Organometallic Compounds of Transition Metals ‘As was shown in Part 1, the Main Group elements are divided into the s-block (alkali and alkaline earth elements) and the p-block (from the boron group to noble gases) which, as this classification signifies, use s- and p-orbitals for chemical bonding. Inserted between these element blocks are the so-called transition metals. Going from left to right across the Periodic Table, from Period 4 (main quantum number 4) onwards, d-orbitals begin to be filled. The elements arising from the filling of the 3d, 4d, and 5d shells are commonly referred to as first row, second row, and third row transition metals, respectively, and each group of elements with identical d-electron count forms a triad. As shown in the following table, the group number is conveniently identical to the number of d-electrons of each ele- ment in its zero-valent state. The five d-orbitals are aligned along, as well as in between, the x, y, and z axes: day dy.» dy dey and dy, so that orbitals with appropriat® symmetry are available for bonding in many directions. From Group 4 onwards the chemical differences between transition and Main Group’elements become very pronounced, and the term ‘transition metals’ is therefore usually applied to the elements in Groups 4-10, These elements show a particularly rich chemistry of varying oxidation states and the most unusual bonding and reactivity patterns, in organometallic chemistry, which is the reason for their importance in industrial processes and a multitude of catalytic applications. Starting with scandium (Sc) the 3d energy level sinks below the 4s level, so that the chemistry of scandium and the following elements is dictated by the occupancy of these d-levels. Group 3 7 8 9 0 1 2 ne [a [a Ja Ja aS a5) ase Mn [Fe _(Co_INi [cu [2 5 id Pd_lag_|co 6 ta IRe los Pt [au [Hg 7 |ae oP pop pp pp po pe pe pe pe reer [Nd_[em [Sm feu [Gd [Te [oy [Wo Jer [tm [Yb [uw mh [Pa_|U —|Np [Pu [Am [cm Jak [cf [es [Fm _Imd_[No_ [tr Structure of the n block82 Electron configuration: oo Nora's? Lanthanides Actinides Lanthanide contraction Coinage metals 2: ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS Box 2.0 Electron configuration and d-electron count ‘The electron configuration of transition metal atoms is often quoted as [noble gas] nd(n+1) 8) (ie. [Ar]3d°4s" for titanium, for example). However, as stated above, from scandium ‘onwards the d level of transition elements sinks below the s level. Also for the metal fons, d is always below s. In a chemical environment the effective electron configuration of transition metals is therefore d', a, d®, ... d'°: ‘The number of d-electrons of the element is identical to the group number. ‘The table on the preceding page gives the d-electron configuration for each transition metal triad (oxidation state 0). This will be important for understanding electron counting and the magnetism of transition metal complexes. Up to Group 7 the number of d-electrons gives the maximum obtainable oxidation state. ‘The elements in Group 3 (Sc, Y, and La) exist mainly as M°* ions; their chemistry is not unlike that of aluminium, and they are referred to as ‘rare earth’ metals. ‘The element lanthanum is followed by 14 elements which arise by filling the seven 4f orbitals, The elements from cerium (Ce) to lutetium (Lu) are called the ‘lanthanides: £-Orbitals are not significantly involved in covalent bonding, and these elements form mainly compounds based on M°* ions which are chemically generally similar. They do how- ever provide some interesting magnetic and electrophilic organometallic compounds. In analogous fashion, the elements following actinium (Ac) are referred to as the ‘acti- ides: Of these, only thorium and uranium have a significant organometallic chemistry. Since 14 elements are inserted between lanthanum and the next transition element, haf- nium, the increased nuclear charge leads to a significant contraction of the atomic and ionic radii: the so-called ‘lanthanide contraction’ This has important structural and bonding consequences: the radii of Row 6 transition metals tend to be smaller than those of Row 5 