Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies concerning the chemistry of rare earth elements in organic environments Crease, Allan Edward 1973

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1974_A1 C74.pdf [ 6.8MB ]
Metadata
JSON: 831-1.0099929.json
JSON-LD: 831-1.0099929-ld.json
RDF/XML (Pretty): 831-1.0099929-rdf.xml
RDF/JSON: 831-1.0099929-rdf.json
Turtle: 831-1.0099929-turtle.txt
N-Triples: 831-1.0099929-rdf-ntriples.txt
Original Record: 831-1.0099929-source.json
Full Text
831-1.0099929-fulltext.txt
Citation
831-1.0099929.ris

Full Text

STUDIES CONCERNING THE CHEMISTRY OF RARE EARTH ELEMENTS IN ORGANIC ENVIRONMENTS by ALLAN EDWARD CREASE B . S c , Univers i ty of Sussex, 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n the Department of Chemistry We accept this thesis a.s. conforming ^to the required standard THE UNIVERSITY OF BRITISH COLUMBIA NOVEMBER, 1973 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f the r e q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by the Head o f my Department o r by h i s r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f C A l g K l S T f c Y The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8, Canada Date S O N o v f i ^ b Q_T f °> ^ ABSTRACT i i The infrared and proton magnetic resonance spectra of a number of compounds containing s i tes of Lewis b a s i c i t y change in the presence of selected cyclopentadienyllanthanides, R 3 Ln. These spectral changes indicate that the lanthanide der ivat ives can act as Lewis acids towards bases such as bridging and terminal carbonyl l igands , terminal n i t r o s y l l igands , appropriate t r a n s i t i o n metals and a carbon-carbon t r i p l e bond. Several s o l i d adducts can be isolated and characterized. The preparation of the new compounds R 2 LnM(_ 5 -Cp)(C0) 3 [Ln = Dy, Ho, Er or Yb; M = Mo or W] i s described, and evidence for the existence of isocarbonyl linkages i n these complexes i s presented. As a synthetic route to compounds containing apparent lanthanide-t r a n s i t i o n element bonds, the d i r e c t reaction of elemental metals with t r a n s i t i o n metal organometal1ics i s s tudied. Finely divided lanthanide and other metals react with substrates such as Mn(C0) 5Br, ( _ 3 - C 3 H 5 ) F e ( C 0 ) 3 I , [ (_ 5 -Cp)Mo(C0) 3 ] 2 and (_5-Cp)Cr(C0)3HgCl i n THF to y i e l d extremely reac t ive , a i r - and moisture-sensit ive so lut ions . Some of the chemistry of these solutions and other synthetic approaches to them are o u t l i n e d . The v e r s a t i l i t y of these solutions as reactive intermediates during the preparation of organometallic compounds i s described. TABLE OF CONTENTS ABSTRACT TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES ABBREVIATIONS AND COMMON NAMES ACKNOWLEDGEMENTS CHAPTER I , GENERAL INTRODUCTION CHAPTER I I , LEWIS ACIDITY OF CYCLOPENTADIENYLANTHANIDES 2.1 INTRODUCTION 2.2 EXPERIMENTAL 2.3 RESULTS AND DISCUSSION CHAPTER I I I , REACTIONS INVOLVING ELEMENTAL METALS 3.1 INTRODUCTION 3.2 EXPERIMENTAL 3.2a REACTIONS INVOLVING MANGANESE CARBONYLS 3.2b REACTIONS INVOLVING ALLYLTRICARBONYLIRON IODIDE . . . 3.2c REACTIONS USING CYCLOPENTADIENYLTRICARBONYL-MOLYBDENUM DIMER AND YTTERBIUM METAL 3.2d REACTIONS INVOLVING MERCURY COMPOUNDS AND YTTERBIUM METAL 3.3 RESULTS 3.3a REACTIONS INVOLVING MANGANESE CARBONYLS 3.3b THE MAUVE SOLUTION FROM ALLYLTRICARBONYLIRON IODIDE AND A METAL 3.3c CYCLOPENTADIENYLMOLYBDENUM DIMER REACTIONS 3.3d REACTIONS WITH MERCURY COMPOUNDS iv Table of Contents (cont'd) Page 3.4 DISCUSSION 102 3.4a REACTIONS INVOLVING MANGANESE CARBONYLS 102 3.4b REACTIONS INVOLVING ALLYLTRICARBONYLIRON IODIDE 108 3.4c REACTIONS INVOLVING CYCLOPENTADIENYLTRICARBONYL-MOLYBDENUM DIMER 1 1 J 3.4d REACTIONS INVOLVING MERCURY COMPOUNDS 1 1 5 CHAPTER IV, CONCLUDING REMARKS 1 1 7 REFERENCES . . . 1 1 9 APPENDIX THE- DETERMINATION OF LANTHANIDES IN ORGANOMETALLIC COMPLEXES BY THE CLOSED OXYGEN FLASK METHOD 125 LIST OF TABLES Table Page I Infrared spectra i n the carbonyl and n i t r o s y l s tretching region 3 3 II Infrared spectra of adducts formed from [ (_ 5 -Cp)Fe(C0) 2 ] 2 and some Lewis acids 44 III Changes induced by (MeCp)3Nd i n the proton magnetic resonance spectra of various Lewis bases 46 IV Conductivity measurements i n THF 51 V Infrared spectra of R 2 Ln[(_ 5 -Cp)M(C0) 3 ] species 5 2 VI Conductivity resul ts for manganese carbonyl reactions 9 ^ VII Conductivity measurements of s ta r t ing reagents and solvents 91 v i LIST OF FIGURES Figure Page 3-1 D i l u t i o n of the Red Solution 89 3-2 Formation of the Mauve Solution 95 3-3 D i l u t i o n of the Mauve Solution 96 3-4 Mossbauer Spectrum of green s o l i d from mauve solut ion and t r i phenyl phosphine .' 