43: Synthesis of [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23). 135

152 III. esults and discussion [{Pd(dvds)} 2 {μ 2 -Al(DDP)} The 1 H M spectrum of 23 shows a set of signals for a Al(DDP) ligand with reduced (C s ) symmetry, indicated by the two septet signals for the i Pr-CH protons at 3.25 ppm and 2.96 ppm, as well as four distinct doublet signals at 1.26 ppm, 1.12 ppm, 1.06 ppm and 0.90 ppm for the i Pr-Me groups, respectively. Additionally, one broad signal for the SiMe 2 groups at room temperature is observed, indicating a fluctional process intermolecularly exchanging dvds ligands. Indeed, at -60 C, this broad signal splits into six distinct, yet partially overlapping singlets. The 13 C-M spectrum is in accordance with this result, also exhibiting a broad signal for the SiMe 2 groups at 3.2 ppm. Figure 30: Molecular structure of [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23). For structural parameters see Table 37. The molecular structure of 23 (Figure 30) consists of a central Pd 2 Al-core with two {(dvds)pd}-moieties bridged by one Al(DDP) ligand. Selected bond lengths and angles are shown in Table 37. The Pd-Pd distance of Å is significantly longer than in Pd 2 (GaCp*) 5 (2.609 Å), [55] but similar to the one in Pd 3 (GaCp*) 8 (2.843 Å), [144] which was reported to be a bonding interaction. The Al(DDP) ligand is almost symmetrically located between the two Pd centers, exhibiting Pd-Al bond distances of Å (Pd1-Al) and Å (Pd2-Al), 136

153 III. esults and discussion [{Pd(dvds)} 2 {μ 2 -Al(DDP)} respectively. These bond distances are significantly longer than the terminal bond distances in the monomeric compound [Pd(AlCp*) 4 ] (2.2950(9) Å) [218] or the dimeric cluster [Pd 2 (AlCp*) 5 ] (Pd-Al: 2.32 Å). [55] Similar bond distances can be found for the μ 2 -bridging AlCp* ligands in the latter (1.44 Å), which also exhibit a nearly symmetric Pd-Al-Pd bridge. Due to unsymmetric coordination of the μ²-alcp* units in [Pd 3 (AlCp*) 2 (μ²-alcp*) 2 (μ 3 -AlCp*) 2 ], [55] one distance was found to be significantly longer (2.47 Å Å), whereas the associated one is similar (2.40 Å 2.42 Å). Table 37: Selected bond lengths (Å) and angles ( ) for [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23). Pd1-Pd (7) C201-C (9) Pd1-Al (18) C203-C (9) Pd2-Al (18) C101-C (9) Pd1-Al-Pd (5) C103-C (9) 1-Al (2) The two planes, each formed by the olefinic carbon atoms of one dvds ligand and one Pd center are almost exactly perpendicular to each other (torsion angle 89 ), therefore forcing the Al(DDP)-ring slightly out of plane (deviation 28.7 ). The C=C bond lengths of 1.40 Å and 1.42 Å are similar to the ones observed in the monomeric complexes [(dvds)pd{e(ddp)] (E = Al (16): 1.405(4) Å and 1.411(4) Å; E = Ga [159] : Å and Å) and also comparable to the monomeric HC complex [(HC)Pd(dvds)] (HC = (Mes) 2 C 3 H 2 ), C=C: Å and Å). [250] This results point to a rather high degree of π-backbonding in 23, again exemplifying Al(DDP) to be not only isolobal to HC's but also beeing a similarly strong σ-donor ligand. As mentioned above, the metal-to-ligand ratio is crucial in this reaction. A Pd:Al ratio of 2:1 leads to the formation of the dimeric, Al(DDP)-bridged complex [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23), whereas a Pd:Al ratio of 1 : > 1 yield in the formation of [(dvds)pd{ga(ddp)}] (16). Thereupon, the question arises, if the formation of monomeric 16 proceeds via the dimeric intermediate 23, simply by addition of another Al(DDP) ligand to the Pd 2 Al-core. To clearify this point, the reaction of 23 with one additional equivalent of Al(DDP) was performed, indeed showing the formation of 16 (Scheme 44). It should be noted, that only on addition of an excess of Al(DDP) a complete reaction of the dimer to the monomer was observed, 137

154 III. esults and discussion [{Pd(dvds)} 2 {μ 2 -Al(DDP)} indicating an association-dissociation equilibrium in solution at ambient temperature, which is shifted towards the monomer on increase of the Al(DDP) concentration. [Pd 2 (dvds) 3 ] + 1 eq. Al(DDP) Me 2 Si O Si Me 2 Pd Al 23 Pd SiMe 2 O SiMe E(DDP) >2 eq. Al(DDP) or 1 Ga(DDP) E Pd Me 2 Si O Si Me 2 E = Al (16), Ga Scheme 44: eactivity of [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23) towards E(DDP). Similarly, compound 23 also reacts with Ga(DDP) to give a 1:1 mixture of 16 and [(dvds)pd{ga(ddp)}], as proven by 1 H-M spectroscopy. 138

155 III. esults and discussion [Pd 2 (GaCp*) 4 {Al(DDP)}] Ligand substitution in [Pd 2 (dvds) 2 (μ 2 [Lit. 5, Chapter VII.3] -Al(DDP)] (23). As seen in the case of GaCp*, the formation of di- and trimeric cluster complexes of palladium can be readily performed by substitution of the coordinated olefins in [Pd 2 (dvds) 3 ]. [55] The reaction with E(DDP) (E = Al, Ga) also leads to substitution, but only the bridging dvds ligand can be replaced. For Al(DDP), the intermediate [Pd 2 (dvds) 2 (μ 2 -Al(DDP)] (23) can be isolated (see previous chapter), which reacts further with a second E(DDP) to give the respective monomeric complexes [(dvds)pd{e(ddp)}]. For the dimeric complexes [M 2 (GaCp*) 5 ], several ligand exchange reactions can be performed at the Ga(DDP)-bridged Pd 2 core. Thus, the terminal GaCp* ligands can be replaced by other ligands L e.g. phosphines, CO or t BuC to give [M 2 (GaCp*) 3 (L) 2 ]. Also, the bridging GaCp* ligands undergo substitution with AlCp* to give [Pt 2 (AlCp*) 3 (GaCp*) 2 ], again retaining the M 2 core. These substitution reactions are not limited to ECp*. The coordinated dvds ligand in E(DDP) containing compounds like [(dvds)pd{ga(ddp)}] can also be replaced by e.g. strong π-acceptor ligands. Thus, the dvds-ligands in 23 might be suitable for ligand-exchange reactions as well. Si O Si Pd Al Al + 4 GaCp* Pd Si O -2 dvds *CpGa Pd Si Ga Cp* Pd GaCp* GaCp* Scheme 45: Synthesis of [Pd 2 (GaCp*) 2 (μ 2 -GaCp* ) 2 {μ 2 -Al(DDP)}] (24). The reaction of 23 with an excess of GaCp* was performed in hexane at ambient temperature. Upon addition of GaCp*, the former yellow solution turns into deep red and the dimeric cluster compound [Pd 2 (GaCp*) 2 (μ 2 -GaCp* ) 2 {μ 2 -Al(DDP)}] (24) is formed (Scheme 45). Compound 24 is the first example of a heterobimetallic cluster exhibiting both group 13 DDP and Cp* ligands. Also, a sterically encumbered Al(DDP) ligand located in a bridging position between to metal centers was only once reported so far, confirming once more the strong preference of Al(I) for a bridging coordination mode, as already seen for AlCp*. [55] 139

156 III. esults and discussion [Pd 2 (GaCp*) 4 {Al(DDP)}] The 1 H-M spectrum of 24 at room temperature shows the typical set for a symmetric coordinated Al(DDP) moiety. In contrast to the related compound [Pd 2 (GaCp*) 5 ], two sets of signals for the bridging and the terminal GaCp* moieties (1.96 ppm and 1.97 ppm, 30H each) can be observed. However, on heating the solution to 60 C, only one signal for the Cp* moieties can be observed, pointing to a fast exchange of the bridging and the terminal GaCp* ligands on the M timescale, as previously discussed for the [M 2 (ECp*) 5 ] congeners. [55] At -80 C in toluene-d 8 a splitting of the GaCp* signals is found giving four distinguishable resonances (1.92, 1.94, 1.98 and 2.02 ppm), suggesting an asymmetric solution structure similar to the molecular structure observed in the solid state (vide infra). Complex 24 crystallizes in the monoclinic space group P2(1)/n. Similar to the homoleptic cluster compounds [M 2 (ECp*) 5 ], the molecular structure of 24 (Figure 31) consists of a central Pd 2 unit surrounded by two terminal and two bridging GaCp* as well as one bridging Al(DDP) ligand, leading to a distorted, dipalladium-centered trigonal-bipyramidal structure. The Pd-Pd bond distance is significantly shorter (2.582 Å) compared to 23, the homoleptic compound [Pd 2 (GaCp*) 2 (μ 2 -GaCp*) 3 ] (2.609 Å) or the CO-bridged complex [{( 2 PC 2 H 4 P 2 )Pd} 2 (μ 2 -CO)] (2.71 Å). [251] As pointed out above, the Al(DDP) ligand is located in a bridging position. This fact is consistent with ligand exchange reactions with AlCp* in M 2 (GaCp*) 5 (M = Pd, Pt) giving [Pt 2 (GaCp*) 2 (μ 2 -AlCp*) 3 ] and [Pd 2 (AlCp*) 2 (μ 2 -AlCp*) 3 ] (vide supra), respectively. In contrast to 23, the Al(DDP) ligand in 24 is unsymmetrically coordinated between the two Pd centers with Pd-Al bond distances of 2.457(3) Å (Pd1-Al) and 2.559(3) Å (Pd2-Al), respectively. For selected bond lengths and angles see Table 38. Whereas the Pd-Al bond distances in [Pd 2 (AlCp*) 5 ] are with av Å slightly shorter (due to symmetric coordination of the AlCp* units), a similar unsymmetric coordination can be found in [Pd 3 (μ 3 -AlCp*) 2 (μ 2 -AlCp*) 2 (AlCp*) 2 ], also showing significantly shorter bond distances (2.41 Å and 2.49 Å) for the μ 2 -bridged AlCp* moieties. [55] Similar Pd-Al bond distances are solely reported for the μ 3 -bridging AlCp* units in the latter complex (2.55 to 2.59 Å). The terminal Pd-Ga bond lenghts in 24 (2.376 and Å) are slightly longer than in [Pd 2 (GaCp*) 5 ] (2.358 and Å) or the monomeric compound [Pd(GaCp*) 4 ] (2.366 Å), [54] but comparable to those in [Pd 3 (GaCp*) 8 ] (2.399 and Å). [144] The bridging Pd-Ga bonds are distinctly longer (2.496 and Å) than the terminal, but similar to those in 140

157 III. esults and discussion [Pd 2 (GaCp*) 4 {Al(DDP)}] [Pd 2 (GaCp*) 5 ] (2.494 and Å). The Ga2-Pd2-Pd1 unit is almost linear, whereas the Pd2- Pd1-Ga1 angle is significantly distorted (168.7 ), possibly an effect of the unsymmetric orientation of the Al(DDP) ligand. Figure 31: Molecular structure of [Pd(GaCp*) 4 {μ 2 -Al(DDP)}] (24). Table 38: Selected bond lengths (Å) and angles ( ) for [Pd(GaCp*)4{μ2-Al(DDP)}] (24). Pd1-Pd (15) Pd2-Ga (15) Pd1-Al 2.456(3) Pd2-Al 2.559(3) Pd1-Ga (17) Al (8) Pd1-Ga (17) Al (8) Pd1-Ga (15) Pd2-Ga (17) Ga1-Pd1-Pd (5) Pd2-Ga (2) Pd1-Pd2- Ga (5) 1-Al (4). 141

158 III. esults and discussion [Pd 2 (GaCp*) 4 {Al(DDP)}] Summary. In contrast to the homoleptic complexes [M(ECp*) 4 ], which are found to be kinetically inert towards a variety of ligands (P 3, CO, isocyanides, alkenes etc), the Pd(0) compound [(dvds)pd{ga(ddp)}] readily reacts with the strong π-acceptor ligand t BuC to give the dimeric cluster compound [Pd{μ 2 -Ga(DDP)}( t BuC)] 2 (22). Similar to the related CO complex [Pd{μ 2 -Ga(DDP)}(CO)] 2, the Ga(DDP) ligands are bridging the two Pd centers in a µ 2 -coordination mode. The difference between Al(I) and Ga(I) ligands, as previously reported for the E I CP* congeners, again becomes visible in the case of E(DDP). Whereas the reaction of [Pd 2 (dvds) 3 ] with Ga(DDP) only leads to the monomeric complex [(dvds)pd{ga(ddp)}], the 1:1 reaction with Al(DDP) gives the dimeric complex [Pd 2 (dvds) 2 (μ 2 -Al(DDP)] (23), with the Al(DDP) being located in a bridging position between the two Pd(0) centers. Only on addition of an excess of E(DDP) to 23, the related monomeric complex [(dvds)pd{ga(ddp)}] (16) can be obtained. This result is in well agreement with the previously described strong preference of AlCp* to bridge transition metal centers, as the substitution of the bridging GaCp* units in [Pt 2 (GaCp*) 5 ] to give [Pt 2 (µ 2 -AlCp*) 3 (GaCp*) 2 ] reveals. According to this, stronger σ-donor abilities towards transition metal centers can be suggested for Al(DDP) than for Ga(DDP). Moreover, the substitution of the terminal coordinated dvds-ligands in 23 by GaCp* leads to the formation of [Pd 2 (GaCp*) 2 (μ 2 -GaCp* ) 2 {μ 2 -Al(DDP)}] (24), the first example of a heterobimetallic cluster with both E(DDP) and ECp* ligands. In contrast to the related [Pt 2 (GaCp*) 5 ], fluxionality of the GaCp* units can be observed only at higher temperature (60 ). The increased steric demand of the DDP vs. Cp* ligand at the Al(I) center is a reasonable explanation for this. 142

159 III. esults and discussion Dimeric compounds with a Pt 2 core [Lit. 7, Chapter VII.3] Dimeric compounds with a Pt 2 core eaction of [(1,3-cod)Pt{Ga(DDP)} 2 ] with t BuC. As seen in chapter III , the monomeric compound [(dvds)pd{ga(ddp)}] readily reacts with strong π-acceptor ligands like CO or t BuC to give the dimeric clusters [Pd{μ 2 -Ga(DDP)}(L)] 2 (L = CO, t BuC (22)) in high yields. Whereas the reaction of [(1,3-cod)Pt{Ga(DDP)}] with CO gives the analogue [Pt{μ 2 -Ga(DDP)}(CO)] 2, it is likewise to examine the reaction of [(1,3-cod)Pt{Ga(DDP)}] with t BuC. [(1,3-cod)Pt{Ga(DDP)} 2 ] xs. C t Bu - 1,3-cod -[(DDP)Ga( t Bu)(C)] Ga Pt Pt C t Bu Ga Bu t C 25 Scheme 46: Synthesis of [Pt{μ 2 -Ga(DDP)}( t BuC)] 2 (25). As expected, the olefin in [(1,3-cod)Pt{Ga(DDP)}] can also be substituted by isocyanide ligands. The dimeric compound [Pt{μ 2 -Ga(DDP)}( t BuC)] 2 (25) is formed on reaction with t BuC and can be isolated as deep red crystals in high yields according to Scheme 46. Likewise 22, compound 25 is insoluble in all common organic solvents and thus no solution M spectroscopic data are available. The molecular structure of 25 was confirmed by x-ray and is discussed below, the crystals being obtained by slow diffusion of t BuC into a solution of [(1,3-cod)Pt{Ga(DDP)}] in hexane or THF. When the reaction of [(1,3-cod)Pt{Ga(DDP)}] and t BuC is monitored by 1 H M spectroscopy in solution, the formation of a side product is revealed, which can be identified as [(DDP)Ga( t Bu)(C)] (8). Obviously, Ga(DDP) is liberated in the course of the reaction, which readily reacts with t BuC giving the oxidative addition product 8. otably, Ga(DDP) reacts with t BuC giving 8 by insertion of the gallium diyl into (CH 3 ) 3 C also in the absence of any platinum containing compound, as already discussed in chapter III

160 III. esults and discussion Dimeric compounds with a Pt 2 core Compound 25 crystallizes in the monoclinic space group C2/m. The molecular structure of 25 is depicted in Figure 32 and is similar to the one found for to the carbonyl compound [Pt{Ga(DDP)}(CO)] 2, [159] showing two Pt centers bridged by two Ga(DDP) ligands with an almost ideal square planar arrangement (angular sum: ). Furthermore, it can be regarded as the Ga(DDP) analogue of the compound [Pt 2 ( t BuC) 2 (μ 2 -GaCp*) 3 ], with the steric demand of the Ga(DDP) ligand preventing the coordination of a third Ga(DDP). Selected bond lengths and angles are shown in Table 39. Figure 32: Molecular structure of [Pt( t BuC){μ 2 -Ga(DDP)}] 2 (25). For structural parameters see Table 39. The Pt-Ga distances are almost equal (Pt1-Ga1: Å; Pt1*-Ga1: Å) and slightly shorter than in the CO analogue (2.49 Å to 2.52 Å). Similar bond length are reported for the dimeric complexes [Pt 2 (μ 2 -GaCp*) 3 (L) 2 ] (L = GaCp* [143] ; Pt-Ga: 2.45 Å to 2.47 Å; L = Ph 3 P [55] ; Pt-Ga: 2.46 to 2.49 Å). The Ga- bond length are somehow elongated (1.995 Å) 144

161 III. esults and discussion Dimeric compounds with a Pt 2 core compared to the CO analogue (1.953 Å), but similar to the previously discussed Ga- bond distances of coordinated Ga(DDP) ligands. Also, the Pt-C distance is shortend (1.897 Å) with respect to [Pt{Ga(DDP)}(CO)] 2 (1.968 Å), but comparable to those in other Pt-isonitril complexes (e.g. in [Pt 3 ( t BuC) 3 (μ 2 - t BuC) 3 ]: 1.88 to 1.92 Å). [252] The C bond distance (1.078 Å) is rather short, pointing to a low degree of π-backbonding in 25 caused by the bridging mode of the 2e-donor Ga(DDP)-ligands. Table 39: Selected bond lengths (Å) and angles ( ) for [Pt{μ 2 -Ga(DDP)}( t BuC)] 2 (25). Pt-Ga (11) 1-Ga-1* 92.4(2) Pt-Ga* (12) 1-Ga-Pt (11) Pt-Pt* (8) 1-Ga-Pt* (10) Ga (4) Ga-Pt-Ga* (3) Ga-1* 1.995(4) Pt-Ga-Pt 64.73(3) Pt-C (7) C20-Pt-Pt* 174.1(2) C (9) Pt-C (8) 145