metals, while the metal-carbon bonds become stronger, for example. Group 11 elements (Cu, Ag, Au), the coinage metals, behave in their higher oxidation states like transition metals, while their M* ions have a d'” electron configuration, ie, a com- pletely filled d-shell, and therefore have Main Group characteristics. The Group 12 elements, finally, are best regarded as Main Group elements; they have therefore been included in Part 1 General Trends: Electronegativity and Bond Strength The following table lists the Pauling electronegativities (EN) of transition metals. For com- parison, the electronegativities of C and H are 2.5 and 2.2, respectively. Ti v Cc oMn Fe GON 151606 aS zr Nb Mo Te Ru Rh PAB increasing increasing 16 16189222222. stability M-C bond ofhigher strength Ho Taw Re 0st Pt AU oxidation eles ete eel erg e192 22) 2 2i et esta more polar M-Co-bonds, <=—=—————————> more covalent M-C o-bonds2: ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS ‘Metals are more electropositive than carbon and hydrogen, and in M-X bonds the metals Mare positively polarized. However, the Row 5 and 6 elements in Groups 8-11 from ruthe- njum to gold are collectively referred to as the noble metals (shaded), and they are char- acterized by high electronegativities and covalent bonding to carbon, such that an Au-C a-bond, for example, is non-polar and generally stable to air and moisture. ‘The metal-carbon bond strength in the transition metal block increases as one goes down a given triad from Rows | to 3. This is opposite to the bond energy trend described for Main Group elements and reflects the involvement of d-orbitals in M~C bonding, Third row met- als tend to make the most thermodynamically stable complexes, and are usually less reactive and less involved in catalysis than second row metals (with some exceptions). 8384 2: ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS 2.1 Ligand Types Whereas the organometallic chemistry of Main Group elements is dominated by compounds with metal-carbon 9-bonds, transition metals are capable of binding to a much greater vari- ty of ligands, including those that interact with the metal through their n-electron systems. Asa very general classification, ligands come in two types: (1) L-type ligands donate an electron pair to the metal, and (2) X-type ligands, where X is an electronegative element that forms M-X o-bonds; these ‘may be polar or covalent and can be expressed by a resonance form that accumulates, negative charge on X: ‘There are also more complex forms of interaction where a ligand has both L- and X- characteristics, LX and LyX. Ligand types are indicated in the following table. Ligands contribute electrons to the metal centre, and the stability and reactivity of such complexes depend strongly on the total electron count: total number of electrons per metal centre = number of d-electrons + sum of ligand electrons — overall charge. ‘There are various ways of determining the electron count and the number of electrons pro- vided by a given ligand. Here, we will use the ‘neutral ligand’ approach and consider each ligand asa neutral entity. This implies that for the purpose of electron counting only the metal is con- sidered as zero-valent, so that there is no need to worry about its oxidation state at this stage, Ligand type electron ligands: xetype 2electron ligands: Ltype ‘Type Examples olen i yan, Ca . terminal halides, alkoxides, amides: ~Cl, ~Br, 1, ~OR, -NR> odor 4 Fi 4 u— Tmt mf ayers, sane an atane composes 4 ae non, R reacceptor lpr ey wc] , h 7 cnbenes sens ronan) an acceptor R ligands Re alkenes:2.1 LIGAND TYPES 85 Ligand type 3-electron ligands: Letype 4-clectron ligands: 3 <<, “§ Ren ome “he = a and LX type S-electron ligands: Lx ype 6electron ligands 7-electron ligands B-electron ligands Type Examples terminal and bridging ap-allyl,1?-enyl carbyne (alkylidyne) eae w-indenyl DA A 0 M& NR! bridging halide, alkoxide, amide ni-benzyt nP-acyl — / orrD, VE. oe a V ~~ " nt-diene, n2,n?