99 v i i ABBREVIATIONS AND COMMON NAMES The fo l lowing l i s t of abbreviations and common names, most of which are commonly adopted in chemical research l i t e r a t u r e , w i l l be employed i n th is t h e s i s . Cp : C5H5 cyclopentadienyl MeCp : C 6 H 7 methylcyclopentadi enyl COT : cyclooctatetraene Me : CH 3 methyl Et : C 2 H 5 ethyl i -Bu : (CH 3) 2CHCH 2 isobutyl Ph : C 6 H 5 phenyl E t 20 : ( C 2 H 5 ) 2 0 diethyl ether THF : C^HgO tetrahydrofuran py : C5H5N pyridine bipy : C 1 0 H 8 N 2 2,2 ' - b i p y r i d y l phen : C 1 2 H 8 N 2 1,10-phenanthroline Ln : a lanthanide element R 3A1 : a t r ia lkyla luminium Ph 3P ( C 6 H 5 ) 3 P t r i phenylphosphi ne TMS : (CHgKSI tetramethylsi lane CNDO/2 : Complete Neglect of D i f f e r e n t i a l Overlap - Method Two. SCF : Se l f Consistent F i e l d n .m.r . : nuclear magnetic resonance e . s . r . : e lectron spin resonance ppm : parts per m i l l i o n J : coupling constant h : hour(s) v i i i Abbreviations and Common Names (Cont'd) Hz : Herz v : s t retching frequency cm" 1 : wave numbers, rec iprocal centimetres M : molar c a l c . : calculated Q.S. : quadrupole s p l i t t i n g I .S . : isomer s h i f t i x ACKNOWLEDGEMENTS I would l i k e to thank the various members of the technical and academic s t a f f of the Chemistry Department for the helpful suggestions they have made. In p a r t i c u l a r , I am grateful to Mr. P. Borda for making a s i g n i f i c a n t contr ibut ion which enabled me to invest igate more thoroughly one aspect of my work. I would l i k e to thank El i n Sigurdson and John Mali to for technical assistance and thoughtful advice. May I also express my gratitude to Ms. J . Houlden, who typed the manuscript. In a d d i t i o n , I would l i k e to thank Dr. F. Aubke and Dr. W.R. Cullen who read t h i s thesis and suggested several improvements. The receipt of a National Research Council of Canada predoctoral scholarship i s g r a t e f u l l y acknowledged. F i n a l l y , may I record my thanks and indebtedness to Dr. P. Legzdins, whose understanding, good humour and encouragement were invaluable during the l a s t four years . CHAPTER I GENERAL INTRODUCTION The work presented i n t h i s thesis i s p r i m a r i l y concerned with the chemistry of the lanthanides or rare earths, that i s elements 57-71. Scandium and yt tr ium are often included for discussion with these elements, but we shal l only be concerned with y t t r i u m . The coordination chemistry of the 1 2 lanthanides has been an area of intense study, * whereas only a l imi ted amount of work has been reported i n the f i e l d of organolanthanides. Indeed, at the time t h i s project began the only organic der ivat ives known were those containing cyclopentadienyl 3*77,103 HJndeny 1 4 or phenyl 5 groups, plus some i l l -6 5 defined cyclooctatetraene and methyl compounds. Since that time a new series 7 8 9 of cyclooctatetraene compounds, ' some Grignard-type ytterbium species, a propynide of europium,^ and some cerium(IV) d e r i v a t i v e s , ^ whose reported synthesis and properties must be ser iously questioned, have been prepared and reported. Very recent ly , some carbonyls of ytterbium and neodymium have been 12 detected, by infrared spectroscopy, in argon matrices at 10°K. Generally speaking, organolanthanides are a i r - and moisture-sens i t ive s o l i d s ; they are readi ly attacked by halogen-containing or a c i d i c solvents and are only soluble i n strongly donor solvents which often solvate 3 the product. In f a c t , the f i r s t reported organolanthanides, Cp 3 Ln, were almost 13 c e r t a i n l y prepared unwitt ingly as ammonia solvates . The species which can be most readi ly prepared and manipulated are those containing ligands which can s t a b i l i z e a negative charge (e .g . Cp" and COT 2 ") . The attempts to prepare phenyl and methyl d e r i v a t i v e s 5 have, with one notable exception, namely ( ( L i ( C t t H 8 0 ) i + ) ( L u ( C 8 H 9 ) l + ) , resulted i n polymeric mater ia ls . This behaviour i s 15 16 i n contrast with the recently prepared a l k y l t r a n s i t i o n metal d e r i v a t i v e s , 5 which are monomeric so l ids displaying good s o l u b i l i t y in many organic solvents . 