162 III. esults and discussion Dimeric compounds with a Pt 2 core Synthesis of the dimeric hydride complex [Pt{μ 2 -Ga(DDP)}(H 2 )] 2 (26). In the hydrogenation reaction of [(1,3-cod)Pt{Ga(DDP)} 2 ], both Ga(DDP) ligands remain coordinated to the platinum center forming the trans-dihydride complex 17 as the thermodynamically most stable product. How about hydride complexes having only one Ga(DDP) ligand coordinated to the platinum center, in accordance to the substitution products 25 and the respective CO complex? Since a diolefin complex with only one Ga(DDP) ligand similar to [(dvds)pd{ga(ddp)}] or [i(olefin) 2 {Ga(DDP)}] (olefin = C 2 H 4 (13), styrene (14), dvds (15)) is still unknown for platinum, a 1:1 mixture of [Pt(cod) 2 ] and Ga(DDP) in THF was treated in situ with hydrogen by applying increased pressure (2 bar) of hydrogen to the degassed solution. After stirring the deep red reaction mixture for 6 h at room temperature, a yellow precipitate of [Pt{μ 2 -Ga(DDP)}(H) 2 ] 2 (26) is formed (Scheme 47). Compound 26 is soluble in common organic solvents like hexane, THF and benzene, but stable only for several days in an inert gas atmosphere, even at -30 C, and decompose in air or moisture immediately. [Pt(cod) 2 ] H H + Ga(DDP), H 2 Pt Ga Ga - 2 C 8 H 16 H Pt H 26 Scheme 47: Synthesis of [Pt{μ 2 -Ga(DDP)}(H) 2 ] 2 (26). As expected, the Ga(DDP) moiety gives rise to a set of signals for a C 2v symmetric ligand, i.e. the septet signal of the i Pr-CH at 3.35 ppm and two doublet signals at 1.11 ppm and 0.80ppm, respectively, as seen in the 1 H-M spectrum. Together with the broad signal at ppm, which exhibis an integration of 4H with respect to the heterocycle-ch proton signal of the DDP-ligand at 4.89 ppm, this observation indicates the coordination of two hydrides and one Ga(DDP) moiety to each Pt center. The 13 C-M spectrum is in good agreement to this, again showing only signals for a Ga(DDP) moiety with local C 2v symmetry. 146

163 III. esults and discussion Dimeric compounds with a Pt 2 core Figure 33: Molecular structure of [Pt{μ 2 -Ga(DDP)}(H) 2 ] 2 (26). Table 40: Selected bond lengths (Å) and angles ( ) for [Pt{μ 2 -Ga(DDP)}(H) 2 ] 2 (26). Pt-Ga (9) Pt-Ga-Pt* 74.09(3) Pt-Ga* (10) Ga-Pt-Ga* (3) Pt-Pt* (8) 1-Ga (19) Ga (4) Ga (5) The mercury plot of the molecular structure of the dimeric tetrahydrido complex 26 is shown in Figure 33, whereas selected bond lengths and angles can be found in Table 40. Crystals suitable for x-ray analysis could be obtained by recrystallization of the crude product from THF at -30 C. Compound 26 crystallizes in the monoclinic space group P2 1 /n. The molecular structure shows a coplanar arrangement of the Pt- and the Ga-centers with the Ga(DDP) being in a bridging position between the two Pt-centers (Pt-Ga distances 2.46 and 2.48 Å), similar to [Pt{μ 2 -Ga(DDP)}( t BuC)] 2 (25). Whereas the hydrides could not be located in the refinement of the structure, the Pt-Pt distance can be measured to be Å and thus, significantly longer than in 25 (2.634 Å), the related carbonyl complex [Pt{Ga(DDP)}(CO)] 2 (2.614 Å), or the dimeric compound [Pt 2 (GaCp*) 5 ] (2.58 Å). The Ga- distances in 26 are Å and Å and therefore quite similar to [Pt{Ga(DDP)}(CO)] 2 ( Å). 147

164 III. esults and discussion Compound with a i 3 -core [Lit. 8, Chapter VII.3] 4.2. Trinuclear compounds Cluster compounds with a i 3 core. As mentioned in chapter III , the reaction of Ga(DDP) with [i(c 2 H 4 ) 3 ] leads to the dimmeric, Ga(DDP) bridged i 2 cluster compounds [{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19) and [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 -C 2 H 4 )] (20), respectively. However, when heating the reaction mixture above room temperature (e.g. 60 C), decomposition of the latter complexes occurs accompanied by liberation of ethylene and the trimeric cluster [{{μ 2 - Ga(DDP)}i(C 2 H 4 )} 2 i(μ 2 -CH=CH 2 )(H)] (27) is formed, showing a i/ga ratio of 3:2. It should be noted, that the reaction of [i(c 2 H 4 ) 3 ] and Ga(DDP) always leads to the formation of 27 upon heating to ca. 60 C, independent of the ratio of the reactants (1:1, 2:1, 3:2) (Scheme 48). This result suggests, that the decomposition of 19 proceeds via 20 and finally lead to 27, which is the thermodynamic product in these cluster formation processes and thus, the most stable i cluster in this series. H i + Ga(DDP) 2 i( ) 3 Ga H i H H + i Ga H H i H H ΔT i i Ga i H Ga 27 Scheme 48: Synthesis of [{{μ 2 -Ga(DDP)}i(C 2 H 4 )} 2 i(μ 2 -CH=CH 2 )(H)] (27). Compound 27 can readily prepared by adding two equivalents of Ga(DDP) to a cooled solution of three equivalents of [i(c 2 H 4 ) 3 ] (-5 to -10 C) (in situ prepared by dissolving

165 III. esults and discussion Compound with a i 3 -core mg (0.307 mmol) of i(cdt) in 5ml of cold hexane (-10 C) and applying 3 bar of ethylene for 1h) in hexane in ethylene atmosphere. When the reaction mixture is allowed to warm up to room temperature, heated to ca. 60 C for several minutes and cooled to room temperature again, the formation of a deep red solid occurs after 1-2 hours. ecrystallisation from hot hexane leads to the formation of deep red needles of 27 in high yields (Scheme 48). Alternatively, 27 can also be prepared by reaction of [i(cdt)] and [i(c 2 H 4 ) 2 {Ga(DDP)}] (13) (3:2) in ethylene atmosphere upon warming to ca. 60 C, as shown by 1 H M spectroscopy. The 1 H-M-spectrum of 27 shows asymmetrically (C s ) coordinated Ga(DDP) ligands, indicated by significant splitting of the i Pr-Me groups to eight distinct doublets (see experimental section), whereas the two terminal bonded ethylene ligands are equivalent. They appear as a multiplet signal centered at 3.26 ppm (8H). The bridging vinyl ligand gives rise to three distinct signals in the expected regions, whereas in contrast to the complexes 19 and 20, a hydrid signal is found at ppm, indicating an activation of the bridging ethylene molecule in the course of the reaction. The 13 C-M spectrum also shows the expected signals. Compound 27 was found to be rather stable in solution as well as in the solid state. Also, variable temperature M spectroscopy does not show any changes in the M spectra in the range of -60 C up to +70 C. Figure 34: Molecular structure of [{{μ 2 -Ga(DDP)}i(C 2 H 4 )} 2 i(μ 2 -CH=CH 2 )(H)] (27). For structural parameters see Table

166 III. esults and discussion Compound with a i 3 -core Deep red needles of 27 suitable for x-ray single crystal diffraction studies were grown by cooling a hot saturated solution of 27 in hexane to room temperature. Complex 27 crystallizes in the orthorhombic space group Pccn. Selected bond lengths and angles are shown in Table 41. The molecular structure in the solid state consists of a triangular arrangement of the i atoms (i-i-i: 90.0 ), with two short (2.5531(5) Å) and one longer i-i contact (3.611 Å). The two short i-i bonds are bridged by Ga(DDP) ligands, with the Ga-i bond distances of 2.29 Å (to the central i2) and 2.38 Å (to the terminal i1) being distinctly longer than the terminal i-ga distances found in [i(gacp*) 4 ] (i-ga: 2.21 Å), [136] but also considerably longer than the distances for bridging ligands found in [{(C(SiMe 3 ) 3 )Ga} 2 (icp) 2 ] (Ga-i: 2.28Å). [88] Each terminal i-center (i1 and its symmetry equivalent) are coordinated by an ethylene molecule with a C=C double bond distance of 1.40 Å, which is similar to those found in 13 or [(Cy 3 P)i(C 2 H 4 ) 2 ] (1.41 and 1.39 Å). [228] The hydride ligand could not be located in the refinement. The vinyl group is σ-bound to the central i2, which presumably also bears the hydride ligand, indicated by the open space trans to the vinyl group. The vinyl group is π-coordinated by both terminal nickel atoms, placing it exactly in between these two metal centers, with a relatively long C-C distance of 1.44 Å, being a result of the strong π-backbonding. The vector of the vinylic C=C bond is tilt towards the plane spun by the three nickel atoms by an angle of 28, which leads to the C s symmetry of the Ga(DDP) ligands found in the 1 H-M spectrum. Table 41: Selected bond lengths (Å) and angles ( ) for [{{μ 2 -Ga(DDP)}i(C 2 H 4 )} 2 - i(μ 2 -CH=CH 2 )(H)] (27). Ga-i (5) Ga (2) Ga-i (3) C30-C (5) i1-i (5) C40-C (7) i1-c (3) i1-c (4) 1-Ga (9) i1-c (7) i1-ga-i (18) i1-c (6) i1-i2-i1* 90.02(2) i2-c (6) C40-C41-i (4) Ga (2) 150

167 III. esults and discussion Compound with a i 3 -core Preliminary DFT calculations were performed on the B3LYP/LAL2DZ level of theory, with [Ga{(MeCMe) 2 CH}] as a model ligand. As a result, the zero point energy difference between activated and non-activated ethylene molecules in the i 3 Ga 2 cluster 27 show a preference for the C-H activated species by 20.1 kcal/mol, which is in good accordance to the 1 H-M spectrum of 27, which accordingly shows the expected hydride signal at ppm (vide infra). Summary. The high steric demand of the DDP moiety prevents the formation of kinetically inert homoleptic i/ga complexes or clusters and thus allows the coordination of additional olefin ligands. Thus, the two nickel-gallium clusters with a i 2 Ga core ([{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19) and [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 -C 2 H 4 )] (20)), synthesized by the reaction of i(c 2 H 4 ) 3 with Ga(DDP) in a molar ratio of 2:1, exhibit one ethylene molecule in a π-/agostic mixed coordination mode, bridging the two nickel centers. Additionally, the Ga(DDP) ligand is also coordinated in a bridging coordination mode between the two ni centers. The complexes 19 and 20 are very reactive at ambient temperature and decompose (also on heating to 60 C) under liberation of ethylene and formation of the i 3 Ga 2 complex [{{μ 2 -Ga(DDP)}i(C 2 H 4 )} 2 i(μ 2 -CH=CH 2 )(H)] (27). The concomitant decrease of the i-c bond distance of the bridging ethylene during the expected decomposition process (19 20 ) finally leads to a C-H activation of the bridging ethylene molecule, giving the vinyl-hydride cluster 27. These results show, that the concept of tuning the reactivity of mixed metal cluster complexes by an increase of the steric bulk of the ligands E I has been successfully applied. The presence of the strongly electron donating ligand Ga(DDP) which is additionally able to bridge the nickel centers and thus establishes the cluster formation, is crucial for the activation reaction. The chemistry of the novel i/ga mixed metal clusters indicate a potential for further studies directed towards the activation of small molecules such as olefins at transition metal centers supported by carbenoid group 13 metal ligands. Moreover, these mixed metal cluster compounds can serve as model complexes for alloy materials and thus lead towards tailored alloy compositions for more controlled and selective reactivities of heterogenous catalysts. 151

168 III. esults and discussion The cluster [Sn 17 {ClGa(DDP)} 4 ] 4.2.2: Synthesis of the large cluster [Sn 17 {ClGa(DDP)} 4 ]. Polyhedral homoatomic group 14 cluster compounds, first observed as the anionic [Pb 9 ] 4- species in 1890 by Joannis, are in the focus of research for many years. [253] These compounds are known as Zintl ions after the German chemist Eduard Zintl, whose work was dedicated to the understanding of such ligand-free polyanions. The Zintl ions are generally formed by reduction of the corresponding element with alkali metals, leading to a negative average oxidation state of the group 14 element. After the first ligand-stabilized polyhedral cluster compound of the general formulae (E) n E = Si, Ge, Sn; = ligand) was synthesized by Matsumoto et al. in 1988, [254] this new field developed rapidly, leading to several clusters with up to ten tetrel atoms in the cluster core. In the last few years a third class of group 14 cluster compounds, the metalloid cluster compounds of the general formulae E n m with n > m (E = Si, Ge, Sn, Pb; = ligand), could be established, in which naked, i.e. not ligand stabilized, metal atoms are present as well as ligand-bound metal atoms and the number of metal metal bonds exceeds those to the ligands. The average oxidation state of the metal atoms in the cluster core of such clusters is between 0 and 1, thus, the metalloid clusters can be regarded as being intermediates on the way to the corresponding element. [255] Up to now, only few examples for tin-containing clusters are known, including the Zintl-anions Sn 2-5 and Sn 4-[256] 9 and clusters with coordinated aryl or silyl ligands like [Sn 8 4 ], [257] [Sn 8 6 ], [258] [Sn 9 3 ], [Sn 10 3 ], [259] as well as the largest reported cluster so far, [Sn 15 Z 6 ] (Z = (2,6- i Pr 2 C 6 H 3 )(SiMe 2 X); X = Me, Ph). [260] Several synthetic strategies to synthesize such clusters are known, including the reductive coupling of EX or EX 3 precursors ( = bulky ligands) with an adequate reducing agent to give metalloid group 14 cluster compounds. It was found, that the ligand was thought to be necessary for the protection of the cluster core against the exterior influences, e.g. agglomeration and cluster/particle growth. [255] 152

169 III. esults and discussion The cluster [Sn 17 {ClGa(DDP)} 4 ] In this respect, with its reducing ability, as seen in the insertion into different metal-halogen bonds, and its steric demand, which is suitable for stabilizing reactive intermediates, as seen in the formation of the h complex [(Ph 3 P) 2 h{ga(ddp)}(μ-cl)], the use of the sterically demanding Ga(DDP) ligand might provide another suitable way for the formation of group 14 cluster compounds. Thus, Ga(DDP) readily inserts into the Sn-Cl bond and forms the stable Sn(IV) compound [Me 2 Sn{ClGa(DDP)} 2 ] (6) in high yields. However, the use of Sn(IV) compounds seems to be inappropriate for such cluster formation, since the reported syntheses of the known Sn x -clusters (x = 8, 9, 10, 15) all involve the use of a Sn(II) precursor. Therefore, SnCl 2 was used as a precursor for the reaction with Ga(DDP). Cl (DDP)Ga Sn Cl Ga(DDP) Sn Sn SnCl2 + excess Ga(DDP) THF - (DDP)GaCl 2 Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn Sn (DDP) Ga Sn Sn Sn Ga(DDP) Cl Cl [Sn 17 {ClGa(DDP)} 4 ] (28) Scheme 49: Synthesis of [Sn 17 {ClGa(DDP)} 4 ] (28). When mixing an excess (two equivalents) of Ga(DDP) with one equivalent of SnCl 2 in THF at -30 C, the colour of the solution turns into deep red / violet after several seconds. Stirring the solution for 60 minutes, concentrating followed by storing the solution at -30 C overnight afforded once deep red crystals of [Sn 17 {ClGa(DDP)} 4 ] (28) in moderate yields (Scheme 49). Whereas other examples of Sn 17 clusters with interstitial metal atoms in the Sn 9 subunits are known (e.g. the ion [i 2 Sn 17 ] 4- ) [261], compound 28 represents the largest polyhedral homoatomic group 14 cluster compound reported so far. However, decomposition of the compound occurs, as indicated by a colour change of the deep red / violet solution into brown when storing it for more than 12 h, also at -30 C. During this process, an insoluble solid precipitates which could not yet analyzed in more detail. Also, several attempts to reproduce [Sn 17 {ClGa(DDP)} 4 ] (28) in different solvents or at different reaction conditions failed so far, leading to the brown solution and the insoluble solid in all cases. 153

170 III. esults and discussion The cluster [Sn 17 {ClGa(DDP)} 4 ] evertheless, single crystal x-ray diffraction analysis was performed on a deep red crystal of 28. Selected bond lengths and angles are shown in Table 42, whereas the molecular structure is shown in Figure 35. Figure 35: Molecular structure of [Sn 17 {ClGa(DDP)} 4 ] (28). For structural parameters see Table 42. It can be best described as two coupled Sn 9 clusters that share a common vertex (Sn3). The two Sn 9 subunits are crystallographic equivalent and both consist of a distorted tricapped trigonal prism, with the trigonal prisms coordinated perpendicular towards each other (dihedral angle between Sn1-Sn9-Sn3 and Sn1*-Sn9*-Sn3: ) and the common vertex Sn atom being one of the capping ligands (Figure 36). The other capping Sn atoms bearing a {ClGa(DDP)} moiety (Sn4, Sn6), which might result from the insertion of Ga(DDP) into the Sn-Cl bond of a SnCl 2 moiety, which can be suggested to be an intermediate in the reaction. However, the distortion of the trigonal prism from D 3h symmetry is expressed in the difference of the three prism heights: Sn1-Sn9 = Å, Sn2-Sn5 = Å and Sn7-Sn8 = Å. Also, the edge lengths are different, ranging from Å (Sn5-Sn7) to Å 154