-diene, nf-cyclobutadienyl ‘nf-trimethylenemethane conjugated sanconugated R 5 eed YA S oe wT \ > nfantelooctadiene bridging atyne ees Ge i n’-cyelopentadieny! n®-indenyl —_—_n®-pentadienyl_nf-cyclohexadienyl_ boratabenzene complex 3 Gar) M M nf-arene ‘eycloheptatriene _®-cyclododecatriene ™ n®-cyclooctatetraenyl86 2: ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS 2.2 Common Types of Organometallic Complexes ‘The ligands classified in the previous section form various types of complexes with transi- tion metals, as well as the lanthanides and actinides. Their structures, thermal stability, and reactivity depend on the electronic characteristics of the metal centre on the one hand, and ‘on the steric requirement of the ligands on the other. Complexes that contain only one type of ligand are called homoleptic; compounds containing only two types of atoms are called binary. Ligands may occupy one, two, three, or more coordination sites and are correspondingly termed mono-, bi, tri-, and polydentate ligands. Ligands attached with more than one atom to the same metal tend to form significantly more stable complexes than monodentate li gands (the ‘chelate effect’). While the most common coordination number is 6 (octahedron or trigonal prism) and 7-, 8, and 9-coordinate complexes are found, steric or electronic reasons may impose al- ternative and lower coordination numbers. Complexes of metal centres with a d* electron configuration are predominantly square-planar, this holds true particularly for second and third row metals and if the ligands are strong donors. Where steric factors prevail, the com- plexes will be tetrahedral since in this geometry steric repulsion is minimized, Metal centres in positively charged complexes are more Lewis acidic than their neutral counterparts and ofien have a higher coordination number, Representative types of complexes with monodentate ligands are the examples of halide, alkyl, CO, and phosphine complexes shown in Table 2.2.1 For coordination numbers of 5 and 6, two alternative geometries are possible. For CN 5, trigonal-bipyramidal (tbp) and square-pyramidal geometries are usually separated by only a rather small energy difference, which provides a pathway for ligand scrambling. 1-Donor and clectronegative ligands favour tbp (as in NbCl,Me;), whereas the o-only complex Tae, is square-pyramidal in the gas phase. For CN 6 and d’-d’ electron configurations supported by strong o-donors (methyl ligands), a trigonal-prismatic geometry is favoured over the conventional octahedron. However, in most other cases, and certainly with 1-acceptor li- gands, the octahedral geometry is more stable. ‘The importance of steric hindrance. Complexes with coordination number 3 are rare. As can be seen from the examples in Table 2.2.1, in order to be able to isolate complexes of low coordination number (as opposed to generating them at low temperatures for spectroscopic analysis), ligands of high steric hindrance are required. ‘These provide kinetic shielding. For example, whereas TiMe, is thermally unstable, the bulkier Ti(CH,SiMe,), is stable at room temperature. “The bulk of the ligand often dictates the coordination number and the metal oxidation state; for example, alkylation of CrCl; gives three-coordinate Cr(II1) products CrR; if CH(SiMe;), but Cr(IV) alkyls if R = CH,SiMe,, Steric hindrance is often used to control coordination number, particularly for metals that act as precursors for catalysts. For Pd(0) complexes, which generally form coordina- tively and electronically saturated, 18 VE phosphine complexes PdL,, itis often useful to employ bulky phosphines to facilitate dissociation of L (to give PdLs or PdL) or to prevent2.2 COMMON TYPES OF ORGANOMETALLIC COMPLEXES, 87 Table 2.2.1 Common coordination geometries for organometallic transition metal complexes MXqy My n=2 n=3 n=4 Geometry wa linear trigonal-planar tetrahedral square-planar (sap) Halide HgCly TiCl, [Coch (pact)? Metal alkyls RyC-MIn-CR, R =SiMey (CHR); R SiMe, TiMey {Puc=cR),)* Ti(CH,SiMe,), Metal carbonyl (Au(CO),I" d¥,14 VE [CuCO)sI @°,16VE NICO), 18VE _RH(CO),]" d, 16 VE Phosphine PAIPBUGPH)2 (RSPIAUCL ——_NILPEL)), Ni(PMe) on [RH(PMe)l', complex NiCL(PPh,), NiCl,(PMe,), MX, Ml 5 n n ne tna 7 vile hos “| 7N ale trigonal-bipyramidal top) _square-pyramidal octahedral trigonal-prismatic Halide MoCl, (Tici?