2 A var ie ty of physical techniques have been employed to determine the nature of the bonding i n the compounds, Cp 3 Ln. The general consensus of opinion J 7 > 1 8 » 1 9 based upon spectral and chemical propert ies , i s that these compounds are almost e n t i r e l y i o n i c a l l y bonded with l i t t l e or no 4f o r b i t a l involvement i n any covalent in terac t ions . Bearing i n mind the character of organolanthanides, we made some preliminary attempts to prepare new der ivat ives based upon the l igands : f l u o r e n y l , phenyl acetylenyl and tr iphenyl methyl. Although various solvents were used, inc luding l i q u i d ammonia, i n which ytterbium metal i s so lub le , we were not able to reproducibly i s o l a t e any new compounds, i n spite of observing d i s t i n c t colour changes i n the reaction mixture and obtaining new infrared spectra of the impure products. The major d i f f i c u l t y , aside from extensive non-stoichiometric s o l v a t i o n , was the separation of the products from the s t a r t i n g mater ia l s , say Grignard reagents, because of s i m i l a r s o l u b i l i t y and r e a c t i v i t y propert ies . In view of these d i f f i c u l t i e s , i t was decided to use the known properties of i s o l a b l e organolanthanides i n our s tudies . The greatest potential f o r organolanthanides appeared to be i n the u t i l i z a t i o n of 20 t h e i r a b i l i t y to act as Lewis acids and so t h i s property was examined fur ther . S p e c i f i c a l l y , invest igat ions were carr ied out using some unusual base s i t e s , i n p a r t i c u l a r the oxygen atom of a t r a n s i t i o n metal carbonyl or n i t r o s y l . The resul ts of these studies led us into the f i e l d of metal-metal bonds and hence elemental metal reactions were undertaken as a possible route to compounds displaying th i s kind of bond. The whole approach to the work has been essen-t i a l l y of a survey nature, namely to determine what can be done with the rare earth elements i n an organic environment. 3 CHAPTER II LEWIS ACIDITY OF CYCLOPENTADIENYLLANTHANIDES 2.1 INTRODUCTION The Lewis electron pair bonding diagram for carbon monoxide reveals both ends of the molecule to possess a lone pair of e lectrons. Carbon monoxide normally bonds to t r a n s i t i o n metals by attachment of i t s carbon atom, leaving the oxygen atom coordinat ively unsaturated. There are various j u s t i f i c a t i o n s for th is o r i e n t a t i o n , d i f f e r i n g only i n t h e i r level of s o p h i s t i c a t i o n . For 21 example, Ballhausen and Gray point out that "the two electrons with the highest energy are mainly l o c a l i z e d on the carbon nucleus" and that th i s o r b i t a l i s of sui table symmetry and energy for overilap with t r a n s i t i o n elements r e s u l t i n g i n l i n e a r M-CeO arrangement. Another explanation i s that the carbon end of CO i s the softer Lewis base par t , whereas the oxygen terminal i s much harder. In accord with the general c r i t e r i o n for s o f t - s o f t and hard-22 hard interact ions for acids with bases, i t i s carbon which attaches to the t r a n s i t i o n elements. The l a t t e r are regarded as sof t i f t h e i r oxidation 2? state i s low or zero. An analogous argument can be made for n i t r i c oxide bound to t r a n s i t i o n metals. Hence, for t r a n s i t i o n element carbonyls or n i t r o s y l s and t h e i r substituted d e r i v a t i v e s , the oxygen atom of the carbonyl or n i t r o s y l group may be viewed as a potential base s i t e . Formally i soe lec t ronic with CO and N0 + i s the CN" group which i s known to be bidentate towards Lewis acids i n various environments, (e .g. CH3CNBF3 and K t f[Fe(CNB.F 3) 6] ) . In a l l cases, upon complex formation, the s tretching frequency of the CN group i s observed to increase. Simple inductive arguments would