171 III. esults and discussion The cluster [Sn 17 {ClGa(DDP)} 4 ] (Sn1-Sn2). As recently described for other Sn 9 complexes, the distances between ligand bound and naked tin atoms are suggested to be shorter than those between naked atoms. [255] Figure 36: view of the Sn17-core in [Sn 17 {ClGa(DDP)} 4 ] (28). In accordance to this, the capping Sn atoms bearing a {ClGa(DDP)} moiety (Sn4 and Sn6) showing Sn-Sn bond distances of 2.87 Å 2.95 Å. However, similar short bond distances are found for the naked Sn bond distances Sn5-Sn7 and Sn2-Sn8 (2.90 Å and 2.94 Å). Expectedly, the bond distances of the central Sn3 atom, which is located in the center of a distorted cube, to the corners of this cube are longer and within a range of 3.08 Å to 3.13 Å These bond lengths are close to the average value of 3.10 Å in gray tin (α-sn, diamond lattice) and similar to the ones found in the Sn 15 Cluster [Sn 15 Z 6 ] (Z = (2,6- i Pr 2 C 6 H 3 )(SiMe 2 X); X = Me, Ph) (av Å). [260] Additionally, the edge length of this distorted cube varies from short (av Å, Sn5-Sn7 and Sn2-Sn8) to long Sn-Sn interactions (3.23 Å 3.36 Å (Sn8-Sn8*, Sn5-Sn5*, Sn7-Sn2* and Sn2-Sn7*), as well as non-bonding interactions (4.64 Å and 4.44 Å; Sn5-Sn2 and Sn7-Sn8). Similar long Sn-Sn bond distances of 3.15 Å and 3.65 Å are also found in the aforementioned Sn 15 cluster. The Sn-Ga bond distances are almost identical with Å and 2.59 Å, which is significantly shorter than the reported Sn-Ga bond distances in the anionic complexes [ 2 SnGa{[(Ar)C(H)] 2 }] - (2.72 Å) or the neutral complexes [(Giso)SnGa{[(Ar)C(H)] 2 }] (2.68 Å) [67] and [Me 2 Sn{ClGa(DDP)} 2 ] (6) (av Å). As expected, the Ga-centers are coordinated in a distorted tetrahedral fashion by the chlorine, a 155

172 III. esults and discussion The cluster [Sn 17 {ClGa(DDP)} 4 ] tin and two nitrogen atoms, respectively. The Ga-Cl bond distances of Å and Å are comparable to the ones found in free [(DDP)GaCl 2 ] (2.22 Å) [175] or the tin compound 6 (2.247 Å). The Ga- bond distances are rather short (1.93 Å to 1.96 Å), which again points to an electrophilisation of the ligand in 28. Table 42: Selected bond lengths (Å) and angles ( ) for [Sn 17 {ClGa(DDP)} 4 ] (28). Sn1-Sn (14) Sn3-Sn (13) Ga10-Sn (16) Sn1-Sn (12) Sn3-Sn7* (10) Ga (11) Sn1-Sn (12) Sn3-Sn (10) Ga (12) Sn1-Sn (13) Sn4-Sn (13) Ga10-Cl (4) Sn1-Sn (13) Sn4-Sn (13) Ga11-Sn (17) Sn2-Sn (8) Sn4-Sn (13) Ga (11) Sn2-Sn (13) Sn5-Sn (12) Ga (11) Sn2-Sn (12) Sn5-Sn (13) Ga11-Cl (4) Sn2-Sn7* (14) Sn5-Sn (14) Sn3-Sn2* (8) Sn6-Sn (13) 1-Ga (5) Sn3-Sn (13) Sn7-Sn (13) 3-Ga (5) Sn3-Sn5* (13) Sn7-Sn2* (14) Cl1-Ga11-Sn (11) Sn3-Sn8* (13) Sn8-Sn8* (18) Cl2-Ga10-Sn (11) 156

173 IV. Conclusion and outlook. IV. Conclusion and outlook. This study mainly focuses on the synthesis and characterisation of new transition metal complexes and compounds of the sterically encumbered aluminium- and gallium-heterocycles E I (DDP) in the low oxidation state +I. Within the limits of this study, new monomeric and unsaturated transition metal compounds as well as complexes with higher nuclearity (two and more metal centers in the cluster core) are synthesized and characterised in more detail, thus authenticating the mentioned conjecture of E I (DDP) being suitable ligands in organometallic chemistry due to their increased steric demand (cone angle 147 vs. 112 ) as well as their different electronic properties compared to ECp*. In this respect, the formation of M-E bonds was achieved by using classical synthetic strategies of organometallic chemistry like inserting E I (DDP) into metal-halide and metal-methyl bonds or via substitution of easily replaceable ligands. 1. Insertion reactions of Ga(DDP) into metal-halide and metal methyl bonds. Generally, the Ga-center of the Ga(DDP) ligand becomes more electrophilic on coordination to a metal center. This fact is reflected in a decrease of the Ga- bond lengths observed in all complexes reported so far compared to those in free Ga(DDP) (2.05 Å). If a halide or methyl group X is present at the metal center, the electrophilic Ga-center competes with the metal for the electrons of the group X and, with only one exception, yields the full migration of the halide or methyl group to the gallium center to give new complexes with a coordinated {(DDP)GaX} motif. The known complexes of this type are summarized in Table 43. Additionally, besides single-insertion, also double insertions into M-X bonds are possible, if two or more ligands X are coordinated at the metal center, as seen in the formation of the compounds 2, 5, and 6. However, based on the results of this study, these double insertion reactions seem to depend on two different properties of the metal centers used; the size of the metal atoms (i.e. the covalent radius) and their coordination number. Thus, comparably small metal centers (e.g. Si, covalent radius of 117 pm) with a high coordination number of 4 only allow mono-insertions (compound 7) according to structure a in Scheme 50, whereas for bigger metal atoms like Sn (covalent radius: 140 pm), a double insertion is observed (structure b, represented by compound 6,). Consequently, if the metal atom has an average size (e.g. Ga, Zn, covalent radius 125 pm) and a low coordination number (coord. number: 2), a double insertion is preferred (structure c; compound 2), whereas an increase in coordination number 157

174 IV. Conclusion and outlook. complicates (coord. number: 3, structure d and e; compounds 4 and 5) or completely prevents a double insertion (coord. number: 4, structure a, compound 3). Ga X a M X X X + MX 4 X Ga X M Ga c + MX 2 + MX 3 Ga(DDP) X X M d X Ga Ga X X X M b X Ga + MX 4 + MX 3 Ga X X M e X Ga Scheme 50: structural motives of the products of insertion reactions into M-X bonds (X = CH 3, chloride, = 2,6- i Pr 2 -C 6 H 3 ), depending on the metal size and its coord. number. In contrast to the previously reported insertion chemistry of ECp* compounds (E = Al, Ga, In), that either show reduction of the metal (e.g. M 2+ to M 0 ) with formation of Cp*EX 2, Cp* transfer and coordination of EX 2 to the metal center or a halide atom bridging two or three coordinated GaCp* units in the complexes (Chapter II), the coordinated {(DDP)GaX} moiety is less acidic compared to its counterpart {Cp*GaX} and thus remains coordinated at the metal center. This lower acidity is also reflected in the comparably long Ga-Cl bond lengths (2.247 Å to Å in the complexes 1, 3 and 6) compared to those in free [(DDP)GaCl 2 ] (2.222 Å). 158

175 IV. Conclusion and outlook. Table 43: known compounds synthesized via insertion reactions into M-X bonds. Compound M / E Lit. [(COE)(η 6 -benzene)h{(ddp)gacl}] (1) 1 / 1 This study [(Ph 3 P) 2 h{ga(ddp)}(μ-cl] 1 / 1 [{(DDP)GaMe} 2 Zn] (2) 1 / 2 This study [{(DDP)GaCl}ZnCl(THF) 2 ] (3) 1 / 1 This study [{(DDP)GaMe}GaMe 2 ] (4) 1 / 1 This study [{(DDP)GaMe} 2 GaMe] (5) 1 / 2 This study [Me 2 Sn{ClGa(DDP)} 2 ] (6) 1 / 2 This study [Cl 3 Si{ClGa(DDP)}] (7) 1 / 1 This study [(DDP)Ga( t Bu)(C)] (8) 1 / 1 This study [(DDP)Ga( t Bu)(Cl)] (9) 1 / 1 This study [(Ph 3 P)Au{Ga(DDP)Cl}] 1 / 1 [{(DDP)Ga}Au{Ga(DDP)Cl}] 1 / 2 [(Ph 3 P)Au{Ga(DDP)Me}] 1 / 1 [81] [80] [80] [80] 2. Cationic complexes featuring M-Ga(DDP) bonds. The rather long Ga-Cl bond lengths (vide supra), accompanied with the observed fast chlorine exchange in 3 also indicate a comparably weak coordination of the chlorides to the gallium center, thus allows its abstraction by a[bar F ] to give cationic complexes of the type [L n M{Ga(DDP)} m ] + (L = additional ligand). These cationic complexes, exemplified shown by the isolation of the linear symmetric cationic complex [{(DDP)Ga} 2 Au][BAr F ] (10) from the reaction of [{(DDP)Ga}Au{ClGa(DDP)}] with a[bar F ] in flourobenzene, exhibit an rather strong electrophilicity of the coordinated Ga-center in the cation, that becomes visible in the axially coordination of a THF molecule to each gallium center in the isolated complexes [{(DDP)Ga. THF} 2 Au][BAr F ] (10. 2THF) and [{THF. Ga(DDP)}Zn(THF)(μ-Cl)] 2 [BAr F 2] (11) (Chapter III.2.). These results point to the general possibility to synthesize cationic transition metal-ga(ddp) complexes by abstraction of the respective chlorides in [L n M{ClGa(DDP)} m ] compounds by a[bar F ], which is generally accompanied by the increase in electrophilicity of the coordinated Ga-centers. 159

176 IV. Conclusion and outlook. 3. Monomeric complexes featuring M-Ga(DDP) bonds. The substitution of labile olefin ligands is another and very common way to establish TM-E I bonds. In this work, a variety of different monomeric transition metal complexes with either one or a maximum of two E(DDP) moieties coordinating the metal center are obtained depending on the transition metal used. Table 44 gives an overview of the known monomeric complexes synthesized via substitution of labile ligands by E(DDP), which are so far restricted to the d 10 metals i, Pd and Pt. Table 44: monomeric complexes synthesized via substitution of labile ligands by E(DDP). Compound M / E Lit. [{(DDP)Ga} 2 i(1,3-cod)] 1 / 2 This study [{(DDP)Ga}i(cdt)] (12) 1 / 1 This study [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13) 1 / 1 This study [{(DDP)Ga}i(styrene) 2 ] (14) 1 / 1 This study [{(DDP)Ga}i(dvds)] (15) 1 / 1 This study [{(DDP)Al}Pd(dvds)] (16) 1 / 1 This study [{(DDP)Ga}Pd(dvds)] 1 / 1 [{(DDP)Ga} 2 Pt(1,3-cod)] 1 / 2 [159] and [160] [159] and [160] trans-[{(ddp)ga} 2 Pt(H) 2 ] (17) 1 / 2 This study cis-[{(ddp)ga} 2 Pt(H)(SiEt 3 )] (18) 1 / 2 This study Whereas the olefins of the Lewis-acidic i(0) center in i(cdt) can not be replaced by the σ-donor ligand Ga(DDP) and instead leads to the formation of the 18VE Lewis acid base adduct [(cdt)i{ga(ddp)}] (12) in nearly quantitative yields, the substitution of labile olefins in other d 10 -metal olefin complexes like i(c 2 H 4 ) 3, Pd 2 (dvds) 3 or Pt(cod) 2 is readily observed. However, the coordinated cdt-ligand in 12 can easily be replaced by other olefins like ethylene, styrene or dvds giving the compounds [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13), [{(DDP)Ga}i(styrene) 2 ] (14) and [{(DDP)Ga}i(dvds)] (15), respectively. 160

177 IV. Conclusion and outlook. As a result of this study, three general structural motives of possible (i.e. isolable) products of these substitution reactions can be identified, which are depicted in Scheme 51. It seems, that the nature of the metal center (i.e. its size, reflected by its covalent radius) and the co-ligands present in the reaction mixture are relevant factors for the observed structures of the products in such reactions. L L M f Ga Ga g M M = small metal, L = non-chelating olefin + ML 3 Ga(DDP) + MLL 2 M = d 10 metal L = L 2 or olefin L 2 = chelating olefin + i(cdt) + M(cod) 2 i Ga 12 Lewis-acidic metal Ga M h Ga M = big metal Scheme 51: structural motives of the products of substitution reactions of olefins at d 10 metal centers ( = 2,6- i Pr 2 -C 6 H 3 ). Thus, due to its tremendous steric demand, a coordination of more than one Ga(DDP) ligand to the rather small i 0 center (covalent radius: 115 pm) similar to structure h in Scheme 51 is not possible, leading to decomposition of the formed complex, as already discussed for [{(DDP)Ga} 2 i(1,3-cod)] (Chapter III ). Instead, only one olefin is substituted and stable complexes of the type [{(DDP)Ga}iL 2 ] (L = non-chelating olefins, structure f, Scheme 51) are synthesizable. However, increasing the size of the metal center, i.e. when going from i 0 to Pt 0 (covalent radius: 129 pm), the complex [{(DDP)Ga} 2 Pt(1,3-cod)] [159, 160] is isolable. Obviously, the Pt-center is large enough to allow coordination of two Ga(DDP) ligands, giving a structure similar to h. 161

178 IV. Conclusion and outlook. Whereas the covalent radii of Pt and Pd are almost identical (129 pm vs. 128 pm), one might suggest, that the substitution of olefins in a Pd 0 complex like Pd 2 (dvds) 3 lead to the coordination of two Ga(DDP) ligands and thus, a structure similar to h for the reaction product. The synthesis of the complexes [{(DDP)Ga}i(dvds)] (15) and [{(DDP)E}Pd(dvds)] (E = Ga, Al(16)) however indicate, that also the co-ligands are crucial for the outcome of these substitution reactions. Accordingly, if a chelating olefin like dvds is present in the reaction mixture, the size of the metal center becomes neglible and only compounds of the type [{(DDP)Ga}ML] (L = chelating olefin, structure g) are obtained. The chelating effect of the co-ligand dvds might be responsible for the stability of those complexes. Additionally, the elongation of the C=C bond distances in all reported complexes indicate, that Ga(DDP) is a rather strong σ-donor ligand and thus comparable to electron rich phosphines or - heterocyclic carbenes, HCs. These results show, that, in contrast to the well examined ECp* ligands, the low-valent, bulky group 13 bis-imidinates E I (DDP) exhibit a great potential in terms of synthesis and stabilisation of unsaturated transition metal fragments. Further studies will now mainly focus on a detailed examination of the properties of these ligands to control the reactivity of small, electron rich and coordinatively unsaturated complexes of the type [M(E I ) n ] towards typical building blocks in organic chemistry (like olefins, H 2, C-X, Si-X and in general element-x bond activation reactions) and a comprehensive comparison with well established ancilliary ligands like phosphins or -heterocyclic carbenes HCs. In this respect, going from compounds containing d 10 metal centers like i(0), Pd(0) or Pt(0) to d 8 -complexes of e.g. h(i), Ir(I), u(0), Pd(II) or Pt(II) seems to be very promising. Thus, related compounds with phosphines or HCs have shown to be very reactive in bond activation reactions. As seen in previous studies, the six-membered group 13 bis-imidinates E(DDP) also exhibit a comparable reactivity towards such transition metal fragments. Based on M-spectroscopic studies, it is suggested, that e.g. the reaction of [(Ph 3 P) 3 ucl 3 ] with Ga(DDP) yields in a C-H bond activation and the formation of one or more orthometallated species, although these compounds could not be crystallographically investigated in more detail so far. The catalytic hydrogenation of cyclooctadiene at the Pt(II) complex [{(DDP)Ga} 2 Pt(H) 2 ] (17) (chapter III.3.3.) clearly points in this direction. In order to have a catalytic activation of a strong element-x bond, both thermodynamic, as well as kinetic premises must be fulfilled. For a thermodynamic preferred process, new strong 162

179 IV. Conclusion and outlook. bonds must be established, if strong bonds are catalytically broken during the activation reaction. From a kinetically point of view, the relevant intermediates must be reactive enough to activate such strong bonds. In this respect, the prevention of the possible alkyl- or aryl group transfer in such activation reactions from the transition metal in [(E I ) n M( )] to the group 13 metal center to give [(E I ) n-1 M(E )] is of general importance. Otherwise, the formation of the thermodynamically preferred E- bond is in conflict with a catalytic bond activation. One possibility to overcome this problem is the use of ligand systems, which are designed in such a way, that these alkyl transfer reactions are electronically unfavourable or sterically impossible. The specific features of the six-membered heterocycles E I (DDP) make them advantageous in comparison to their ECp* congeners. Besides a sterically blocking, these ligands also exhibit a sp 2 hybridized E I -system, in which no energetically favourable, free and electrophilic orbital is available at the group 13 metal center. Thus, together with the comparably less reductive properties, a catalytic course of the reaction may be possible. The nicely reversible oxidative addition of HSiEt 3 to the unsaturated fragment [{(DDP)Ga} 2 Pt], as well as the obviously catalytic H/D exchange reaction with the solvent C 6 D 6, which is suggested to proceed via a C-D activation of C 6 D 6 and the formation of a Pt(IV) intermediate (chapter III.3.3.), support, that these ligands possess a notably potential for bond activation reactions, which might stimulate further studies. 4. complexes with higher nuclearities featuring M-Ga(DDP) bonds. Multi-atom, naked metal clusters are increasingly the subject of investigation by complex physical and quantumchemical methods, for example, in nanotechnology because of their special properties in the transitional region between molecular and solid-state chemistry. [262] Also, metalloid cluster compounds can be seen as intermediates on the way to the corresponding element. [255] In this respect, the synthesis of mixed-metal homoleptic and heteroleptic, E I substituted complexes and clusters of the general type [M a E b (E I ) c ] m (with a,b,c,m = 0, 1, 2, ; M = different metals; E = different group 13 metals) is a synthetic challenge also in organometallic chemistry. The general possibility of the E I ligands ( = Cp*, alkyls ) to stabilize large clusters of this type was disclosed in 2004, when Schnöckel and co-workers reported on the synthesis of the large Al 50 cluster [Al 38 (AlCp*) 12 ], which is stabilized solely by AlCp* ligands. [263] The synthetic approach of Schnöckel et al. is the controlled disproportionation reaction of metastable Al(I) and Ga(I) halide solutions and the 163