>, Wel Aly! NbCIMes Tame, [Rhmeg?™ Carbonyl Fe(CO), d3,18 VE Cr(CO), 46,18 VE Phosphine Fethte.)(Me,PC! RiC=RUCIL(PR)2 [Fe(mesPCHy complex CHPMe:), CHsPMe,)5I”" Rh(+(CO}(PPH)s WC1,(PMe;Ph)s higher coordination altogether, as in the two-coordinate Pd(PBu'sPh),, Three-coordinate complexes of d® metals retain a square geometry but with one site unoccupied, which results in a T-shaped structure. Steric hindrance can also be used to suppress certain reaction pathways or enable com- pletely different reactivity. For example, iron complexes of bis(imino)pyridines with small phenyl substituents on N give bis(ligand) complexes [Fe(NNN)9I?* which have long been studied for their redox behaviour, However, bulky imine- substituents allow the coordina- tion of only one ligand per metal, and the complexes become highly active ethylene polym- erization catalysts.88 2; ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS K cSyvacant site less active more active 20 Ss ™ N an NY inactive as catalyst highly active r-Ligands. Organometallic complexes often contain neutral, anionic, or formally cat onic open or closed unsaturated ligands which bind to the metal through their n-electron systems. Their structure can usually be derived from an octahedron where two or three coordination sites are occupied by s-ligands. Ligands with negative charge are classed as LX (3-electron donor) or 1X (5-clectron donor) types. The most important of these, the cyclopentadienyl anion CH" (Cp"), falls into the LX category. Complexes where the metal is enclosed between two such planar cyclic w-ligands are re- ferred to as sandwich complexes, while those containing a cyclic r-ligand as well as one or ‘more monodentate ligands are described as half-sandwich complexes, or as complexes with ‘piano sto? geometry. There are systematic series of open and closed 1-systems: Open mligands: A PX\-FAT) OY aP-allyl nfdiene, enediyl npentadienyl P-cyclohexadienyt Closed r-ligands: 2m 6n “There are isoelectronic series of sandwich and half-sandwich complexes, in various li- gand combinations, depending on the electron count of the metal centre and the number of s-clectrons supplied by the ring ligands, Large ions like U** and Th**can also form sandwich complexes with the large CgH,*" ring, a 107 Hiickel aromatic system.2.2 COMMON TYPES OF ORGANOMETALLIC COMPLEXES Sandwich complexes: Qa st @& aby Kas) G5) a) ‘The Cp ligands can be substituted as required. In order to achieve steric stabilization of otherwise very labile species, bulky substituents on the Cp ring are frequently used. One of the most important bulky Cp ligands is pentamethyl-Cp, CsMes” (usually abbreviated Cp’). ‘The introduction of five methyl substituents doubles the effective diameter of the ligand and its thickness; each Me substituent also shifts the redox potential by an incremental amount, so that MCp*; complexes are easier to reduce than their MCp, congeners. Commonly used Cp ligands: CH = = “cm, €O OR? fluorenyl A particularly widely-used class of sandwich complexes are those with two Cp rings (MCp», ‘metallocenes’), especially those carrying additional ligands, CpM(X)(Y). To ac- commodate these, the two Cp rings must be tilted, which affects the energies of the frontier orbitals. Complexes with a wide variety of ligands, Cp,ML;, CpyMXL, and Cp,MX,, exist and are used for a myriad of functionalizations, and in synthetic and catalytic applications. Connecting the two Cp ligands via a linker prevents Cp rotation and controls the coordi- nation gap aperture, for enhanced reactivity. Types of metallocene complexes: | coon bent mealocene ansa-metallocene A, B= X- or Ltype ligand ansa (Latin) = handle Z= bridging group90 Extending the 18-electron rule: see Section 2.5.2. 