predict a lowering of t h i s frequency as a resu l t of electron withdrawal from the C-N bond towards the a c i d . Various authors have offered 25 explanations of t h i s phenomenon; Purcel l i n a series of papers has made the 4 most deta i led study and his f indings are summarised here. He was able to show that kinematic coupling of the C-N and N-Lewis acid vibrat ions could not account for a l l of the increase i n v ^ . Using an extended Hlickel theory method he determined that increased strengthening of the C-N l i n k upon coordination was pr imar i ly caused by N2s overlap with the C2s and C2po o r b i t a l s ; that i s the c system was responsible for extra s t a b i l i s a t i o n of the CN group. Whilst comparing a and IT bonding effects in the coordination of CO and CN" using a SCF ca lcu la t ion according to the CNDO/2 scheme, he investigated HC0+ and C0H+. He found that the lone pair molecular o r b i t a l s of carbon and oxygen taken together respond s i m i l a r l y to protonation at e i ther end with an increase i n CO bond order. However, C0H+ d i f f e r s from HC0+ i n the d i f f e r e n t response of the 3a o r b i t a l (mainly 02s in character) which i s v i r t u a l l y unperturbed for carbon a d d i t i o n , but polarised i n the case of oxygen addit ion to decrease the CO bond order. In terms of Valence Bond theory one may say that (I) makes an enhanced cont r ibut ion . :.C = Ot • :C = 0: (I) ( I D to C0H+, w h i l s t (II) i s preferred by HC0+. This increased p o l a r i z a t i o n of electrons by oxygen when coordination occurs at that atom led Purcel l to speculate: " . . . the CO a bond order of C0H+ i s also appreciably reduced and leads us to suspect that , should attempts to prepare Lewis acid adducts of metal carbonyl complexes i n which the acid coordinates to the oxygen of carbon coordinated CO, be success fu l , the infrared spectra of the adduct w i l l be characterised by a pronounced lowering, r e l a t i v e to the carbonyl complex, of the CO stretching frequency. Furthermore, enchanced back bonding from the metal to CO would compliment p o l a r i z a t i o n of the CO TT dens i ty , par t i cu lary i n 5 view of the ant ic ipated increase in CO distance. The decrease of CO frequency should, i n any event, amount to at least a few hundred wave numbers". Recently, i t has been shown that t h i s predict ion can be rea l i sed i f s u f f i c i e n t l y hard Lewis acids are employed. These adducts, i n which both ends of the CO or NO molecule are bonded, are said to contain an " isocarbonyl" or " i s o n i t r o s y l " l inkage respect ive ly . The evidence for these types of interact ions has r e l i e d heavily upon infrared data, although a few crys ta l structures have been determined and one u l t r a - v i o l e t study was made. Complex formation does resul t i n a lowering of the infrared stretching frequency of the C-0 (or N-0) group. Not a l l the V C Q (or V ^ Q ) are lowered, however, i f there are other non-complexing carbonyls (or n i t r o s y l s ) in the donor molecule. The absorptions of these ligands are observed to r i s e s l i g h t l y i f they change at a l l . The rat ionale i s that as electrons are removed towards the Lewis a c i d , the electron density on the t r a n s i t i o n metal i s lowered,result ing in less IT back donation into the uncomplexed carbonyls 1 (or n i t r o s y l s ' ) TT* o r b i t a l s and so causing a higher s tretching frequency to be observed for th i s l i g a n d . I t i s readi ly seen that t h i s addit ional e f fect i s secondary i n terms of s ize of A V C Q / N 0 ' because °^ the distance from the s i t e of adduct formation. In early 1971, when our studies began, the l i t e r a t u r e contained only four short c o m m u n i c a t i o n s 2 6 ' 4 4 ' 4 6 ' 4 7 dealing with isocarbonyl behaviour. There are now twenty papers published, over half of these i n the l a s t twelve months, concerning t h i s mode of bonding. The l i t e r a t u r e review below summarises a l l work i n the f i e l d , other than our own, up to the end of September 1973. The f i r s t reported example of an isocarbonyl l inkage , which was recognized as such, was [ ( j T 5 - C p ) F e ( C 0 ) 2 ] - 2 A l E t 3 2 6 . [(h. 5-Cp)Fe(CC^] 2 i s known