180 IV. Conclusion and outlook. concomitant substitution of the halides by appropriate ligands to trap the resulting clusters. However, by doing so, this modus operandi is somehow limited to pure main group metal clusters. An alternative way comes from the idea, to use small transition metal complexes of the group 13 organyls ECp* as building blocks for the systematic synthesis of larger cluster compounds. This type of synthesis was recently embarked by our group and yielded in the controlled synthesis of homoleptic clusters of the type [M 2 (E I Cp*) 5 ]. [54] In contrast to that, the high steric demand of the DDP moiety prevents the formation of homoleptic M/Ga(DDP) complexes or clusters, thus leading to heteroleptic dinuclear complexes of the type [M 2 {Ga(DDP)} 2 L n ], with one or more additional ligands coordinating the metal center (chapter III.4.). In these reactions, the choice of the transition metal olefin precursor as well as the stoichiometry of the reactants is most often crucial for the formation of the new compounds, which are summarized in Table 45. Table 45: dinuclear compounds with coordinated E(DDP). Compounds M / E Lit. [{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19) 2 / 1 This study [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 -C 2 H 4 )] (20) 2 / 1 This study [(cod)i 2 {μ 2 -Ga(DDP)}{μ 2 -(PhC CPh)}(PhC CPh)] (21) 2 / 1 This study [{(DDP)Ga}Pd(CO)] 2 1 / 1 [159] and [160] [Pd{μ 2 -Ga(DDP)}( t BuC)] 2 (22) 2 / 1 This study [{Pd(dvds)} 2 {μ 2 -Al(DDP)}] (23) 1 / 1 This study [Pd 2 (GaCp*) 2 (μ 2 -GaCp*) 2 {μ 2 -Al(DDP)}] (24) 2 / 4 / 1 This study [{(DDP)Ga}Pt(CO)] 2 1 / 1 [159] and [160] [Pt{μ 2 -Ga(DDP)}( t BuC)] 2 (25) 1 / 1 This study [Pt{μ 2 -Ga(DDP)}(H) 2 ] 2 (26) 1 / 1 This study Based on the results of this study, four general structural motives can be identified for the dinuclear transition metal-e(ddp) complexes, which are shown in Scheme 52. Whereas in all dinuclear examples synthesized in this study the E I (DDP) ligand is found in a bridging position between the two transition metal centers, it is most likely, that the nature of the coligand affects the structure of the formed dinuclear complexes containing a E I (DDP) moiety, i.e. the arrangement of the additional ligands surrounding the cluster core. Thus, as represented by the compounds 22 and 25 and their related CO complexes [{(DDP)Ga}M(CO)] 2 (M = Pd, Pt), the use of strong π-acceptors like CO or isonitrils yields 164

181 IV. Conclusion and outlook. in the formation of a dinuclear compound with two E(DDP) ligands, which, due to their strong σ-donor capabilities, are bridging the two metal centers, whereas the π-acceptor ligands are terminally coordinated to the metals (structure i). L 1 L M Al M L L 1 M Ga M L 2 L 2 L L j L = strong σ-donors E I (DDP) L 1 = strong π-acceptors, 2e or 4e π-donors L 2 = L 1 or chelating olefins k Ga L M M L Ga L = strong π-acceptors, weak σ-donors i Al M l M chelating olefins Scheme 52: structural motives for dimeric complexes featuring M-E I (DDP) bonds. In contrast to that, the use of strong (however sterically less demanding) σ-donors (e.g. GaCp*) gives a M 2 {μ 2 -E(DDP)}-core, which is additionally coordinated by two terminal and two bridging donors, giving the structural analogue to [M 2 (ECp*) 5 ] (M = Pd, Pt) (structure j, complex 24). As seen in the formation of the complexes 19, 20 and 21, the use of π-donor ligands (like olefins or acetylenes) gives dinuclear complexes with one bridging and two (20) or three terminal (19) π-donors, which could also be chelating olefins (21) (structure k). Finally, if only chelating olefins are present in the reaction mixture, the formation of a complex with two chelated metals bridged by a E(DDP) moiety (structure l), comparable to 23, can be expected. 165

182 IV. Conclusion and outlook. Additionally, the difference between Al(I) and Ga(I) ligands, which was reported in both theoretical [49] as well as experimental manner for the E I Cp* compounds [45, 55], is also reflected in the bulky bis-imidinates E(DDP). Whereas the monomeric complex [(dvds)pd{ga(ddp)}] is the exclusive product in the reaction of [Pd 2 (dvds) 3 ] with Ga(DDP), the 1:1 reaction with Al(DDP) gives the dimeric complex [Pd 2 (dvds) 2 (μ 2 -Al(DDP)] (23), with the Al(DDP) again being located in a bridging position between the two Pd(0) centers. The monomeric compound [(dvds)pd{al(ddp)}] (16) can only be synthesized by using an excess (2.5 eq.) of Al(DDP) in this reaction. This observed preference of Al(DDP) to be coordinated in bridging mode is also observed for the AlCp* congener, as the previously described substitution of the bridging GaCp* units in [Pt 2 (GaCp*) 5 ] by AlCp* to give [Pt 2 (µ 2 [45, 55] -AlCp*) 3 (GaCp*) 2 ] reveals. According to this, comparably stronger σ-donor abilities towards transition metal centers can be suggested for Al(DDP) than for Ga(DDP). However, the general problem of a limitation in the cluster size by the steric bulk of the ligands, as already discussed in a comprehensive study for ECp* by Tobias Steinke in 2005, [264] is also obvious from the results of this study. Besides one exception, only compounds with two transition metals in the cluster core are obtained, in which the E I (DDP) ligand is always located in a bridging position rather than showing a terminal coordination mode. Further supported by the fact, that in all of these dimeric complexes only small coligands (linear carbonyls or isocyanides, H-atoms) are coordinated in terminal position, which is presumably because of steric bulk of the bridging E(DDP) ligands, a building block synthesis of larger clusters [M a {E(DDP)} b ] (with a > 3) seems to be hopeless. In consideration of this, a decrease in steric demande, e.g. by substituting the 2,6- diisopropylphenyl groups of the DDP-ligand with smaller groups like phenyl- (PPP) or t Bugroups (TTP), might help to overcome this problem. However, such compounds of the lowvalent group 13 metals Al(I) or Ga(I) are unknown so far. evertheless, taking into account, that the naked Ga + ion [Cp*Ga 2 ][BAr F ] or related [Ga][Otf] are nowadays quite well accessible and could serve as Ga + sources, [112, 126] the synthesis of the respective Ga(PPP) or Ga(TTP) compounds is only a question of time and will, most likely, open the door for a more facile synthesis of new and larger cluster compounds by this building block approach. 166

183 IV. Conclusion and outlook. Another route for the synthesis of larger cluster compounds, as already known and commonly applied for the tetrel atoms, is the reductive coupling reaction of a MX or MX 3 precursor (M = e.g. tetrel atom) with an adequate reducing agent (e.g. alkaline earth metals like KC 8 ), in which the naked metal atoms are inserted by adding a M(II)-halide, that is completely reduced and inserted into the cluster core. In the case of the group 14 elements, bulky ligands are used for the kinetic stabilizing of the cluster core. [255] Both required properties, the reducing power as well as the stabilizing capability, are combined in the group 13 heterocycles E(DDP), thus allowing the synthesis of the largest pure tin cluster [Sn 17 {ClGa(DDP)} 4 ] (28) reported so far. This result points to a general possibility to use the specific features of this ligands, most often described as being exotic, for the synthesis and stabilization of large cluster compounds. A more detailed analysis of the formation process itself, as well as optimization of the reaction conditions and the identification of suitable metal (also transition metal) precursors might be a suitable way to synthesize large clusters and thus get a deeper understanding of the development of the bonding situation on the way to the metallic state. [255] 167

184 V. Experimental section. V. Experimental section. 1. General remarks. All manipulations were carried out in an entirely air- and moisture free atmosphere of purified argon (cleaned over Cu-catalysts and dried with molecular sieve 4 Å, O 2 and H 2 O content < 1 ppm) using standard Schlenk and glove box techniques. All components were dried in high vacuum (p < mbar). The crystallization of the organometallic compounds was generally carried out by cooling a saturated solution of the compound or by slow diffusion of unpolar solvents into a saturated solution of the compounds in polar solvents (e.g. hexane into flourobenzene). Solvents. Hexane, pentane, THF, toluene and diethylether were dried using an mbraun Solvent Purification System. The final H 2 O content in all solvents was checked by Karl-Fischer- Titration and did not exceed 5 ppm. Flourobenzene and benzene were dried using an activated Al 2 O 3 column (heated prior to use). 2. Analytical methods. Elemental analysis. Elemental analysis were performed by the Microanalytical Laboratory of the uhr-universität Bochum (Supervisor: K. Bartholomäus), using a CHSO Vario EL 1998 analyzer fabricated by Elementar, Hanau. M-Spectroscopy. M spectra were recorded on a Bruker Avance DPX-250 spectrometer ( 1 H, MHz; 13 C, 62.9 MHz) in deuteratet solvents at 298 K, if not stated otherwise. The deuterated solvents (THF-D 8, C 6 D 6, toluene-d 8, CDCl 3 ) are purchased from Merck (Uvasol ) or Deutero GmbH, saturated with argon and dried over molecular sieve (4 Å) prior to use. Chemical shifts are given relative to TMS and were referenced to the solvent resonances as internal standards. 168

185 V. Experimental section. Single-crystal x-ray diffraction. The cell constants of suitable crystals of the compounds 1 to 28 were measured on a Saphire2- CCD-difractometer (Oxford Diffraction). The data collection had been carried out in a ω-scan mode. Data collection and reduction was performed using the program package CrysAlis Pro, with CrysAlis CCD used for data collection and CrysAlis ED for data reduction. Absorption correction of the data was carried out by indexing the planes of the crystals following the Gauss-method (compounds 1, 2, 4, 13, 14, 16, 17,22,23,25) or by using SADABS (compounds 3, 5-12, 15, 18-21, 24, 26-28). The space groups were determined using the program XPEP, followed by solving the structure with SHELXS. efinement of the structure, introduction of anisotropic displacement parameters and determination of the position of the Protons are conducted using the program SHELXL. The hydrogen atoms are fixed at geometric positions. The refinement was performed by minimizing the function Σw(F 2 obs. - F 2 calc. ) 2 with the full-matrix least-squares methods against F 2. The agreement between the crystallographic model and the diffraction data is discribed by the values, which are defined as followed: = 1 Σ( Fobs. - Fcalc. ) / Σ F obs. w / 2 2 = [ Σ w ( Fobs. Fcalc. ) / Σ w ( Fobs. ) ] with w = [σ 2 (F obs. 2 )+(ap) 2 +(bp)] -1 and P = 2 Max( F obs,0) + 2 F 3 2. calc. whereas a and b are to be refined. ), instability factor, Goodness-of-Fit: GOF = / 2 [ Σ w ( Fobs. Fcalc. ) /( obs. calc )] Due to disorder of solvent molecules and the resulting bad reflex/parameter ratio, the disordered solvent molecules were deleted and the data of the compounds 22 and 26 were corrected using the Squeeze program out of the PLATO program package. 169

186 V. Experimental section. 3. Computational Details. All calculations were performed with the Gaussian98 (rev. A11) program package. DFT calculations were carried out using the hybrid exchange-correlation functional B3LYP together with the Los Alamos ational Laboratory double-ζ LanL2DZ basis set. All structures were fully optimized without symmetry constraints. Vibrational frequencies were calculated for all stationary points to ensure that local minima were located. In order to verify that the overall picture is not affected by the choice of the comparably small LAL2DZ basis set, single point calculations were performed using a significantly larger basis (6-311G** for C, H, Ga and Stuttgart SC 1997 ECP for i). The results indicate that the qualitative trends in energies and activation barriers are reasonably well reproduced by the LAL2DZ basis set. A respective comparative table together with a depiction of the calculated structureas and their xyz-coordinates are given in the appendix (chapter VIII.1.). 4. Compounds. The synthesis of the starting compounds was performed using commercial available basis chemicals. Thus, ZnMe 2, ZnCl 2, [SnMe 2 Cl 2 ], SiCl 4, t BuC, t BuCl, ethylene, styrene, dvds, HSiEt 3, PhCCPh,and SnCl 2 are used as received without any further purification. The following compounds are synthesized according to literature methods: Ga(DDP), [31] Al(DDP), [30] [(COE) 2 hcl)] 2, [265] [{(DDP)Ga}Au{ClGa(DDP)}], [80] a[bar F ], [266] [(Ph 3 P)AuCl], [267] [i(cdt)], [226] [i(cod) 2 ], [268] [Pd 2 (dvds) 3 ], [162] [(1,3-cod)Pt{Ga(DDP)}], [159] [(dvds)pd{ga(ddp)], [159] GaCp*, [22] [Pt(cod) 2 ]. [269] 170

187 V. Experimental section. 5. Syntheses. h(coe)(c 6 H 6 )(µ-clga(ddp)) (1). 100 mg of [(COE) 2 hcl)] 2 (0.14 mmol) and 136 mg Ga(DDP) (0.28mmol) in 5 ml benzene were stirred at room temperature overnight. The solvent was removed in vacuo. ecrystallization of the deep green solid out of hexane at - 30 C gave deep green crystals of 1 in 60% yield (135 mg). 1 H-M (THF-d 8, T): δ (ar, 6H); 5.65 (s, 6H, C 6 H 6 ); 5.19 (s, 1H, CH); 3.80 (m, 2H, CHCH3); 3.28 (m, 2H, CHCH3); 2.67 (d, 2H, cyclooctene); 1.83 (d, 2H, cyclooctene); 1.76 (s, 6H, CCH 3 ); 1.43 (d, 6H, CHCH 3 ); 1.41 (d, 6H, CHCH 3 ); 1.22 (d + m, 12H, CHCH 3 + cyclooctene); 1.03 (d + m, 10H, CHCH 3 + cyclooctene); 13 C-M (THF-d 8, T): δ (C); (ar); (ar); (ar); (ar); (ar); (ar); (ar); 99.2 (γ- C); 98.7 (C 6 H 6 ); 57.7 (C=C); 57.5 (C=C); 35.7 (cyclooctene); 32.9 (cyclooctene); 30.4(CH(CH 3 ) 2 ); 28.3 (CH(CH 3 ) 2 ); 27.6 (cyclooctene); 27.1 (cyclooctene); 26.5 (cyclooctene); 25.2 (CH(CH 3 ) 2 ); 25.1 (CH(CH 3 ) 2 ); 24.6 (CH(CH 3 ) 2 ); 23.8 (CMe); 1 H-M (C 6 D 6, T): δ (ar, 12H); 5.12 (s, 1H, CH); 4.08 (m, 2H, CHCH 3 ); 3.36 (m, 2H, CHCH 3 ); 2.83 (d, 2H, cyclooctene); 1.89 (d, 2H, cyclooctene); 1.69 (d, 6H, CHCH 3 ); 1.68 (s, 6H, CCH 3 ); 1.44 (d, 6H, CHCH 3 ); 1.41 (m, 4H, cyclooctene); 1.25 (d, 6H, CHCH 3 ); 1.20 (m, 4H, cyclooctene); 1.09 (d, 6H, CHCH 3 ); 1.00 (m, 4H, cyclooctene); 13 C-M (C 6 D 6, T): δ (C); (ar); (ar); (ar); (ar); (ar); (ar); (ar); 98.9 (γ-c); 96.1 (t, coord. C 6 D 6 ); 55.8 (C=C); 55.6 (C=C); 34.0 (cyclooctene); 31.3 (cyclooctene); 28.5 (CH(CH 3 ) 2 ); 28.4 (CH(CH 3 ) 2 ); 26.7 (cyclooctene); 26.6 (cyclooctene); 25.4 (cyclooctene); 25.3 (cyclooctene); 25.1 (CH(CH 3 ) 2 ); 25.0 (CH(CH 3 ) 2 ); 23.8 (CH(CH 3 ) 2 ); 23.7 (CH(CH 3 ) 2 ); 23.6 (CH(CH 3 ) 2 ); 23.5 (CH(CH 3 ) 2 ); 23.1 (CH(CH 3 ) 2 ); 23.0 (CH(CH 3 ) 2 ); 22.3 (CMe); 22.1 (CMe). Elemental analysis calculated (found) for hga 2 C 43 H 61 Cl: C, (63.44); H, 7.55 (7.58);, 3.44 (3.46) [{(DDP)GaMe} 2 Zn] (2). Ga(DDP) (200 mg, 0.41 mmol) was dissolved in 5 ml of F-Ph at room temperature ml of a M solution of ZnMe 2 in F-Ph was slowly added. The reaction mixture turned to deep orange and was stirred for an additional 1h at room temperature. The solvent was removed in vacuo and the orange-yellow solid was redissolved in ca. 2 ml. Cooling the solution to -30 C overnight to give deep orange-red crystals of 2 in 178 mg yield (81 %). 171

188 V. Experimental section. 1 H-M (CDCl 3, 25 C): δ (m, 6H, Ar), 5.24 (s, 1H, γ-ch), 3.55 (sept, 2H, CH(Me) 2 ), 3.07 (sept, 2H, CH(Me) 2 ), 1.85 (s, 6H, CH 3 ), 1.28 (d, 6H, CH(Me) 2 ), 1.24 (d, 6H, CH(Me) 2 ), 1.22 (d, 6H, CH(Me) 2 ), 1.09 (d, 6H, CH(Me) 2 ), (s, 3H, GaMe). 13 C-M (CDCl 3, 25 C): δ (C), (CMe), (Ar), (Ar) (Ar), (Ar), (Ar), 97.1 (γ -C), 28.9 (CHMe 2 ), 27.7 (CHMe 2 ), 26.8 (CMe), 24.8 (CMe), 24.1 (CHMe 2 ) (CHMe 2 ), 23.7 (CHMe 2 ), 1.0 (GaMe); Elemental analysis calculated (found) for C 60 H 88 Ga 2 4 Zn: C, (67.69); H, 8.29 (8.57);, 5.24 (5.06). [{(DDP)GaCl}Zn(Cl)(THF) 2 ] (3). ZnCl 2 (55 mg, 0.41 mmol) and Ga(DDP) (200 mg, 0.41 mmol) were mixed and dissolved in 5 ml of THF at room temperature. The pale yellow solution was stirred for 1h, the solvent reduced to ca. 2ml and stored at -30 C overnight. Big colourless prismatic crystals of 3 could be obtained in 250 mg yield (79 %). 1 H-M (CDCl 3, 25 C): δ (m, 6H, Ar), 5.27 (s, 1H, γ-ch), 3.83 (br. t, 8H, THF), 3.24 (sept, 4H, CH(Me) 2 ), 1.90 (br., 14 H, overlapping signal of THF and CH 3 ), 1.33 (d, 12H, CH(Me) 2 ), 1.18 (d, 12H, CH(Me) 2 ). 13 C-M (CDCl 3, 25 C): δ (C), (CMe), (Ar), (Ar) (Ar), 97.1 (γ -C), 68.7 (THF), 28.5 (THF), 25.5 (CMe), 25.1 (CMe), 24.8 (CHMe 2 ), 23.9 (CHMe 2 ). Elemental analysis calculated (found) for C 37 H 57 Cl 2 Ga 2 O 2 Zn: C, (57.89); H, 7.48 (7.56);, 3.65 (3.98). [{(DDP)GaMe}GaMe 2 ] (4). Ga(DDP) (200 mg, 0.41 mmol) was dissolved in 5 ml of hexane at room temperature ml of a M solution of GaMe 3 in hexane was slowly added. The reaction mixture turned to deep yellow and was stirred for an additional 1h at room temperature. The solvent was reduced to ca. 2 ml and stored at -30 C overnight to give deep yellow crystals of 4 in 200 mg yield (81 %). 1 H-M (C 6 D 6, 25 C): δ (m, 6H), 4.78 (s, 1H, γ -CH), 3.52 (sept, 2H, CH(Me) 2 ), 3.04 (sept, 2H, CH(Me) 2 ), 1.52 (s, 6H, CH 3 ), 1.35 (d, 6H, CH(Me) 2 ), 1.13 (d, 12H, CH(Me) 2 ), 1.08 (d, 6H, CH(Me) 2 ), 0.37 (br, 6H, GaMe 2 ), (s, 3H, GaMe). 13 C-M (C 6 D 6, 25 C): δ (C), (CMe), (Ar), (Ar) (Ar), (Ar), (Ar), 96.2 (γ - C), 28.6 (CHMe 2 ), 25.7 (CHMe 2 ), 24.8 (CMe), 24.6 (CMe), 23.9 (CHMe 2 ) (CHMe 2 ), 1.4 (GaMe), -6.1 (GaMe 2 ). Elemental analysis calculated (found) for C 32 H 50 Ga 2 2 : C, (63.89); H, 8.37 (8.41);, 4.65 (5.08). 172