2: ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS Asis explained in Chapter 2.3, sandwich complexes are most stable if they attain an elec- tron count of 18. This ‘magic number is one ofa series: capping a sandwich like Cr(benzene)) with another [Cr(benzene)] unit gives a so-called triple-decker complex, Cr,(benzene)3; here the electron count is extended from 18 to 30 valence electrons (VE). Other triple- decker structures also exist, with up to 34 VE. Higher-order metallocenes: triple-decker complexes oO _ & 18VE e 2 o S. 30VE 30-34 VE 30VE Half-sandwich complexes, Transition metal fragments bearing donor or anionic ligands, such as CO, phosphines, halides, hydride, etc., form half-sandwich complexes with n-donor ligands. Coordination enhances the reactivity, and half-sandwich complexes probably con stitute the largest and most diverse class of complexes with extended n-systems. Some repre sentative examples of neutral and charged half-sandwich complexes with 7-ligands ranging from n}-allyl to aromatic ring systems are shown in the following diagram. Open-ligand half-sandwich complexes: a We 1 ! : 2 Pd? oN co oe TN oN Py NO oc Ico oct Lco oc” | Sco ry PR, oC co co co co co Representative examples of half-sandwich complexes: A,B,C = variously Lorxtypeligands EY \ Phzp “CL L=CO, PR,2.3 ELECTRON COUNTING AND THE 16/18-ELECTRON RULE 2.3 Electron Counting and the 16/18-Electron Rule 2.3.1 Counting Conventions. ‘As was illustrated in Part 1, for compounds of Main Group elements the octet rule in- dicates particular stability: the available s- and p-orbitals are fully occupied and form four 2-electron-2-centre bonds, which together give an electron octet. The equivalent for transition metals is the 18-electron rule: it, too, predicts a particular stability whenever the sum of d-electrons, ligands, and positive or negative charges reaches the ‘magic num- ber’ of 18 valence electrons (VE). It is sometimes also referred to as the 16/18-electron rule because some metals, for well-understood electronic reasons, prefer an electron count of 16. “The 18-clectron rule is best considered as a rule-of-thumb and a useful guide. It is most consistently followed by complexes with ligands like CO, which bind strongly through donor and acceptor interactions and exert strong ligand fields. A common explanation for the par- ticular stability associated with an electron count of 18 is that 18 electrons are required to fill one s-, three p-, and five d-orbitals in a metal’s valence shell, so that a noble gas configuration can be reached, More recent theoretical considerations have cast doubt on the involvement of p-orbitals in ligand bonding and suggest that the metal-ligand o*-orbitals are populated instead; however, the result is the same. Electron counting can follow two conventions. In the list of ligands in Chapter 2.1 we have considered each ligand as a neutral entity, so that a methyl or a chloride is considered as a neutral radical delivering one electron to the metal, In this way there is no need to assign oxidation states Alternatively, one can consider all ligands that are known to form stable anions as nega- tively charged, and compensate for the charges by applying the appropriate oxidation state to the metal. While this does not change the outcome, the assignment of oxidation states is not always straightforward, so itis usually simpler not to worry about the electronic structure of complexes for electron counting purposes. QJ Covalent counting convention _ tonic counting convention 1 Gy 50° Gti be fe Fe(a) d® Fe* 6 ZB Gh se Ce 6e" Ferrocene = 18VE “18 VE bis(nS-cyclopentadienybiron In compounds with metal-metal bonds, the electrons donated from one metal to another are taken into account. For example, manganese carbonyl has the structure (OC);Mn-Min(CO);. With a d-electron configuration of d” and two electrons donated by each CO ligand, the electron count of each Mn(CO); fragment is d’ + 5 x 2 = 17 electrons; the additional electron from the second manganese atom makes this 18 VE for each metal centre. 