to e x i s t i n the s o l i d phase with two bridging and two terminal carbonyls, 27 28 whether a c i s or trans arrangement of (h_5-Cp) ligands i s the case. The E t 3 A l molecules were shown by an X-ray crysta l lographic study to be coordin-ated to the oxygen of the bridging carbonyls i n the c i s isomer of C(h 5 -Cp)Fe(C0) 2 ] 2 . Et 3 h5_CP>v C h 5 -Cp 0 E t o A l ^ ° 29 The A l - 0 distance was measured to be 1.98(2)A, which i s close to the value o of 2.02(2)A observed for the donor-acceptor bond i n the bis- ( tr imethylaluminium)-30 ° dioxane adduct, u but i s s i g n i f i c a n t l y longer than the A l - 0 distance of 1.89A in A 1 ( H 2 0 ) 6 3 + 3 1 and 1.82A i n A 1 C 1 3 - C 6 H 5 C 0 C l . 3 2 In the l a s t two cases, however, the ligands attached to aluminium are far more electronegative than ethyl groups and would enhance the hard character of th is metal , thereby helping to shorten the A l - 0 bond length. The Al-O-C bond a n g l e 2 9 i s 155°, which i s compatible with the locat ion of two lone pairs on the carbonyl oxygen, i f 29 one pair i s involved i n bonding. The A l - C bond distances are s i m i l a r to values found i n the (Me3Al ) 2 - C i t H 8 0 2 3 0 adduct, w h i l s t the [(h_5-Cp)Fe(C0) 2] 2 part of the complex i s almost unchanged r e l a t i v e to the free molecule, except o for a s l i g h t l y shorter (0.04A) Fe-Fe distance. In conjunction with th i s crysta l data , infrared evidence substantiates the same mode of bonding i n s o l u t i o n . In heptane the infrared spectrum shows a 112 c m " 1 , lowering for the bridging carbonyl and a 40 cm" 1 r a i s i n g of the terminal carbonyl abs rption, 33 r e l a t i v e to the parent i ron compound. Further support for the formulation of t h i s compound as an adduct i s the reaction of [(h_ 5-Cp)Fe(C0) 2] 2 '2AlEt 3 with a 7 s l i g h t molar excess of tr iethylamine causing complete regeneration of [(_ 5 -Cp)Fe(C0) 2_2' In a l a t e r p a p e r 3 4 Shriver studied the solut ion infrared spectrum of t h i s system i n more depth. By adding the Lewis acid incrementally and measuring the associated changes i n pos i t ion and i n t e n s i t y of the carbonyl absorptions, he was able to ascertain the f o l l o w i n g : (a) addit ion is stepwise v i a a 1:1, then to the 1:2 adduct and no f u r t h e r ; (b) the 1:2 adduct i s probably symmetrically arranged, because the symmetric stretch of the bridging carbonyl remains weak as observed i n the parent i ron compound; (c) although both _i_s and trans forms of [ (_ 5 -Cp)Fe (C0) 2]2 are present, the proportion of the c i s form increases from parent to 1:1 to 1:2 adduct. In.order to determine point ( c ) , Shriver made use of the fact that the r e l a t i v e i n t e n s i t i e s of the v„,_ and v, of the terminal CO stretch sym asym 27 r e f l e c t the c i s - t r a n s r a t i o i n the s t a r t i n g i ron dimer, and that t h i s donor apparently i s hardly changed s t r u c t u r a l l y in : the adduct. The only comment one might make about the infrared data i s the in terpreta t ion of the 2v C Q bridging bands for the 1:1 adduct. Shriver remarked that t h i s species lacks a centre of symmetry and thus should "d isplay two prominent CO stretching absorptions" . Whilst not questioning the number of bands or that a 1:1 adduct has formed, i t would seem reasonable to expect 2v C Q bands on the grounds that one carbonyl group i s complexed whi l s t the other i s not. Closer scrut iny of the two br iding bands shows one raised and one substan-t i a l l y lowered^ r e l a t i v e to the parent [ (_ 5 -Cp)Fe (C0) 2 ] 2 , which i s to be expected i f only 1 group were forming an isocarbonyl l i n k . Further evidence 3 4 for both ci_s and trans isomers in solut ion was obtained from n.m.r . experiments which showed two resonances for the C 5 H 5 pro-^ In fact th is absorption i s lowered to the exact pos i t ion of the 1:2 adduct, \ although i t i s weaker in i n t e n s i t y of course. 