189 V. Experimental section. [{(DDP)GaMe} 2 GaMe] (5): 150 mg Ga(DDP) (0.031 mmol) were dissolved in 5 ml of hexane at room temperature. On slow addition of 0.49 ml of a M solution of GaMe 3 in hexane, the colour of the solution turns to intense yellow (after the addition of ½ eq. GaMe 3 ) and changes to orange after the addition was completed. Stirring the reaction mixture for 1h at room temperature and cooling it to -30 C overnight, a mixture of bright yellow and deep orange crystals were obtained. Separating the orange crystals of 5 manually gave a yield of 29 mg (25 %). 1 H-M (C 6 D 6, -40 C): δ (br, m, 12H), 4.64 (s, 2H, γ -CH), 3.35 (br, sept, 8H, CH(Me) 2 ), 1.41 (br, s, 12H, CH 3 ), 1.09 (br, d, 48H, CH(Me) 2 ), 0.63 (br, 6H, (DDP)GaMe), 0.03 (br, s, 3H, GaMe). 13 C-M (C 6 D 6, -40 C): δ (C), (CMe), (CMe), (CMe), (CMe), (Ar), (Ar), (Ar), (Ar), (Ar), 98.8 (γ -C), 37.6 (CHMe 2 ), 35.0 (CHMe 2 ), 31.3 (CHMe 2 ), 31.2 (CHMe 2 ), 30,7 (CHMe 2 ), 29.8 (CHMe 2 ), 28.2 (CHMe 2 ), 27.4 (CHMe 2 ), 27.1 (CHMe 2 ), 26.4 (CHMe 2 ), 26.2 (CHMe 2 ), 26.1 (CHMe 2 ), 17.4 (CMe), 14.6 (CMe), 4.1 (GaMe), 0.86 (GaMe). Elemental analysis calculated (found) for C 61 H 91 Ga 3 4 : C, (67.24); H, 8.42 (8.37);, 5.17 (5.41). [Me 2 Sn{ClGa(DDP)}] (6): 45 mg (0.205 mmol) [SnMe 2 Cl 2 ] and 200 mg Ga(DDP) (0.41 mmol) were dissolved in toluene at romm temperature. The colour of the former yellow solution turns slightly orange after several minutes. Concentrating and cooling the solution to -30 C overnight gave 6 as colourless crystals. Yield: 125 mg (86 %). 1 H-M (C 6 D 6, T): δ (6H), 4.96 (s, 1H, γ -CH), 3.84 (sept, 2H, CH(Me) 2 ), 3.10 (sept, 2H, CH(Me) 2 ), 1.48 (s, 6H, CH 3 ), 1.31 (d, 6H, CH(Me) 2 ), 1.24 (d, 6H, CH(Me) 2 ), 1.21 (d, 6H, CH(Me) 2 ), 0.93 (d, 6H, CH(Me) 2 ), (SnMe J = 21.7 Hz). 13 C-M (C 6 D 6, T): δ 169.0, 146.5, 142.3, 125.5, 123.4, 99.3, 29.5, 28.7, 28.1, 24.8, 24.2, 24.1, 24.0, Elemental analysis calcd. (found) for C 60 H 88 Cl 2 Ga 2 4 Sn: C, (60.35); H, 7.43 (7.80);, 4.69 (4.54). [Cl 3 Si{ClGa(DDP)}] (7): 200 mg Ga(DDP) (0.41 mmol) were dissolved in 5 ml hexane and an excess of SiCl 4 (1 ml) was added via a syringe to the solution. Heating the solution to reflux and stirring for 10 min at room temperature afforded a colourless precipitate. All volatiles were removed in vacuo, the colourless precipitate redissolved in 3 ml hexane, again heated to reflux and slowly cooled to room temperature to give colourless crystals. Yield: 130 mg (48 %). 173

190 V. Experimental section. 1 H-M (C 6 D 6, T): δ (6H), 4.88 (s, 1H, γ -CH), 3.74 (sept, 2H, CH(Me) 2 ), 3.18 (sept, 2H, CH(Me) 2 ), 1.54 (s, 6H, CH 3 ), 1.51 (d, 6H, CH(Me) 2 ), 1.38 (d, 6H, CH(Me) 2 ), 1.18 (d, 6H, CH(Me) 2 ), 0.99 (d, 6H, CH(Me) 2 ). 13 C-M (C 6 D 6, T): δ 170.9, 145.9, 142.6, 140.2, 126.7, 124.1, 123.6, 98.5, 29.9, 28.2, 26.6, 24.9, 24.5, 23.8, Elemental analysis calculated (found) for C 29 H 41 Cl 4 Ga 2 Si: C, (53.75); H, 6.29 (5.89);, 4.26 (4.00). [(DDP)Ga( t Bu)(C)] (8). Ga(DDP) (58 mg, 0.12 mmol) was dissolved in hexane (5 ml) at room temperature. Excess t BuC (10 mg, 0,12 mmol) was added via a syringe. The reaction mixture was heated to 60 C and stirred for 1h. All volatiles were removed in vacuo and the crude product recrystallized from hexane (-30 C, overnight). Yield 60 mg (89 %). 1 H-M (C 6 D 6, T): δ (6H), 4.88 (s, 1H, γ -CH), 3.66 (sept, 2H, CH(Me) 2 ), 3.21 (sept, 2H, CH(Me) 2 ), 1.56 (s, 6H, CH 3 ), 1.53 (d, 6H, CH(Me) 2 ), 1.26 (d, 6H, CH(Me) 2 ), 1.20 (d, 6H, CH(Me) 2 ), 0.98 (d, 6H, CH(Me) 2 ), 0.75 (s, 9H, t Bu). 13 C-M (C 6 D 6, T): δ 170.1, 145.6, 142.5, 142.0, 125.5, 123.7, 98.0, 30.7, 29.5, 28.3, 26.1, 24.8, 23.5, Elemental analysis calculated (found) for C 34 H 50 Ga 3 : C, (71.35); H, 8.83 (8.95);, 7.37 (7.02). [(DDP)Ga( t Bu)(Cl)] (9). Ga(DDP) (220 mg, 0.45 mmol) was dissolved in hexane (10 ml) at room temperature. Excess t BuCl (1 ml, 1,5 mmol) was added via a syringe. The reaction mixture was heated to 60 C and stirred for 20 min. All volatiles were removed in vacuo and the crude product recrystallized from hexane (-30 C, overnight). Yield 195 mg (89 %). 1 H-M (C 6 D 6, T): δ (6H), 4.97 (s, 1H, γ -CH), 3.92 (sept, 2H, CH(Me) 2 ), 3.30 (sept, 2H, CH(Me) 2 ), 1.61 (s, 6H, CH 3 ), 1.50 (d, 6H, CH(Me) 2 ), 1.31 (d, 6H, CH(Me) 2 ), 1.17 (d, 6H, CH(Me) 2 ), 1.00 (d, 6H, CH(Me) 2 ), 0.84 (s, 9H, t Bu) C-M (C 6 D 6, T): δ 169.7, 145.9, 142.6, 142.5, 125.6, 123.6, 98.3, 30.8, 29.6, 28.0, 25.7, 25.0, 24.9, 23.7, Elemental analysis calculated (found) for C 33 H 50 ClGa 2 : C, (67.86); H, 8.69 (8.83);, 4.83 (5.56) [{(DDP)Ga} 2 Au][BAr F ] (10): 90 mg of [{(DDP)Ga}Au{ClGa(DDP)}] (0,075 mmol) and 80 mg a[bar F ] (0.09 mmol) in 3 ml C 6 H 5 F were stirred at room temperature for 1h. The white precipitate of acl was filtered off and the solvent of the resulting yellow solution was removed in vacuo. The remaining colourless solid was washed twice with hexane and dried in vacuo. Yield 145 mg (95 %). 174

191 V. Experimental section. 1 H-M (C 6 H 5 F/C6D6, T): δ 8.36 (br, 8H, BArF), 7.65 (s, 4H, BArF), (ar, 12H,), 5.26 (s, 2H, CH), 2.50 (m, 8H, CHCH 3 ), 1.61 (s, 12H, CCH 3 ), 1.09 (d, 24H, CHCH 3 ), 0.85 (d, 24H, CHCH 3 ). 13 C-M: Due to the large 13 C M signals for the solvent, no satisfying 13 C- M spectrum of this sample could be recorded. EA analysis was not performed. [{(DDP)Ga. THF} 2 Au][BAr F ] (10. 2THF). The reaction was performed in a similar way described for 10 using 47 mg of [(Ph 3 P)AuCl)] (0.1 mmol), 100 mg Ga(DDP) (0.2 mmol), 90 mg a[bar F ] (0.1 mmol) and THF as a solvent. Yield 150 mg (73 %). Crystals, suitable for X-ay analysis were obtained by slow diffusion of hexane into a saturated THF-solution of 10. 2THF at room temperature. 1 H-M (THF-d 8, T): δ 7.79 (m, 8H, BArF), 7.57 (s, 4H, BArF), (ar, 12H), 5.36 (s, 2H, CH), 2.79 (m, 8H, CHCH 3 ), 1.79 (s, 12H, CCH 3 ), 1.11 (d, 24H, CHCH 3 ), 0.95 (d, 24H, CHCH 3 ). 13 C-M (THF-d 8, T): δ (C), (q, BArF-ar), (ar), (ar), (ar), (q, BArF-ar), (q, BArF-ar), (ar), (q, j C-F = 272 Hz, BArF-CF 3 ), (ar), (BArF-ar), 99.7 (γ-c), 29.2 (CHCH 3 ), 24.4 (CHCH 3 ), 24.1 (CMe). 1 H-M (C 6 H 5 F/C 6 D 6 : 10:1, T): δ 8.25 (m, 8H, BArF), 7.55 (s, 4H, BArF), (ar, overlaid by C 6 H 5 F), 5.17 (s, 2H, CH), 3.43 (m, 8H, THF), 2.43 (m, 8H, CHCH 3 ), 1.53 (s, 12H, CCH 3 ), 1.44 (m, 8H, THF), 1.02 (d, 24H, CHCH 3 ), 0.77 (d, 24H, CHCH 3 ). Elemental analysis calculated (found) for AuGa 2 4 BO 2 C 98 H 110 F 24 : C, (53.81); H, 5.09 (5.24);, 2.57 (2.85). [{(DDP)Ga. THF}Zn(µ-Cl)(THF)] 2 [BAr F ] 2 (11). A mixture of 100 mg of 3 (0.127 mmol) with 1 eq. a[bar F ] (100 mg, mmol) in THF or FPh was stirred at room temperature for 1h. Cooling the colourless solution to -30 C overnight, colourless needles of 9 could be obtained in 72 % yield (173 mg). 1 H-M (C 6 H 5 F/C 6 D 6 7:1, 25 C): δ 8.33 (br, 16H, BArF), 7.63 (s, 8H, BArF), (overlying signals of FPh and aromatic protons, not interpretable), 5.24 (s, 2H, γ-ch), 3.51 (br. s, 16H, THF), 2.78 (sept, 8H, CHMe 2 ), 1.60 (s, 12 H, CCH 3 ), 1.53 (br. m, 16H, THF), 1.16 (d, 24H, CHMe 2 ), 1.09 (d, 24H, CHMe 2 ). 13 C-M (C 6 H 5 F /C 6 D 6 7:1, 25 C): Because of the large 13C M signals for the solvent, no assignment of the 13 C M signals of the aromatic carbons of this sample could be done. δ 98.7 (γ-c), 68.5 (THF), 28.5 (CHMe 2 ), 25.5 (THF), 25.1 (CMe), 23.9 (CHMe 2 ), 23.5 (CHMe 2 ). Elemental analysis calculated (found) for C 138 H 138 B 2 Cl 2 F 48 Ga 2 4 O 4 Zn 2 : C, (52.08); H, 4.36 (4.35);, 1.76 (1.66). 175

192 V. Experimental section. [{(DDP)Ga}i(cdt)] (12). A mixture of [i(cdt)] (91 mg, 0.41 mmol) and Ga(DDP) (200 mg, 0.41 mmol) was dissolved in hexane (5 ml) and stirred for 1 h. Cooling the reaction mixture to -30 C overnight gave deep red crystals of 12. Yield 256 mg (88%). 1 H-M (C 6 D 6, 25 C): δ 7.15 (m, 6H), 5.17 (s, 1H, γ -CH), 4.96 (br, 4H, olefinic H), 4.30 (m, 4H, olefinic H), 3.17 (sept, 4H, CH(Me) 2 ), 2.15 (d, 4H, cdt-ch 2 ), 1.99 (s, 8H, cdt-ch 2 ), 1.71 (s, 6H, CH 3 ), 1.32 (d, 12H, CH(Me) 2 ), 1.13 (d, 12H, CH(Me) 2 ). 13 C-M (C 6 D 6, r.t.): δ (C), (CMe), (Ar), (Ar), (Ar), (Ar), 99.4 (C=C), 98.1 (γ- C), 40.2 (CH 2 CH 2 ), 32.7 (CH 2 CH 2 ), 28.9 (CHMe 2 ), 25.1 (CHMe 2 ), 24.1 (CMe), 23.8 (CHMe 2 ). Elemental analysis calculated (found) for C 41 H 59 Ga 2 i: C, (69.67); H, 8.40 (8.58);, 3.95 (4.15). [{(DDP)Ga}i(C 2 H 4 ) 2 ] (13). (Method A). A cooled solution (-30 C) of Ga(DDP) (200 mg, 0.41 mmol) in hexane was added to [i(ethylene) 3 ], synthesized in situ by bubbling ethylene through a solution of [i(cdt)] (90.6 mg, 0.41 mmol) in hexane at -10 C for ca. 30 min. The colour of the solution turned to deep red. Solid residue was removed by means of cannulation and deep red crystals of 13 were obtained by crystallisation overnight at -30 C. Yield: 88 mg (35.63 %). (Method B). Ga(DDP) (200 mg, 0.41 mmol) and [i(cdt)] (90.6 mg, 0.41 mmol) were dissolved in 5 ml hexane and stirred for 30 min. Ethylene was bubbled through this solution for 20 min and the deep red solution stored under ethylene-atmosphere (~1 bar) at - 30 C overnight. Deep red crystals of 13 could be obtained in 45% yield (110 mg). The same yield is obtained by applying ethylene pressure (2-3 bar) to a solution of 12 in hexane for approx. 20 min. 1 H-M (rt, C 6 D 6 ): δ 5.20 (s, 1H, γ-ch ), 3.11 (sept, 4H, CHMe 2 ), 2.29 (s, 8H, ethylene-h), 1.68 (s, 6H, CH 3 ), 1.21 (d, 12H, Me 2 CH), 1.09 (d, 12H, Me 2 CH). 13 C-M (rt, C 6 D 6 ): δ (C), (Ar), (Ar), (γ-c), 42.5 (C 2 H 4 ), 31.9 (CH(CH 3 ) 2 ), 29.0 (CH(CH 3 ) 2 ), 24.5 (CH(CH 3 ) 2 ), 24.0(CH(CH 3 ) 2 ), 23.9 (CMe). Elemental analysis calculated (found) for C 33 H 49 Ga 2 i: C, (64.75); H, 8.20 (8.06);, 4.65 (4.94). emark: The EA-values are consistant with the calculated values for [{(DDP)Ga}i(C 2 H 4 )] (C, 64.85; H, 7.90;, 4.88), supposedly formed by decomposition of 13 and liberation of one C 2 H 4 ligand prior to measurement. 176