91 metal-metal bonds F°-co f2-c0 comico92 6°, square-planar: 16VE 2: ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS As was discussed in Chapter 1.1, a bridging halide (or another heteroatom with at least one lone electron pair) contributes a total of three electrons (X = halide, OR, SR, NR,, etc., L = general neutral donor ligand): i L era Kite aN", Complexes of metal centres with a d® electron configuration adopt mostly a square- planar geometry with an electron count of 16. This is particularly true for second and third row transition metals, such as Rh(I), Ir(1), Pd(H1), Pt(ID), and Au(IID, and for complexes of strong-field ligands, such as cyanide CN", and strong donor ligands like PRs. The reason is the increased stabilization of the d-shell as the atomic number increases, so that the occu- pied d,2 orbital no longer participates in ligand bonding, Some examples: z : . one Pd Pt ch, Rh " Rpm cH, ci “i RPO il Note that square-planar geometry is preferred for electronic reasons, whereas the sterically ‘most favourable geometry in complexes with four ligands is tetrahedral. ‘Therefore com: plexes of Ni™-Pt" are mostly square-planar, whereas Ni’-Pt” complexes (d"") are tetrahedral. 2.3.2 Electron Counting and Oxidation States For the purpose of oxidation state assignment, ligands that exist as free molecules are considered neutral, for example CO, ethylene, phosphines, or benzene, while those that are typically present in salts and are more electronegative than the metal are treated as negatively charged, such as Cl” or the cyclopentadienyl anion, CsH” (Cp). On occasions, ligands can also be considered as carrying a positive change, as in linearly bonded NO*, a 2-electron ligand which is isoelectronic with CO. The oxidation state of the metal is given by the charge necessary to compensate the negative charges, taking also account of any overall negative or positive charge of the complex. Oxidation states are useful for det mining the d-electron count ofa metal centre; for example, knowing that an ion is d° (Fe", Col), d (Rh', Pt"), or d'? (Ni°) will suggest adherence to octahedral, square-planar, or tetrahedral coordination geometry, respectively. The following examples illustrate these principles and also introduce different structural types of organometallic compounds. It must be kept in mind, however, that oxidation states are a formalism, not a physical property: they cannot be measured.2.3 ELECTRON COUNTING AND THE 1 © i A ee > co of *co complex _Half-sandwich complex _Half-sandwich complex Chromocene (Piano stool complex’) (7-coordinate) Electron count: Crd® Electron count: Nb d® Electron count: Crd® cp: Se Cp: 5° Cpr2x5=10. 3C0:3x2 400: 4x2=8e" levi Charge ie 18 VE 18 VE Oxidation sta 2Cp anions per Cr, overall no charge: Cr(l) Oxidation state: one Cp overall 1~ charge: Cr(0) anion no overall charge: Nb() . dK DL Ss who ey ® Nickel tetracarbonyt tetrahedral Tris(ethylene)nicket trigonal-planar Bent metallocene Electron count: Electron count: Ni d!° Electron count: Ni d?? ad ACO=+4x2=8e CH 3x2=6e~ 2k 18 VE 16VE cp:2x5 Oxidation state: neutral ligands, no overall charge: Ni(0) Oxidation state: neutral ligands, ro overall charge: NI(0) Oxidation state: 2Cp anions, 2 Ct anions, overall no charge: Zr(V) Oxidation state: one Cp 16/18-ELECTRON RULE 93 ‘complex (9-coordinate) Electron count: Re d” cp: 5e° 6H 6x1 266" 18VE Oxidation state: Cp~ plus 6H", no overall charge: Re(Vil) co Rn, 5° | co oc“ JN, 7 \Neo ok Electron count: Vd shared electrons from V=V bond: ACO=4x2 bridging PRy: 3 e~ iB VE Oxidation state: 2 PR”, no overall charge: V() Sime, ts Messi Sime, Mei, mes Ti Me,SimC —Mn—C-+"'SiMey csscr we bee Ngee ae 82 Nene 2 Sime, Tetramethyltitanium: Amanganese dialkyl, M-M bonded dimer cre tetrahedral very high steric hindrance, shared electrons from Cr-Cr bond he Electron count: Tid* Electron count: Mn d? 4 CH radicals = 4 e™ ave Oxidation state ‘alkyl anions CH no overall charge: TW) Oxidation state: 2anionic alkyl ligands, ‘no overall charge: Ma(i) Oxidation state: Pie, neutral, 2alkyl anions R” per Cr, no overall charge: Cr(t) shared electrons from bridging alkyls: 20.5 = 1 e~ terminal alkyl: 1 ~ 1 phosphine per Cr: 2 iaVE2; ORGANOMETALLIC COMPOUNDS OF TRANSITION METALS R I\ Cle, i € PrP set fp pa — | Me fj / npr, f a pm tech, I cl c a Ss 4 P Mek Square-planar, -Allylcomplex, mer Complex with alkyl, carbene and ‘allyl complex square-planar carbyne ligands Electron count: Pd d*? Electron count: Pd d?