8 tons: a large one at 2.81 ppm and a weaker one at 2.72 ppm u p f i e l d from benzene. The former peak was assigned to the c _ isomer and the l a t t e r to the t rans ; s i m i l a r resul ts have been observed for the uncomplexed [ (_ 5 -Cp)Fe(C0) 2 _2 molecule. a ' Shriver was also able to show from the n.m.r . data that c i s to trans interconversion was slower in the complexed system. This r e s u l t i s to be expected because the interconversion i s thought to occur v ia a non-bridged i n t e r m e d i a t e , 3 ^ 3 ' 3 ^ which would be less favourable for isocarbonyl formation. [(_5-Cp)Ru(C0).2;].2 i s a species which i s known to ex i s t as both 33 38 bridged and non-bridged isomers in so lut ion in almost equal amounts. ' Using a s i m i l a r technique to that used for the iron dimer, S h r i v e r 3 4 showed that the equi l ibr ium could be sh i f ted e n t i r e l y to the bridged side by addit ion of aluminium a l k y l s i n heptane. The infrared data were once again r e a d i l y interpreted as showing a 1:1 and then a. 1:2 adduct forming. Just as i n the i ron case c i s and trans isomers were present with the c i s favoured i n the ad-duct , which could be dissociated by addit ion of a stronger base, such as t r ie thylamine . This example shows quite c l e a r l y the more basic nature of the bridging carbonyl group and that t h i s property can be u t i l i s e d to force a s t ructura l rearrangement. Unlike the previous two examples, [(_ 5 -Cp)Mo(C0) 3 _ 2 has a centro-symmetric molecular structure with a metal-metal bond and s ix terminal carbonyl 3 9 l igands i n the s o l i d s ta te ; moreover, th i s configurat ion i s believed to pers i s t i n s o l u t i o n . 4 0 ' 4 1 Recently, C o t t o n 3 7 used C 1 3 n .m.r . while inves t iga-t ing the p o s s i b i l i t y of carbonyl transfer i n the complex v ia a bridging i n t e r -mediate. His study did not reveal any evidence for such an occurrence. Shriver et a l , 3 4 reacted tr ia lkylaluminiums with [ (_ 5 -Cp)Mo(C0) 3 ] 2 in hydro-carbon solvent and observed new carbonyl bands immediately. The terminal V-Q were sh i f ted to higher frequency and increased i n number (2020w, 1990ms, 9 1979vs, 1944m, 1936m and 1911m c m " 1 ) , w h i l s t two new lower bands at 1774 and 1710 cm" 1 were observed. This species was termed the "equi l ibr ium pro-duct" because a l l the bands increased r e l a t i v e to the parent as the R3A1 concentration was increased and reverted to the parent spectrum upon addit ion of t r ie thylamine. If the mixture was kept at -3°Cffor a few hours, the 1990 and 1710 cm" 1, absorptions increased and bands appeared at 1860 and 1620 c m " 1 . Af ter eight days at room temperature only the 1990, 1860, 1710 and 1620 cm" 1 absorptions remained - t h i s species was termed the " k i n e t i c product". They suggested that the equi l ibr ium product i s consistent with a structure involv ing two R3A1 coordinated carbonyls and four terminal carbonyls. They o went on to remark that the Mo-Mo bond (3.22A in the parent) i s too long to o accommodate a carbonyl bridge because i t i s thought that 2.8A i s the l i m i t i n g i n t e r m e t a l l i c separation to permit such an occurrence. However, they men-o tioned that Mo-Mo bonds are known i n the range 2.89 - 3.09A and i t i s thus "conceivable" that bridging carbonyl formation could occur with a concomitant shortening of the Mo-Mo bond. I t should be remembered at t h i s point that the shortening of the Fe-Fe bond i n [(_ 5-Cp)Fe(C0.) 2] 2 u P o n complexation was only o 0.04A, whereas the Mo case would require a substantial reduction i n the Mo-Mo bond, by an ef fect which i s not considered to be severe. A l l of the examples c i ted so f a r have involved a carbonyl group bridging two metal atoms. If one regards the state of hybr idizat ion of oxygen i n the carbonyl group to be sp, sp 2 and s p 3 for l i n e a r , doubly and t r i p l y bridging s i tuat ions respec t ive ly , then a gradual increase i n b a s i c i t y i s to be expected as the percentage s character of the lone pair o r b i t a l diminishes. An experiment to demonstrate the increasing b a s i c i t y for t r i p l y as opposed to doubly bridging CO was devised by S h r i v e r . 