193 V. Experimental section. [{(DDP)Ga}i(styrene) 2 ] (14). (Method A). A mixture of [i(cdt)] (68 mg, mmol) and Ga(DDP) (150 mg, mmol) was dissolved in hexane (5 ml) and stirred for 1 h. After addition of styrene (1 ml) the solution turned orange-yellow. On stirring the reaction mixture for 1h at room temperature, an orange precipitate was formed. emoval of the solvent by cannulation and recrystallisation from hexane (-30 C overnight) gave orange-yellow crystals. Yield: 166 mg (71%). (Method B). [i(cod) 2 ] (84 mg, mmol), Ga(DDP) (150 mg, mmol) and styrene (1 ml) were dissolved in 5 ml of hexane. Stirring the solution for 1 h and removal of the solvent in vacuo yield in a yellow solid. ecrystallisation out of hexane gave orange-yellow crystals. Yield: 110 mg (47%). 1 H-M (C 6 D 6, 25 C): δ 7.15 (m, 12H), 6.68 (m, 4H), 5.18 (s, 1H, γ -CH), 3.39 (dd, 2H, HC=CH 2, J = 9.3 Hz), 3.07 (sept, 4H, CH(Me) 2 ), 2.48 (dd, 2H, HC=CH 2, 2 J(gem) = 2.7 Hz, 3 J(E) = Hz), 2.25 (dd, 2H, HC=CH 2, 2 J(gem) = 2.7 Hz, 3 J(Z) = 9.25 Hz), 1.67 (s, 6H, CH 3 ), 1.30 (d, 6H, CH(Me) 2 ), 1.09 (d, 6H, CH(Me) 2 ), 1.07 (d, 6H, CH(Me) 2 ), 1.04 (d, 6H, CH(Me) 2 ). 13 C-M (C 6 D 6, r.t.): δ (C), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (γ-c), 67.4 (C=C), 41.2 (C=C), 29.2 (CHMe 2 ), 24.9 (CHMe 2 ), 24.3 (CMe), 23.9 (CHMe 2 ), 23.8 (CHMe 2 ). Elemental analysis calculated (found) for C 45 H 57 Ga 2 i: C, (70.70); H, 7.62 (7.94);, 3.71 (3.80). [{(DDP)Ga}i(dvds)] (15). [i(cdt)] (90.6 mg, 0.41 mmol) and Ga(DDP) (200 mg, 0.41 mmol) were dissolved in hexane (5 ml) and an excess of dvds was added (ca. 0.2 ml, ca mmol). After stirring the reaction mixture for 1 h, the solvent was removed in vacuo. ecrystallisation from hexane (-30 C overnight) gave yellow needles. Yield 196 mg (65%). 1 H-M (C 6 D 6, 25 C): δ 7.15 (m, 6H), 5.20 (s, 1H, γ-ch ), 3.06 (sept., 4H, CHMe 2 ), (complex spin system, 6H, olefinic H), 1.66 (s, 6H, CH 3 ), 1.21 (d, 12H, Me 2 CH), 1.06 (d, 12H, Me 2 CH), 0.50 (s, CH 3 Si, 6H), (s, 6H, CH 3 Si). 13 C-M (C 6 D 6, r.t.): δ (C), (Ar), (Ar), 99.8 (γ-c), 49.2 (CH 2 CH), 44.5 (CH 2 CH), 28.6 (CH(CH 3 ) 2 ), 24.1 (CH(CH 3 ) 2 ), 23.7 (CH(CH 3 ) 2 ), 2.1 (SiMe), -0.9 (SiMe). Elemental analysis calculated (found) for C 37 H 59 Ga 2 OSi 2 i: C, (60.12); H, 8.12 (8.13);, 3.82 (3.83). 177

194 V. Experimental section. [{(DDP)Al}Pd(dvds)] (16). [Pd 2 (dvds) 3 ] (43 mg, 0,045 mmol) and Al(DDP) (100 mg, 0,225 mmol) were dissolved in hexane (10ml) and stirred at room temperature overnight. Traces of black precipitate were removed by means of cannulation and the solvent removed in vacuo. The crude product was recrystalized from hexane (-30 C overnight) to give yellow crystals. Yield: 49 mg (74 %). 1 H-M (C 6 D 6, 25 C): δ 7.15 (6H), 5.21 (s, 1H, γ-ch), 3.02 (sept, 4H, CH(Me) 2 ), (m, 6H, dvds), 1.61 (s, 6H, CH 3 ), 1.27 (d, 12H, CH(Me) 2 ), 1.07 (d, 12H, CH(Me) 2 ), 0.52 (s, 6H, Si(Me) 2 ), (s, 6H, Si(Me) 2 ). 13 C-M (C 6 D 6, r.t.): δ ,, 143,21, , , 10.56, 53.61, 52.94, 31.95, 24.28, 23.98, 23.04, 2.26, Elemental analysis calculated (found) for PdAl 2 Si 2 OC 37 H 59 : C, (59.86); H, 8.06 (8.14);, 3.80 (3.98). [Pt{Ga(DDP)} 2 (H) 2 ] (17). [(1,3-cod)Pt{Ga(DDP)} 2 ] (155 mg, mmol) was dissolved in hexane (5 ml). The solvent was cooled to -78 C, evacuated, warmed to room temperature and filled with hydrogen (2 bar). The reaction mixture was stirred for 3 h until a yellow precipitate was formed. The solvent was removed by means of cannulation, washed with cold hexane and dried in an argon flow. ecrystallisation from hexane gave yellow, rhomboedric crystalls. Yield: 75 mg (52 %). Analoguesly, the reaction of [Pt(cod) 2 ] (100 mg, 0.25 mmol) with two eq. Ga(DDP) (250 mg, 0.51 mmol) and subsequent treating with hydrogen (2 bar) also gives 17 in similar yield and purity. 1 H-M (C 6 D 6, 25 C): δ 7.05 (m, 6H), 5.07 (s, 2H, γ-ch), 3.07 (sept, 8H, CH(Me) 2 ), 1.62 (s, 12H, CH 3 ), 1.17 (d, 24H, CH(Me) 2 ), 1.13 (d, 24H, CH(Me) 2 ), (t, 2H, J Pt-H : 334 Hz). 13 C- M (C 6 D 6, r.t.): δ 166.7, 143.3, 141.1, 126.8, 123.9, 99.2, 28.9, 26.1, 24.3, Elemental analysis calculated (found) for C 58 H 84 Ga 2 4 Pt: C, (58.64); H, 7.23 (7.58);, 4.78 (5.00). [Pt{Ga(DDP)} 2 (H)(SiEt 3 )] (18). [(1,3-cod)Pt{Ga(DDP)} 2 ] (250 mg, mmol) and HSiEt 3 (23 mg, mmol) in hexane (5 ml) were stirred at room temperature for 3 h. The solvent was removed in vacuo and the resulting orange solid redissolved in hexane (2 ml). Cooling the solution to -30 C overnight gave in orange crystals. Yield: 120 mg (49 %). 1 H-M (C 6 D 6, T): δ 7.15 (broad signal, arom, 12H), 4.97 (s, 2H, CH), 3.30 (br. sept., 4H, i Pr-CH), 3.12 (br. Septet, 4H, i Pr-CH), 2.63 (br. Septet, 4H, i Pr-CH), 1.56 (br. s, 6H, CMe), 1.53 (br. s, 6H, CMe), 1.49 (br. d, 12H, i Pr-Me), 1.46 (br. d, 12H, i Pr-Me), 1.33 (br. d, 12H, i Pr-Me), 1.19 (q, 6H, CH 2 CH 3 ), 1.11 (br. d, 12H, i Pr-Me), 0.93 (t, 9H, CH 2 CH 3 ), (s, 1H, 178

195 V. Experimental section. Pt-H, J Pt-H : 530 Hz). 1 H-M (C 6 D 6, 70 C): δ 7.15 (broad signal, arom, 12H), 5.04 (s, 2H, CH), 3.16 (br. sept., 8H, i Pr-CH), 1.55 (br. s, 12H, CMe), 1.24 (q, 6H, CH 2 CH 3 ), 1.15 (br., 48H, i Pr-Me), 0.86 (t, 9H, CH 2 CH 3 ). 13 C-M (C 6 D 6, T): δ 170.1, 166.9, 166.8, 144.6, 144.0, 143.6, 142.5, 129.0, 128.6, 127.8, 124.5, 100.3, 31.9, 29.4 to 28.2 (broad and overlapping signals), 25.0 to 24.3 (broad and overlapping signals), 23.0, 15.9, 14.3, Elemental analysis calculated (found) for PtGa 2 4 C 64 H 98 : C, (59.83); H, 7.68 (8.03);, 4.36 (4.82). Synthesis of [{(C 2 H 4 ) 2 i} 2 (μ 2 -Ga(DDP)] (19) and [{(C 2 H 4 )i} 2 (μ 2 -Ga(DDP))(μ 2 - C 2 H 4 )] (20): [i(ethylene) 3 ] was prepared in situ by dissolving 91 mg (0.41 mmol) [i(cdt)] in 5ml of cold hexane (-10 C) and applying 3 bar of ethylene for 1h. Meanwhile, the colour of the solution turned from deep red to pale yellow. To this solution, a cooled hexane-solution of Ga(DDP) (100 mg, mmol) was added and the deep red solution stirred at -10 C for 1h. Slowly warming to room temperature, a mixture of deep red single crystals of 5 and 6 can be obtained. Unfortunately, neither 19 nor 20 could be synthesized individually. [{{μ 2 -Ga(DDP)}i 2 (μ 2 -PhCCPh)} 2 (PhCCPh)(1,5-cod)] (21). [i(cod) 2 ] (225 mg, 0.82 mmol), Ga(DDP) (200 mg, 0.41 mmol) and PhCCPh (150 mg, 0.84 mmol) were dissolved in hexane and heated to reflux for several minutes. Additional stirring the solution for 1h at room temperature and cooling to -30 C overnight yields deep red, octahedral crystals of 21 in 90% yield (395 mg). 1 H-M (C 6 D 6, r.t.): δ 7.75 (d, 2H, ar), 7.62 (d, 2H, ar), 7.56 (d, 4H, ar), 7.50 (d, 2H, ar), (m, 16H, ar), 5.57 (br. s, 2H, C=C), 5.49 (br. s, 2H, C=C), 5.26 (s, 1H, γ-ch), 3.81 (sep., 2H, ipr-ch), 3.18 (sep., 2H, ipr-ch), 1.85 (m, 8H, CH 2 ), 1.59 (s, 6H, CH 3 ), 1.50 (d, 6H, ipr-ch 3 ), 1.24 (d, 6H, ipr-ch 3 ), 0.82 (d, 6H, ipr-ch 3 ), 0.52 (d, 6H, ipr-ch 3 ). 13 C- M (C 6 D 6, r.t.): δ (C), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (γ-c), 96.6 (acetylene), 91.4 (acetylene), 33.3 (cod), 31.9 (cod), 30.4 (CH 3 ), 30.1 (CHCH 3 ), 28.7 (cod), 28.6 (cod) 28.4 (CHCH 3 ), 26.0 (CHCH 3 ), 25.7 (CHCH 3 ), 25.1 (CHCH 3 ), 25.0 (CHCH 3 ), 23.7 (CHCH 3 ). Elemental analysis calculated (found) for C 65 H 73 Ga 2 i 2 : C, (72.03); H, 6.88 (6.92);, 2.62 (2.75). 179

196 V. Experimental section. [Pd{μ 2 -Ga(DDP)}( t BuC)] 2 (22). [{(DDP)Ga}Pd(dvds)] (100 mg, 0.13 mmol) was dissolved in THF (5 ml) and carefully layered with hexane (2 ml). t BuC (11 mg, 0.13 mmol) was added to the upper layer giving dark red/violet crystals at the interface after 15h at room temperature. Yield: 85 mg (96%). Elemental analysis calculated (found) for Pd 2 Ga 2 4 C 68 H 100 : C, (59.97); H, 7.45 (7.82);, 6.21 (6.50). [{Pd(dvds)} 2 { μ 2 -Al(DDP)}] (23): [Pd 2 (dvds) 3 ] (107 mg, 0.14 mmol) and Al(DDP) (63 mg, 0.14 mmol) were dissolved in 10 ml hexane. The intense orange solution was stirred for 30 minutes, concentrated and stored at -30 C overnight to give yellow-orange crystals Yield: 85 mg (73%). 1 H-M (298 K, MHz, C6D6): δ (m, 6H, Ar), 5.01 (s, 1H, γ-ch), 3.25 (bs, 6H, C2H3), 3.07 (sept, 4H, ipr), 2.96 (sept, 4H, ipr), 2.75 (bs, 6H, C2H3), 1.45 (s, 6H, CMe), 1.26 (d, 6H, ipr-me), 1.12 (d, 6H, ipr-me), 1.06 (d, 6H, ipr-me), 0.9 (d, 6H, ipr-me), 0.29 (bs, 24H, SiMe). 13 C-M (C 6 D 6, r.t.): δ (C), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (γ-c), 31.9 (C2H3), 31.2 (CMe), 29.1 (C2H3), 27.2 (CHMe2), 27.0 (CHMe2), 26.7 (CHMe2), 26.6 (CHMe2), 3.2 (br, SiMe). Elemental analysis calculated (found) for C 45 H 77 Al 2 O 2 Pd 2 Si 4, C, (52.48); H, 7.53 (7.60);, 2.72(2.63). [Pd 2 (GaCp*) 2 (μ 2 -GaCp*) 2 {μ 2 -Al(DDP)}] (24): 50 mg (0.433 mmol) of 23 is dissolved in 5 ml hexane. After addition of 4 eq GaCp* (45 mg, 1.73 mmol), the former yellow solution turns into deep red. Stirring the solution for 10 minutes, concentrating the solution and storing it at -30 C overnight afforded deep red crystals of 24. Yield: 58 mg (90%). 1 H-M (298 K, MHz, Tol-d 8 ): δ (m, 6H, Ar), 4.93 (s, 1H, γ-ch), 3.39 (sept, 8H, ipr), 1.97 (s, 30H, GaCp*), 1.96 (s, 30H, GaCp*), 1.37 (d, 6H, ipr-me), 1.35 (s, 6H, CMe), 1.14 (d, 6H, ipr-me); δc(298 K, 62.9 MHz, C7D8) (C), (Ar), (Ar), (Ar), (Ar), (γ-c), (br, ring atoms GaCp*), (CMe), 28.9 (CMe), 25.2 (CHMe2), 24.8 (CHMe2), 24.5 (CHMe2), 24.2 (CHMe2), 11.5 (Cp*Me), 10.8 (Cp*Me). Elemental analysis calculated (found) for C 69 H 101 AlGa 4 2 Pd 2 : C, (55.87); H, 6.89 (6.65);, 1.90 (1.75). 180

197 V. Experimental section. [Pt{μ 2 -Ga(DDP)}( t BuC)] 2 (25). [(1,3-cod)Pt{Ga(DDP)} 2 ] (130 mg, mmol) was dissolved in THF (5 ml) and carefully layered with hexane (2 ml). t BuC (9 mg, mmol) was added to the upper layer giving dark red/violet crystals at the interface after 15h at room temperature. Yield: 150 mg (96%). Elemental analysis calculated (found) for Pt 2 Ga 2 4 C 68 H 100 : C, (53.12); H, 6.58 (6.74);, 5.49 (6.00). [Pt{μ 2 -Ga(DDP)}(H) 2 ] 2 (26). [Pt(cod) 2 ] (100 mg, 0.24 mmol) and Ga(DDP) (118 mg, 0.24 mmol) were dissolved in THF (5 ml). The solution was cooled to -78 C, evacuated, warmed to room temperature and filled with hydrogen (2 bar). After stirring the reaction mixture for 6h, a yellow precipitate was formed. The solvent was removed by means of cannulation and the crude product recrystalized from THF and dried by argon flow. Yield: 62 mg (38%). 1 H-M (C 6 D 6, 25 C): δ (m, 6H), 4.89 (s, 1H, γ -CH), 3.35 (sept, 4H, CH(Me) 2 ), 1.47 (s, 6H, CH 3 ), 1.11 (d, 12H, CH(Me) 2 ), 0.80 (d, 12H, CH(Me) 2 ), (br, 2H, Pt-H). 13 C- M (C 6 D 6, r.t.): δ 168.7, 144.2, 140.9, 127.1, 124.3, 97.7, 28.6, 25.8, 25.4, Elemental analysis calculated (found) for C 58 H 86 Ga 2 4 Pt 2 : C, (50.40); H, 6.33 (6.17);, 4.09 (4.05). [{{μ 2 -Ga(DDP)}i(C 2 H 4 )} 2 i(μ 2 -CH=CH 2 )(H)] (27). (Method A). [i(ethylene) 3 ] was prepared in situ by dissolving 68 mg (0.307 mmol) of [i(cdt)] in 5ml of cold hexane (-10 C) and applying 3 bar of ethylene for 1h. Meanwhile, the colour of the solution turned from deep red to pale yellow. To this solution, a cooled hexane-solution of Ga(DDP) (100 mg, mmol) was added and the solution turned into deep red. Heating the solution for several minutes to ~60 C yields in the formation of a deep red precipitate after cooling to room temperature. ecrystallisation by repeated heating of the solution to ca. 60 C and slowly cooling to room temperature gave deep red crystals of a in 83% yield (105 mg). (Method B). The addition of 0.5 eq. of [i(cdt)] (4 mg,0.018 mmol) to 19 (22 mg, mmol) in C 6 D 6 and heating the solution to ~60 C gave 27 in high yields, as proven by M spectroscopy. 1 H-M (C 6 D 6, 25 C): δ 9.70 (dd, 1H, ethylene-ch, J H,H1 = 10,3 Hz, J H,H2 =16,6 Hz), 7.01 (m, 4H, ar), 6.91 (t, 4H, ar), 6.85 (d, 2H, ar), 6.72 (d, 2H, ar), 4.98 (s, 2H, γ-ch), 4.49 (dd, 1H, ethylene-ch 2, J H,H = 10.3 Hz), 3.82 (d, 1H, ethylene-ch 2, J H,H = Hz), 3.26 (m, 8H, ethylene-ch 2, J H,H = Hz), 3.15 (m, 4H, ipr-ch), 1.50 (s, 12H, CH 3 ), 1.19 (d, 6H, ipr- CH 3 ), 1.13 (d, 6H, ipr-ch 3 ), 1.03 (d, 6H, ipr-ch 3 ), 1.01 (d, 6H, ipr-ch 3 ), 1.00 (d, 6H, ipr- CH 3 ), 0.90 (d, 6H, ipr-ch 3 ), 0.61 (d, 6H, ipr-ch 3 ), 0.07 (d, 6H, ipr-ch 3 ), (s, 1H, i- 181

198 V. Experimental section. H). 13 C-M (C 6 D 6, r.t.): δ (C), (C), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (Ar), (γ-c), 44.9 (C=C), 44.1 (C=C), 29.0, 28.5, 28.4, 28.2, 26.2, 26.1, 26.0, 24.5, 24.4, 24.2, 24.1, Elemental analysis calculated (found) for C 63 H 92 Ga 2 4 i 3 : C, (61.66); H, 7.56 (7.49);, 4.59 (5.02). [Sn 17 {ClGa(DDP)} 4 ] (28). 2 eq. of Ga(DDP) (0.2 g, 0.41 mmol) were dissolved in 5 ml THF and cooled to -40 C. After addition of 38 mg SnCl 2 (0.205 mmol) and stirring at -40 C for 1h, the colour of the solution turns into deep red / violet. emoving the solvent at -30 C in vacuo, washing the dark solid with hexane and recrystallisation of the crude product out of THF/hexane (1:2) afforded deep red crystals. Unfortunately, no M data are available so far, because the compound could yet not be reproduced. The results of the single crystal x-ray diffraction analysis of this complex is discussed in chapter VI