° Electron count: W dé 2PPh3= 2% e neallyl = 3 e carbyne = 3 € cali = 16° o-bonded C= 1e catbeni ce n-donor Cl=2e alkyl= 16 tee ise 2 phosphine: ie Oxidation state: Oxidation state: Oridation state: PPh3 neutral tally anion per Pa, if earbene is considered neutral, 2 anionic ligands, 1 chloride per Pa carbyne + alkyl suggest WO); no overall charge: PA(, no overallcharge: PaQID, _ifall ligands are considered anionic, W(. typical structure for d® typical structure for d® configuration configuration Whereas in most cases oxidation states can be assigned unambiguously, there are inci dents where there is room for debate. For example, 1,3-butadiene isa neutral n-donor ligand. However, itis able to accept electron density from the metal into its x* orbitals (‘back-bond- ing’), and in the case of early transition metals like zirconium, with energetically high-lying <-orbitals, this back-donation of electron density can be so extensive that the butadiene takes on the characteristics of a butenediyl dianion C,H,”", with two o-bonds to the metal. Zirconocene butadiene complexes can therefore be described as either Zr" or Zr; the bond length distribution suggests that the Zr'Y resonance form makes a major contribution, Such ambiguities are not uncommon, Whereas in the Zr diene case, the bond length distribution is an indicator, other cases are less obvious. Take for example mixed sandwich compounds of the cycloheptatrienyl ligand, such as (1°-Cp)Mo(n’-C;H;). Here Cp may be regarded as Cp” and the seven-membered ring as C;H," (the well-known, isolable tropylium cation); both are familiar planar 6 Hiickel aromatic systems, isoelectronic with benzene. This assignment would leave the metal in the oxidation state 0. On the other hand, spectroscopic and computational studies have shown that the 7-ring carries considerable negative charge and is better described as CjH,?, a 10 Hitckel aromatic system, which would leave the metal in the oxidation state +1V, The real situation is somewhere between these two extremes, 4 o DD sew Galle 68 ge oa | K ~ ~ Mo Mo(o) 6° <> Molly) d? Sh DD wire uri Zr(l) diene complex _Zr(IV) butenediyl complex Fortunately, in mostcases metal oxidation statescan be assigned morestraightforwardly. How- ever, oxidation states are a formalism and not necessarily a good guide to reactivity. Whether a metal centre will undergo a given reaction will be strongly modulated by all the ligands present, and formally similar complexes can display widely differing reactivities. For example, there are2.3 ELECTRON COUNTING AND THE 16/18-ELECTRON RULE numerous thermally stable Ni(0) phosphine complexes Ni(PR;),. However, whereas Ni(PMes), is highly pyrophoric and very easily oxidized, Ni{P(OPh);}, is stable to air and moisture. Both havea d" configuration, but the strongly donating PMe; ligands push the d-orbital levels of the first to much higher energy, so that the energy gain by oxidation is far greater. © Key points Ligands are classified according to the number of electrons they donate to the metal centre, For electron counting purposes only, itis convenient to consider all ligands as neutral and the ‘metal centre as zero-valent. Mononuclear complexes tend towards an electron count of 18. Steric hindrance may reduce this. Complexes with strong n-acceptor ligands (CO, NO) usually obey the 18-electron rule. Metals with d* configuration form square-planar 16-electron complexes. For the assignment of the metal oxidation state, ligands that exist as free molecules are anions are considered negative: Cr, O-, ete. The considered neutral, and ligands that form t oxidation state is then equal to the balancing charge. @ Exercises | 1. Give examples of © ametallocene complex: Gi. ahalf-sandwich complex; ii) anarene complex. 2, Assign the number of electrons donated to metal centres by the following ligands: (0 hydrides Gi) bridging chloride; ii) CO; (iv) a phosphine PRs: ©) ethylene; () 1,3-butadiene; (vil) cycloheptatriene; (iid. cyclopentadienyt 3. What s the electron count ofthe metal centre in the following complexes: © Cpx60; CpaCor's Gi) CpaNis Gv) Cppmin(CO)35 ( Fe(Coy(a,3-butadiene)ss (w) CoWN0y(CO),: (uit) CpNi(ns-allyds (i) Ni(CH20°Ph). 4, What are the formal metal oxidation states in the complexes in question 32 95