3 4 He compared the two molecules [ (_ 5 -Gp)NiC0] 2 and (h_ 5-Cp) 3Ni 3(C0) < 2 as potential donors. The former has only doubly bridging carbonyl groups which are not coplanar with the two nickel atoms, 10 43 hence a formally bent metal-metal bond exis ts in s o l u t i o n . The in terac t ion of t h i s molecule with e i ther tr iethylaluminium or t r i i sobutylaluminium was not strong: only a 1:1 adduct was formed, the s h i f t s of V^Q were not as great as the iron and ruthenium cases considered e a r l i e r and a much higher concen-t r a t i o n of aluminium a l k y l was necessary to bring about adduct formation. On the other hand, (h_ 5 -Cp) 3 Ni 3 (C0) 2 , which contains only t r i p l y bridging carbonyls, exhibited both 1:1 and 1:2 complex formation; the 1:1 adduct formed at one quarter the concentration of alkylaluminium required for the corresponding complex of [(h_ 5 -Cp)NiC0] 2 . In the compound [(h^-CpjFeCO]^ a l l CO ligands are t r i p l y br idging . 34 It was not possible to fo l low successive addit ion of aluminium a l k y l s ; however, a 1:4 adduct was i so la ted as a s o l i d which showed only one V C Q , which was 150 cm" 1 lower than the s ingle absorption of the parent. Kotz and Turnipseed 4 4 reported the f i r s t example of a terminal metal carbonyl involved i n isocarbonyl bonding. They found that Ph 3 PC 5 H 4 Mo(C0) 3 (henceforth ca l l ed cpyl idMo(C0) 3 ) , which i s not very soluble i n toluene, would r e a d i l y dissolve i f trimethylaluminium was added to the s o l u t i o n . A dark brown s o l i d could be i solated from the s o l u t i o n . Tensimetric t i t r a t i o n s and reactions with ethanol indicated a 1:1 adduct had formed. That the base s i t e was an oxygen atom was suggested by the infrared spectrum which showed a low band at 1665 cm' 1 assigned to the isocarbonyl group, w h i l s t the two non-cbmplexed CO groups showed s l i g h t l y higher absorptions than i n c.pylidMo(C0) 3 . They postulated that the increased s o l u b i l i t y of the adduct could be described by the e q u i l i b r i u m : 2cpylidMo(C0) 3 + M e 6 A l 2 2cpylidMo(CO) 3 -AlMe 3 Such an equi l ibr ium would explain why rapid exchange of bridging and terminal methyl groups of the M e 6 A l 2 was found to occur (by n .m.r . ) i n toluene even at 11 -80°C, which i s a temperature well below the coalescence of methyl resonances for M e 6 A l 2 alone. The v a r i a b i l i t y of the s i t e of b a s i c i t y revealed i t s e l f i n the in terac t ion of the anion [(h_5-Cp)W(C0)3]~ with the two Lewis acids Ph 3 In 46 and P h 3 A l . In the case of the indium compound, the symmetry-predicted three terminal CO absorptions, plus two addit ional weak bands from the uncomplexed anion, were observed i n the region 1750 - 1950 cm" 1 . These observations are consistent with the formation of an In-W bond. The Ph 3 Al species , however, showed only two terminal v C Q bands and a t h i r d strong band at 1600 cm" 1 i n d i c a t i v e of isocarbonyl behaviour. To determine whether the [(h_5-Cp)W(C0)3]~ anion was s t i l l i n t a c t , a s l i g h t excess of pyridine was added to the 1:1 complex i n hot toluene and a 95% recovery was ef fected. Presumably the harder aluminium atom i s capable of forming a stronger A l - 0 bond than Al-W bond, whereas the weaker Lewis a c i d , P h 3 l n , prefers the softer base s i t e of the t r a n s i t i o n element. Organoaluminium compounds need not be the Lewis acid for isocarbonyl l i n k s to form. An experiment to demonstrate t h i s point was conducted by Brown and B r o w n . 4 7 They rea l ized that the anion [Fe(CN) 5 X] n ~ in the presence of transi t ion-metal ions i n the s o l i d state adopts the ni troprusside structure i n which a l l cyanides are bridging and the unique l i g a n d , X, would be forced into a bridging s i t u a t i o n . They prepared Co 3 [Fe(CN) 5 C0] 2 -5-7H 2 0 and showed o i t to have a face centered cubic l a t t i c e with a = 10.27 A, which i s consistent with a Prussian blue s tructure . In a d d i t i o n , they compared the infrared spectrum of the s ta r t ing sodium s a l t to that of the cobalt complex: the sodium species which has no bridging CN groups exhibited f i v e bands i n the range 2015 - 2095 cm' 1 [2015m, 2040s, 2055s, 2075s, s h , and 2095m, sh , c m " 1 ] , whereas the cobalt compound has one absorption at 1950 cm" 1 and four others 12 in the region 2025 - 2185 cm"1.. [2025m, 2090s, 2120s, sh , and 2185w cm"1]". 