199 VI. Literature index. VI. Literature index. [1] H. D. Kaesz,. S. Williams,. F. Hicks, J. I. Zink, Y.-J. Chen, H.-J. Müller, Z. Xue, D. Xu, D. K. Shuh, Y. K. Kim, ew J. Chem. 1990, 14, 527. [2] F. B. Passos, D. A. G. Aranda, M. Schmal, J. Catal. 1998, 178, 478. [3] E. S. Shpiro, D. P. Shevchenko, O. P. Tkachenko,. V. Dmitriev, Appl. Catal., A 1994, 107, 147. [4] K. J. Chao, A. C. Wei, H. C. Wu, J. F. Lee, Micropor. Mesopor. Mater. 2000, 35-36, 413. [5] A. Bayer, K. Flechtner,. Denecke, H. P. Steinruck, K. M. eyman,. osch, Surf. Sci. 2006, 600, 78. [6]. Iwasa,. Takezawa, Top.Catal. 2003, 22, 215. [7]. Iwasa, M. Takizawa, M. Arai, Appl. Catal., A 2005, 283, 255. [8]. Iwasa, M. Yoshikawa, M. Arai, Phys. Chem. Chem. Phys. 2002, 4, [9]. ilius, T. M. Wallis, W. Ho, J. Chem. Phys. 2002, 117, [10] S. Penner, B. Jenewein, H. Gabasch, B. Klotzer, D. Wang, A. Knop-Gericke,. Schlogl, K. Hayek, J. Catal. 2006, 241, 14. [11] S. Pick, Surf. Sci. 1999, 436, 220. [12] L. T. Sein, S. A. Jansen, J. Catal. 2000, 196, 207. [13] V. I. Yakerson, Z. L. Dykh, A.. Subbotin, B. S. Gudkov, S. V. Chertkova, A.. adin, E. A. Boevskaya, Z. A. Tertichnik, E. Z. Golosman,. G. Sarmurzina, Kinet. Catal. 1998, 39, 101. [14] S. Ito, Y. Suwa, S. Kondo, S. Kameoka, K. Tomishige, K. Kunimori, Catal. Commun. 2003, 4, 499. [15] H. Braunschweig, M. Colling, Eur. J. Inorg. Chem. 2003, 393. [16] H. Braunschweig, M. Colling, Coord. Chem. ev. 2001, 223, 1. [17] C. Dohmeier, C. obl, M. Tacke, H. Schnöckel, Angew. Chem. Int. Ed. 1991, 30, 564. [18] S. Schulz, H. W. oesky, H. J. Koch, G. M. Sheldrick, D. Stalke, A. Kuhn, Angew. Chem. Int. Ed. 1993, 32, [19] M. Schormann, K. S. Klimek, H. Hatop, S. P. Varkey, H. W. oesky, C. Lehmann, C. öpken,. Herbst-Irmer, M. oltemeyer, J. Solid State Chem. 2001, 162, 225. [20] D. Loos, H. Schnöckel, J. Organomet. Chem. 1993, 463, 37. [21] D. Loos, E. Baum, A. Ecker, H. Schnöckel, Angew. Chem. Int. Ed. 1997, 36, 860. [22] P. Jutzi, B. eumann, G. eumann, H.-G. Stammler, Organometallics 1998, 17, [23] P. Jutzi, L. O. Schebaum, J. Organomet. Chem. 2002, 654,

200 VI. Literature index. [24] W. Uhl, A. Jantschak, J. Organomet. Chem. 1998, 555, 263. [25] W. Uhl, W. Hiller, M. Layh, W. Schwarz, Angew. Chem. 1992, 104, [26] C. Schnitter, H. W. oesky, C. öpken,. Herbst-Irmer, H.-G. Schmidt, M. oltemeyer, Angew. Chem. Int. Ed. 1998, 32, [27] W. Uhl,. Graupner, M. Layh, U. Schütz, J. Organomet. Chem. 1995, 493, C1. [28] M. C. Kuchta, J. B. Bonanno, G. Parkin, J. Am. Chem. Soc. 1996, 118, [29] M. C. Kuchta, H. V.. Dias, S. G. Bott, G. Parkin, Inorg. Chem. 1996, 35, 943. [30] C. Cui, H. W. oesky, H.-G. Schmidt, M. oltemeyer, H. Hao, F. Cimpoesu, Angew. Chem. Int. Ed. 2000, 39, [31]. J. Hardman, B. E. Eichler, P. P. Power, Chem. Commun. 2000, [32] M. S. Hill, P. B. Hitchcock, Chem. Commun. 2004, [33] M. S. Hill,. Pongtavornpinyo, P. B. Hitchcock, Chem. Commun. 2006, [34] S. T. Haubrich, P. P. Power, J. Am. Chem. Soc. 1998, 120, [35] J. Su, X.-W. Li,. C. Crittendon, C. F. Campana, G. H. obinson, Organometallics 1997, 16, [36] X.-W. Li, W. T. Pennington, G. H. obinson, J. Am. Chem. Soc. 1995, 117, [37] B. Twamley, P. P. Power, Angew. Chem. 2000, 112, [38] C. Jones, P. C. Junk, J. A. Platts, A. Stasch, J. Am. Chem. Soc. 2006, 2006, [39]. J. Baker,. D. Farley, C. Jones, M. Kloth, D. M. Murphy, J. Chem. Soc., Dalton Trans. 2002, [40] E. S. Schmidt, A. Jokisch, H. Schmidbaur, J. Am. Chem. Soc. 1999, 121, [41] C. Dohmeier, D. Loos, H. Schnöckel, Angew. Chem. Int. Ed. 1996, 35, 129. [42]. Murugavel, V. Chandrasekhar, Angew. Chem. Int. Ed. 1999, 38, [43]. A. Fischer, J. Weiss, Angew. Chem. Int. Ed. 1999, 38, [44]. J. Baker, C. Jones, Coord. Chem. ev. 2005, 249, [45] C. Gemel, T. Steinke, M. Cokoja, A. Kempter,. A. Fischer, Eur. J. Inorg. Chem. 2004, [46] H. W. oesky, Inorg. Chem. 2004, 43, [47]. A. Fischer, D. Weiss, M. Winter, I. Mueller, H. D. Kaesz,. Froehlich, G. Frenking, J. Organomet. Chem. 2004, 689, [48] A. Y. Timoshkin, G. Frenking, J. Am. Chem. Soc. 2002, 124, [49] J. Uddin, G. Frenking, J. Am. Chem. Soc. 2001, 123, [50] G. Frenking, J. Organomet. Chem. 2001, 635, 9. [51] C. Boehme, G. Frenking, Chem. Eur. J. 1999, 5,

201 VI. Literature index. [52] M. S. Hill, P. B. Hitchcock,. Pongtavornpinyo, Dalton Trans. 2005, 273. [53] M. eiher, A. Sundermann, Eur. J. Inorg. Chem. 2002, [54] C. Gemel, T. Steinke, D. Weiss, M. Cokoja, M. Winter,. A. Fischer, Organometallics 2003, 22, [55] T. Steinke, C. Gemel, M. Winter,. A. Fischer, Chem. Eur. J. 2005, 11, [56] T. Steinke, C. Gemel, M. Cokoja, M. Winter,. A. Fischer, Angew. Chem. Int. Ed. 2004, 43, [57] T. Cadenbach, C. Gemel,. Schmid,. A. Fischer, J. Am. Chem. Soc. 2005, 127, [58] T. Steinke, C. Gemel, M. Cokoja, M. Winter,. A. Fischer, Dalton Trans. 2005, 55. [59] B. Buchin, C. Gemel, A. Kempter, T. Cadenbach,. A. Fischer, Inorg. Chim. Acta 2006, 359, [60] S. P. Green, C. Jones, A. Stasch, Inorg. Chem. 2007, 46, 11. [61] C. Jones,. P. ose, A. Stasch, Dalton Trans. 2007, [62] S. P. Green, C. Jones, D. P. Mills, A. Stasch, Organometallics 2007, 26, [63] P. L. Arnold, S. T. Liddle, J. McMaster, C. Jones, D. P. Mills, J. Am. Chem. Soc. 2007, 129, [64] S. Aldridge,. J. Baker,. D. Coombs, C. Jones,. P. ose, A. ossin, D. J. Willock, Dalton Trans. 2006, [65]. J. Baker, C. Jones, D. M. Murphy, Chem. Commun. 2005, [66] C. Jones, D. P. Mills, J. A. Platts,. P. ose, Inorg. Chem. 2006, 45, [67] S. P. Green, C. Jones, K.-A. Lippert, D. P. Mills, A. Stasch, Inorg. Chem. 2006, 45, [68]. J. Baker, C. Jones, D. P. Mills, D. M. Murphy, E. Hey-Hawkins,. Wolf, Dalton Trans. 2005, 64. [69]. J. Baker, C. Jones, M. Kloth, Dalton Trans. 2005, [70] P. M. Gurubasavaraj, S. K. Mandal, H. W. oesky,. B. Oswald, A. Pal, M. oltemeyer, Inorg. Chem. 2007, 46, [71] H. Zhu,. B. Oswald, H. Fan, H. W. oesky, Q. Ma, Z. Yang, H.-G. Schmidt, M. oltemeyer, K. Starke,. S. Hosmane, J. Am. Chem. Soc. 2006, 128, [72] Z. Yang, X. Ma,. B. Oswald, H. W. oesky, C. Cui, H.-G. Schmidt, M. oltemeyer, Angew. Chem. Int. Ed. 2006, 45, [73] S. embenna, H. W. oesky, S. K. Mandal,. B. Oswald, A. Pal,. Herbst-Irmer, M. oltemeyer, H.-G. Schmidt, J. Am. Chem. Soc. 2006, 128,

202 VI. Literature index. [74] H. Zhu, J. Chai, H. Fan, H. W. oesky, C. He, V. Jancik, H.-G. Schmidt, M. oltemeyer, W. A. Merrill, P. P. Power, Angew. Chem. Int. Ed. 2005, 44, [75] V. Jancik, H. W. oesky, Inorg. Chem. 2005, 44, [76] V. Jancik, L. W. Pineda, A. C. Stueckl, H. W. oesky,. Herbst-Irmer, Organometallics 2005, 24, [77] V. Jancik, M. M. M. Cabrera, H. W. oesky,. Herbst-Irmer, D. eculai, A. M. eculai, M. oltemeyer, H.-G. Schmidt, Eur. J. Inorg. Chem. 2004, [78] G. Bai, Y. Peng, H. W. oesky, J. Li, H.-G. Schmidt, M. oltemeyer, Angew. Chem. Int. Ed. 2003, 42, [79] Stephan Schulz, Andreas Kuczkowski, Daniella Schuchmann, Ulrich Flörke, M. ieger, Organometallics 2006, 25, [80] A. Kempter, C. Gemel,. A. Fischer, Inorg. Chem. 2005, 44, 163. [81] A. Kempter, C. Gemel,. J. Hardman,. A. Fischer, Inorg. Chem. 2006, 45, [82] J. Vollet, J.. Hartig, H. Schnöckel, Angew. Chem. Int. Ed. 2004, 43, [83] W. Uhl, M. Benter, S. Melle, W. Saak, G. Frenking, J. Uddin, Organometallics 1999, 18, [84]. Murugavel, V. Chandrasekhar, Angew. Chem. 1999, 111, [85] J. Uddin, C. Boehme, G. Frenking, Organometallics 2000, 19, 571. [86] D. Weiss, T. Steinke, M. Winter,. A. Fischer,. Froehlich, J. Uddin, G. Frenking, Organometallics 2000, 19, [87] D. Weiss, M. Winter, K. Merz, A. Knüfer,. A. Fischer,. Fröhlich, G. Frenking, Polyhedron 2002, 21, 535. [88] W. Uhl, S. Melle, G. Frenking, M. Hartmann, Inorg. Chem. 2001, 40, 750. [89] M. L. H. Green, P. Mountford, G. J. Smout, S.. Speel, Polyhedron 1990, 9, [90] O. T. Beachley,. Blom, M.. Churchill, K. Faegri, J. C. Fettinger, J. C. Pazik, L. Victoriano, Organometallics 1989, 8, 346. [91] H. Werner, H. Otto, H. J. Krauss, J. Organomet. Chem. 1986, 315, C57. [92] H. Sitzmann, M. F. Lappert, C. Dohmeier, C. Üffing, H. Schnöckel, J. Organomet. Chem. 1998, 561, 203. [93] C. Dohmeier, D. Loos, H. Schnöckel, Angew. Chem. Int. Ed. 1996, 35, 129. [94] J. Gauss, U. Schneider,. Ahlrichs, C. Dohmeier, H. Schnöckel, J. Am. Chem. Soc. 1993, 115, [95] H. Schumann, T. Ghodsi, L. Esser, Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1992, C48,

203 VI. Literature index. [96] D. G. Tuck, Chem. Alum., Gallium, Indium Thallium 1993, 430. [97] C. A. Tolman, Chem. ev. 1977, 77, 313. [98] A. Frazer, B. Piggott, M. B. Hursthouse, M. Mazid, J. Am. Chem. Soc. 1994, 116, [99] B. Twamley, P. P. Power, Angew. Chem. Int. Ed. 2000, 39, [100] M. iemeyer, P. P. Power, Angew. Chem. Int. Ed. 1998, 37, [101] J. Su, X.-W. Li, C. Crittendon, G. H. obinson, J. Am. Chem. Soc. 1997, 119, [102] T. Yamaguchi, K. Ueno, H. Ogino, Organometallics 2001, 20, 501. [103]. J. Hardman,. J. Wright, A. D. Phillips, P. P. Power, J. Am. Chem. Soc. 2003, 125, [104] T. E. Mueller, D. M. P. Mingos, Transition Met. Chem. 1995, 20, 533. [105] D. L. eger, D. G. Garza, A. L. heingold, G. P. A. Yap, Organometallics 1998, 17, [106]. J. Baker, C. Jones, J. A. Platts, Dalton Trans. 2003, [107] E. S. Schmidt, A. Schier, H. Schmidbaur, J. Chem. Soc., Dalton Trans. 2001, 505. [108] Y. Segawa, M. Yamashita, K. ozaki, Science 2006, 314, 113. [109] A. H. Cowley, C. L. B. Macdonald, J. S. Silverman, J. D. Gorden, A. Voigt, Chem. Commun. 2001, 175. [110] J.. Jones, C. L. B. Macdonald, J. D. Gorden, A. H. Cowley, J. Organomet. Chem. 2003, 666, 3. [111] C. L. B. Macdonald, A. H. Cowley, J. Am. Chem. Soc. 1999, 121, [112] B. Buchin, C. Gemel, T. Cadenbach,. Schmid,. A. Fischer, Angew. Chem. Int. Ed. 2006, 45, [113] C. L. B. MacDonald, A. M. Corrente, C. G. Andrews, A. Taylor, B. D. Ellis, Chem. Commun. 2004, 250. [114] W. Uhl, M. Pohlmann,. Wartchow, Angew. Chem. Int. Ed. 1998, 37, 961. [115] A. Sundermann, M. eiher, W. W. Schoeller, Eur. J. Inorg. Chem. 1998, 305. [116] C.-H. Chen, M.-L. Tsai, M.-D. Su, Organometallics 2006, 25, [117] W. A. Herrmann, Angew. Chem. Int. Ed. 2002, 41, [118] C. J. Carmalt, A. H. Cowley, Adv. Inorg. Chem. 2000, 50, 1. [119] D. Bourissou, O. Guerret, F. P. Gabbaie, G. Bertrand, Chem. ev. 2000, 100, 39. [120] A. H. Cowley, Chem. Commun. 2004, [121] J. D. Gorden, C. L. B. Macdonald, A. H. Cowley, Chem. Commun. 2001,

204 VI. Literature index. [122] P. Jutzi, B. eumann, G. eumann, L. O. Schebaum, H.-G. Stammler, Organometallics 2001, 20, [123]. J. Hardman, P. P. Power, J. D. Gorden, C. L. B. Macdonald, A. H. Cowley, Chem. Commun. 2001, [124] H. Jacobsen, H. Berke, S. Doering, G. Kehr, G. Erker,. Froehlich, O. Meyer, Organometallics 1999, 18, [125] Z. Yang, X. Ma,. B. Oswald, H. W. oesky, H. Zhu, C. Schulzke, K. Starke, M. Baldus, H.-G. Schmidt, M. oltemeyer, Angew. Chem. Int. Ed. 2005, 44, [126] B. Buchin, C. Gemel, T. Cadenbach, I. Fernandez, G. Frenking,. A. Fischer, Angew. Chem. Int. Ed. 2006, 45, [127] S. Aldridge, Angew. Chem. Int. Ed. 2006, 45, [128] A. T. T. Hsieh, Inorg. Chim. Acta 1975, 14, 87. [129] P. J. Brothers, P. P. Power, Adv. Organomet. Chem. 1996, 39, 1. [130]. A. Compton,. J. Errington,. C. orman, Adv. Organomet. Chem. 1990, 31, 91. [131] W. Uhl, S. U. Keimling, W. Hiller, M. eumayer, Chem. Ber. 1995, 128, [132] C. Üffing, A. Ecker,. Köppe, H. Schnöckel, Organometallics 1998, 17, [133] W. Uhl, M. Pohlmann, Organometallics 1997, 16, [134] Q. Yu, A. Purath, A. Donchev, H. Schnöckel, J. Organomet. Chem. 1999, 584, 94. [135] P. Jutzi, B. eumann, G. eumann, L. O. Schebaum, H.-G. Stammler, Organometallics 1999, 18, [136] P. Jutzi, B. eumann, L. O. Schebaum, A. Stammler, H.-G. Stammler, Organometallics 1999, 18, [137] J. Weiss, D. Stetzkamp, B. uber,. A. Fischer, C. Boehme, G. Frenking, Angew. Chem. Int. Ed. 1997, 36, 70. [138] E. V. Grachova, P. Jutzi, B. eumann, L. O. Schebaum, H.-G. Stammler, S. P. Tunik, J. Chem. Soc., Dalton Trans. 2002, 302. [139] E. V. Grachova, P. Jutzi, B. eumann, H.-G. Stammler, Dalton Trans. 2005, [140] D. Weiss, T. Steinke, M. Winter,. A. Fischer,. Fröhlich, J. Uddin, G. Frenking, Organometallics 2000, 19, [141] M. Cokoja, T. Steinke, C. Gemel, T. Welzel, M. Winter, K. Merz,. A. Fischer, J. Organomet. Chem. 2003, 684, 277. [142] T. Pechmann, C. D. Brandt, H. Werner, Angew. Chem. Int. Ed. 2000, 39, [143] D. Weiss, M. Winter,. A. Fischer, C. Yu, K. Wichmann, G. Frenking, Chem. Commun. 2000,

205 VI. Literature index. [144] T. Steinke, C. Gemel, M. Winter,. A. Fischer, Angew. Chem. Int. Ed. 2002, 41, [145] A. T. T. Hsieh, M. J. Mays, Inorg. uc. Chem. Lett. 1971, 7, 223. [146] A. T. T. Hsieh, M. J. Mays, J. Organomet. Chem. 1972, 37, 9. [147] A. T. T. Hsieh, M. J. Mays, J. Chem. Soc., Dalton Trans. 1972, 516. [148] F. P. Gabbai, S.-C. Chung, A. Schier, S. Krüger,. ösch, H. Schmidbaur, Inorg. Chem. 1997, 36, [149] C. Dohmeier, H. Krautscheid, H. Schnöckel, Angew. Chem. Int. Ed. 1994, 33, [150] P. Jutzi, B. eumann, L. O. Schebaum, A. Stammler, H.-G. Stammler, Organometallics 2000, 19, [151] M. Cokoja, C. Gemel, T. Steinke, F. Schroeder,. A. Fischer, Dalton Trans. 2005, 44. [152] T. Steinke, C. Gemel, M. Cokoja, M. Winter,. A. Fischer, Chem. Commun. 2003, [153] X. Li, H. Song, L. Duan, C. Cui, H. W. oesky, Inorg. Chem. 2006, 45, [154] H. Zhu, J. Chai, C. He, G. Bai, H. W. oesky, V. Jancik, H.-G. Schmidt, M. oltemeyer, Organometallics 2005, 24, 380. [155] H. Zhu, J. Chai, A. Stasch, H. W. oesky, T. Blunck, D. Vidovic, J. Magull, H.-G. Schmidt, M. oltemeyer, Eur. J. Inorg. Chem. 2004, [156] H. Zhu, J. Chai, Q. Ma, V. Jancik, H. W. oesky, H. Fan,. Herbst-Irmer, J. Am. Chem. Soc. 2004, 126, [157] Y. Peng, H. Fan, H. Zhu, H. W. oesky, J. Magull, C. E. Hughes, Angew. Chem. Int. Ed. 2004, 43, [158] H. W. oesky, S. S. Kumar, Chem. Commun. 2005, [159] A. Kempter, C. Gemel,. A. Fischer, Chem. Eur. J. 2007, 13, [160] A. Kempter, Diploma-Thesis [161]. Jackstell, S. Harkal, H. Jiao, A. Spannenberg, C. Borgmann, D. oettger, F. ierlich, M. Elliot, S. iven, K. Cavell, O. avarro, M. S. Viciu, S. P. olan, M. Beller, Chem. Eur. J. 2004, 10, [162] J. Krause, G. Cestaric, K.-J. Haack, K. Seevogel, W. Storm, K.-. Poerschke, J. Am. Chem. Soc. 1999, 121, [163] U. Anandhi, P.. Sharp, Angew. Chem. Int. Ed. 2004, 43, [164]. J. Baker, C. Jones, J. A. Platts, J. Am. Chem. Soc. 2003, 125, [165]. J. Baker, C. Jones, M. Kloth, J. A. Platts, Organometallics 2004, 23, [166] J. A. Labinger, J. E. Bercaw, ature 2002, 417,

206 VI. Literature index. [167] T. Steinke, M. Cokoja, C. Gemel, A. Kempter, A. Krapp, G. Frenking, U. Zenneck,. A. Fischer, Angew. Chem. Int. Ed. 2005, 44, [168] T. Cadenbach, C. Gemel,. Schmid, S. Block,. A. Fischer, Dalton Trans. 2004, [169] D. J. Patmore, W. A. G. Graham, J. Chem. Soc., Chem. Commun. 1965, 1. [170] D. J. Patmore, W. A. G. Graham, Inorg. Chem. 1966, 5, [171] D. J. Patmore, W. A. G. Graham, Inorg. Chem. 1966, 5, [172] J. Chatt, C. Eaborn, P.. Kapoor, J. Organomet. Chem. 1970, 23, 109. [173] S. M. Hawkins, P. B. Hitchcock, M. F. Lappert, J. Chem. Soc., Chem. Commun. 1985, [174] E. Bittersmann, K. Hildenbrand, A. Cervilla, P. Lahuerta, J. Organomet. Chem. 1985, 287, 255. [175] M. Stender, B. E. Eichler,. J. Hardman, P. P. Power, J. Prust, M. oltemeyer, H. W. oesky, Inorg. Chem. 2001, 40, [176] H. Werner, M. Bosch, M. E. Schneider, C. Hahn, F. Kukla, M. Manger, B. Windmuller, B. Weberndorfer, M. Laubender, J. Chem. Soc., Dalton Trans. 1998, [177] F. P. Gabbai, A. Schier, J. iede, A. Sladek, H. W. Goerlitzer, Inorg. Chem. 1997, 36, [178]. J. Hardman, B. E. Eichler, P. P. Power, Chemical Communications (Cambridge) 2000, [179] S. Singh, H.-J. Ahn, A. Stasch, V. Jancik, H. W. oesky, A. Pal, M. Biadene,. Herbst-Irmer, M. oltemeyer, H.-G. Schmidt, Inorg. Chem. 2006, 45, [180] A. Dashti, K. iediek, B. Werner, B. eumueller, Z. Anorg. Allg. Chem. 1997, 623, 394. [181] A. J. Downs, Editor, Chemistry of Aluminum, Gallium, Indium and Thallium, [182] X.-W. Li, J. Su, G. H. obinson, Chem. Commun. 1996, [183] W. Uhl, A. Ei-Hamdan, G. Geiseler, K. Harms, Z. Anorg. Allg. Chem. 2004, 630, 821. [184] J. B. Hill, S. J. Eng, W. T. Pennington, G. H. obinson, J. Organomet. Chem. 1993, 445, 11. [185] A. Schnepf, C. Doriat, E. Moellhausen, H. Schnoeckel, Chem. Commun. 1997, [186] W. Uhl, L. Cuypers, K. Schuler, T. Spies, C. Strohmann, K. Lehmen, Z. Anorg. Allg. Chem. 2000, 626, [187]. J. Baker, C. Jones, M. Kloth, J. A. Platts, Angew. Chem. Int. Ed. 2003, 42,

207 VI. Literature index. [188] G. Linti, A. odig, W. Kostler, Z. Anorg. Allg. Chem. 2001, 627, [189]. Wiberg, T. Blank, K. Amelunxen, H. oth, J. Knizek, T. Habereder, W. Kaim, M. Wanner, Eur. J. Inorg. Chem. 2001, [190]. J. Hardman,. J. Wright, A. D. Phillips, P. P. Power, Angew. Chem. Int. Ed. 2002, 41, [191] W. Uhl, A. El-Hamdan, A. Lawerenz, Eur. J. Inorg. Chem. 2005, [192] H. Zhu, J. Chai, V. Jancik, H. W. oesky, W. A. Merrill, P. P. Power, J. Am. Chem. Soc. 2005, 127, [193] S. Singh, V. Jancik, H. W. oesky,. Herbst-Irmer, Inorg. Chem. 2006, 45, 949. [194] V. Jancik, Y. Peng, H. W. oesky, J. Li, D. eculai, A. M. eculai,. Herbst-Irmer, J. Am. Chem. Soc. 2003, 125, [195]. J. Hardman, P. P. Power, Inorg. Chem. 2001, 40, [196] Y. Peng, H. Fan, V. Jancik, H. W. oesky,. Herbst-Irmer, Angew. Chem. Int. Ed. 2004, 43, [197] A. J. Arduengo, III, H. V.. Dias, J. C. Calabrese, F. Davidson, Inorg. Chem. 1993, 32, [198] B. Gehrhus, P. B. Hitchcock, M. F. Lappert, Dalton Trans. 2000, [199]. Kuhn, T. Kratz, D. Blaeser,. Boese, Chem. Ber. 1995, 128, 245. [200]. Kuhn, C. Maichle-Mobmer, E. iquet, I. Walker, Z. aturforsch., B: Chem. Sci. 2002, 57, 47. [201] F. Stabenow, W. Saak, M. Weidenbruch, Chem. Commun. 1999, [202] A. Schaefer, M. Weidenbruch, W. Saak, S. Pohl, J. Chem. Soc., Chem. Commun. 1995, [203] W. eimann, H. G. Kuivila, D. Farah, T. Apoussidis, Organometallics 1987, 6, 557. [204]. Wiberg, K. Amelunxen, H. W. Lerner, H. oth, W. Ponikwar, H. Schwenk, J. Organomet. Chem. 1999, 574, 246. [205]. Wiberg, T. Blank, M. Westerhausen, S. Schneiderbauer, H. Schnockel, I. Krossing, A. Schnepf, Eur. J. Inorg. Chem. 2002, 351. [206]. Wochele, W. Schwarz, K. W. Klinkhammer, K. Locke, J. Weidlein, Z. Anorg. Allg. Chem. 2000, 626, [207] A. F. Holleman, E. Wieberg, Lehrbuch der Anorganischen Chemie 1995, Walter de Gruyter Verlag (Berlin, ew York). [208] M. Brookhart, B. Grant, A. F. Volpe, Jr., Organometallics 1992, 11,

208 VI. Literature index. [209] Y. S. Goldberg, I. G. Iovel, M. V. Shymanska, J. Chem. Soc., Chem. Commun. 1986, 286. [210] M. Green, T. A. Kuc, J. Chem. Soc., Dalton Trans. 1972, 832. [211] L. Malatesta, L. aldini, G. Simonetta, F. Cariati, Coord. Chem. ev. 1966, 1, 255. [212] S. Scholz, H. W. Lerner, M. Bolte, Acta Crystallogr., Sect. E: Struct. ep. Online 2002, E58, m586. [213] M. iger, F. Thomas, private communication 2002, CCDC [214] S. D. ogai, H. Schmidbaur, Dalton Trans. 2003, [215] M. anjo, T. Oda, K. Mochida, J. Organomet. Chem. 2003, 672, 100. [216] T. F. Faessler, M. Hunziker, M. E. Spahr, H. Lueken, H. Schilder, Z. Anorg. Allg. Chem. 2000, 626, 692. [217] W. Uhl, M. Benter, M. Prött, J. Chem. Soc., Dalton Trans. 2000, 643. [218] B. Buchin, T. Steinke, C. Gemel, T. Cadenbach,. A. Fischer, Z. Anorg. Allg. Chem. 2005, 631, [219] S. Tobisch, T. Ziegler, J. Am. Chem. Soc. 2002, 124, [220] G. Wilke, Angew. Chem. 1988, 100, 189. [221] B. Bogdanovic, M. Kroener, G. Wilke, Justus Liebigs Annalen der Chemie 1966, 699, 1. [222] G. Wilke, M. Kroener, B. Bogdanovic, Angew. Chem. 1961, 73, 755. [223] K.. Poerschke, G. Wilke, Chem. Ber. 1984, 117, 56. [224] K.. Poerschke, G. Wilke,. Mynott, Chem. Ber. 1985, 118, 298. [225] D. J. Brauer, C. Krueger, J. Organomet. Chem. 1972, 44, 397. [226] G. Wilke, Angew. Chem. Int. Ed. 1988, 27, 185. [227] K. Fischer, K. Jonas, G. Wilke, Angew. Chem. 1973, 85, 620. [228] C. Krueger, Y.-H. Tsay, J. Organomet. Chem. 1972, 34, 387. [229] B. Gabor, C. Krueger, B. Marczinke,. Mynott, G. Wilke, Angew. Chem. Int. Ed. 1991, 30, [230] L. Cronin, C. L. Higgitt,. Karch,.. Perutz, Organometallics 1997, 16, [231] P. B. Hitchcock, M. F. Lappert, C. MacBeath, F. P. E. Scott,. J. W. Warhurst, J. Organomet. Chem. 1997, 528, 185. [232] P.. ylander, Engelhard Industries, Technical Bulletin 1960, 1, 93. [233] J. L. Garnett,. J. Hodges,. S. Kenyon, M. A. Long, J. Chem. Soc., Perkins Trans. 1979,

209 VI. Literature index. [234] E. A. V. Ebsworth, V. M. Marganian, F. J. S. eed,. O. Gould, J. Chem. Soc., Dalton Trans. 1978, [235] G. Linti, H. Schnöckel, Coord. Chem. ev. 2000, 285. [236] B. Buchin, C. Gemel, T. Cadenbach,. A. Fischer, Inorg. Chem. 2006, 45, [237] M. Cokoja, H. Parala, M.-K. Schroeter, A. Birkner, M. W. E. Van den Berg, W. Gruenert,. A. Fischer, Chem. Mater. 2006, 18, [238] T. Steinke, C. Gemel, M. Cokoja, M. Winter,. A. Fischer, Angew. Chem. 2004, 116, [239]. Iwasa, T. Mayanagi,. Ogawa, K. Sakata,. Takezawa, Catal. Lett. 1998, 54, 119. [240] T. Komatsu, K. Inaba, T. Uezono, A. Onda, T. Yashima, Appl. Catal., A 2003, 251, 315. [241] A. Kempter, C. Gemel,. A. Fischer, Chem. Eur. J. 2007, 13, 2990 [242].. Schrock, L. P. H. Lopez, J. Hafer,. Singh, A. Sinha, P. Mueller, Organometallics 2005, 24, [243] T. Schaub, M. Backes, U. adius, Organometallics 2006, 25, [244] P. Macchi, D. M. Proserpio, A. Sironi, J. Am. Chem. Soc. 1998, 120, [245] S. Pasynkiewicz, A. Pietrzykowski, B. Kryza-iemiec, J. Zachara, J. Organomet. Chem. 1998, 566, 217. [246] V. W. Day, S. S. Abdel-Meguid, S. Dabestani, M. G. Thomas, W.. Pretzer, E. L. Muetterties, J. Am. Chem. Soc. 1976, 98, [247]. S. Dickson, J. A. Ibers, J. Organomet. Chem. 1972, 36, 191. [248]. Goddard, B. Apotecher, H. Hoberg, Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1987, C43, [249] X. Chang, K.-E. Lee, I. Jeon Sang, Y.-J. Kim, H. K. Lee, W. Lee Soon, Dalton Trans. 2005, [250]. Jackstell, G. A. Andreu, A. Frisch, K. Selvakumar, A. Zapf, H. Klein, A. Spannenberg, D. ottger, O. Briel,. Karch, M. Beller, Angew. Chem. Int. Ed. 2002, 41, 986. [251]. Trebbe,. Goddard, A. ufinska, K. Seevogel, K.-. Poerschke, Organometallics 1999, 18, [252] M. Green, J. A. Howard, J. L. Spencer, F. G. A. Stone, J. Chem. Soc., Chem. Commun. 1975, 3. [253] T. F. Fassler, Coord. Chem. ev. 2001, 215,

210 VI. Literature index. [254] H. Matsumoto, K. Higuchi, Y. Hoshino, H. Koike, Y. aoi, Y. agai, J. Chem. Soc., Chem. Commun. 1988, [255] A. Schnepf, Chem. Soc. ev. 2007, 36, 745. [256] J. D. Corbett, Angew. Chem. Int. Ed. 2000, 39, 670. [257] B. E. Eichler, P. P. Power, Angew. Chem. Int. Ed. 2001, 40, 796. [258]. Wiberg, H.-W. Lerner, S. Wagner, H. oth, T. Seifert, Z. aturforsch., B: Chem. Sci. 1999, 54, 877. [259] A. F. ichards, B. E. Eichler, M. Brynda, M. M. Olmstead, P. P. Power, Angew. Chem. Int. Ed. 2005, 44, [260] M. Brynda,. Herber, P. B. Hitchcock, M. F. Lappert, I. owik, P. P. Power, A. V. Protchenko, A. uzicka, J. Steiner, Angew. Chem. Int. Ed. 2006, 45, [261] E.. Esenturk, J. C. Fettinger, B. W. Eichhorn, J. Am. Chem. Soc. 2006, 128, 12. [262] A. Schnepf, H. Schnöckel, Angew. Chem. 2002, 114, [263] J. Vollet, J.. Hartig, H. Schnoeckel, Angew. Chem. Int. Ed. 2004, 43, [264] T. Steinke, Dissertation [265] A. Van der Ent, A. L. Onderdelinden, Inorg. Syn. 1990, 28, 90. [266] D. L. eger, C. A. Little, J. J. S. Lamba, K. J. Brown, J.. Krumper,. G. Bergman, M. Irwin, J. P. Fackler, Jr., Inorg. Syn. 2004, 34, 5. [267] P. Braunstein, H. Lehner, D. Matt, Inorg. Syn. 1990, 27, 218. [268] D. J. Krysan, P. B. Mackenzie, J. Org. Chem. 1990, 55, [269] M. Green, J. A. K. Howard, J. L. Spencer, F. G. A. Stone, J. Chem. Soc., Dalton Trans. 1977,

211 VII. Appendix. 1. Computational details. Figure 37: Structures of not-activated species [i 2 (μ-ga(bisimidinate))(c 2 H 4 ) 4 ] (19a) (left) and activated species [i 2 (μ-ga(bisimidinate))(c 2 H 4 ) 3 (H)(C 2 H 3 )] (19b) (right). Figure 38: Structures of not-activated species [i 2 (μ-ga(bisimidinate))(c 2 H 4 ) 3 ] (20a) (left) and activated species [i 2 (μ-ga(bisimidinate))(c 2 H 4 ) 2 (H)(C 2 H 3 )] (20b) (right). 195

212 Figure 39: Structures of not-activated species [i 3 (μ-ga(bisimidinate)) 2 (C 2 H 4 ) 3 ] (27a) (left) and activated species [i 3 (μ-ga(bisimidinate)) 2 (C 2 H 4 ) 2 (H)(C 2 H 3 )] (27b) (right). Table 46: Basis set comparison for the calculation of ΔE (Sum of electronic and Zero-Point- Energy). ΔE in Hartree Δ (xa xb) ΔE in Hartree [6-311G**(C, H, Ga) / Δ (xa xb) (LanL2DZ) In kcal/mol Stuttgart SC 1997 ECP (i)] in Hartree In kcal/mol 1a , b , , ,48 2a , b , , ,88 3a , b , , ,89 196