48 I t had already been established that CN groups exhib i t an increase of t h e i r s tretching frequency when strong bridging occurs, hence the 1950 cm" 1 band was assigned to isocarbonyl behaviour and the remaining absorptions were considered to be bridging cyanides. Further, comparison of the infrared spectra of the sodium and cobalt compounds permitted assignment of the 2040 cm" 1 absorption to CO i n the sodium s a l t . I t should be noted that 49 - , Cotton had at t r ibuted a band -at. 2052 cm 1 to CO i n the very s i m i l a r compounds K 3Fe(CN) 5C0 which displayed almost ident i ca l absorptions. A new class of compounds in which a l l the anionic ligands are bound to the central metal by isocarbonyl bonds has been prepared by 50 "'51 ^53 50 Burl i t c h e t . al_. ' '* The f i r s t compound in the s e r i e s , Al[(J_ 5-Cp)W(C0) 3] 3-3THF was synthesized by a metal exchange reaction of aluminium with Hg[(__ 5-Cp)W(C0) 3] 2 . A s ingle x-ray crysta l study of the product showed aluminium octahedrally coordinated to three THF molecules and three isocarbonyl bonds, one from each (J_5-Cp)W(C0)3 ent i ty , in a fac o conformation. The W-C distances (1.85(2)A) of the carbonyls which l i n k tungsten to aluminium are systematical ly shorter than those of the terminal o carbonyls (1.95(2)A). Conversely, the corresponding C-O distances are o o 1.25(2)A and 1.16(2)A, respect ive ly . These bond length changes are to be expected as electrons are withdrawn towards aluminium and away from tungsten. o The A l - 0 distances to three carbonyls (1.827(9)A) are subs tant ia l ly shorter o than to the THF ligands (1.94(2)A) and also much shorter than the previously discussed [ ( h 5 - C p ) F e ( C 0 ) 2 ] 2 - 2 A l E t 3 case (1.98(2)A). A l l the W-C-0 units are e s s e n t i a l l y l i n e a r (176(3)?.) and a l l 0-A1-0 angles are orthogonal to wi th in 3 ° . The A1-0-C angles vary from 140.4° to 162.9° and average 151°; Burl i t c h a t t r ibuted t h i s spread to packing forces . The shorter and presumably 13 stronger A l - 0 bond to the carbonyl groups as opposed to the THF molecules i s re f lec ted i n the fac t that the infrared spectra i n the s o l i d state (Nujol) and i n THF are i d e n t i c a l , thereby implying no d issoc ia t ion of the [(h_5-Cp)W(C0)3]~ anion from the complex i n THF. In f a c t , only very strongly coordinating solvents such as dimethyformamide cause d issoc ia t ion as evidenced by infrared spectroscopy. Chemically the compound behaves as though i t i s very i o n i c : i t r e a d i l y undergoes displacement by nucleophiles and i s e a s i l y converted to HW(h_5-Cp) (C0) 3 or CF 3 C0W(h 5 -Cp)(C0) 3 . The infrared spectrum i n Nujol i s quite complicated, d isplaying four strong bands in the terminal v c Q region and four others in the range 1570 - 1670 cm"1 a t t r ibuted to isocarbonyl in terac t ions . 51 In a very recent paper, th i s work has been extended to include der ivat ives of magnesium of the general formula m2Mgpy i+, [m = CotCOK, (Jl 5-Cp)Mo(C0) 3 , Mn(C0) 5 , and (h_5-Cp)Fe(C0)2.3 The molybdenum and cobalt der ivat ives were prepared by reacting magnesium amalgam with the appropriate m2Hg i n a so lut ion of pyridine and toluene, w h i l s t the manganese and iron compounds were synthesized using the amalgam d i r e c t l y upon the metal carbonyl dimer i n the same solvent mixture. (The iron required at least a trace of MgCl 2 to be present to promote the reac t ion) . A l t e r n a t i v e l y the molybdenum compound was obtained by a metal exchange reaction with the corresponding m2Hg i n THF and r e c r y s t a l l i z a t i o n from pyr id ine . In a l l cases the complexes showed infrared absorptions ( in p y r i d i n e , toluene or as a Nujol mull) i n the region 1665 - 1720 cm" 1 i n d i c a t i v e of an isocarbonyl l i n k . The molybdenum compound was the subject of a s ingle c rys ta l x-ray determination which revealed an octahedrally coordinated magnesium with four equatorial pyridine o groups and two ax ia l isocarbonyl bonds. The Mg-0 distance (2.047(2)A) i s 50 s l i g h t l y longer than the analogous A l - 0 length mentioned above and apparently 14 th is i s ref lected i n less perturbation of the adjoining C-0 bond now at o 1.189(3)A, but i t i s s t i l l s i g n i f i c a n t l y larger than the mean terminal C-0 distance of 1.157 A. The Mg-O-C angle i s 1 5 5 . 0 ( 2 ) ° , whi l s t the Mo-C-0 angles whether iso (177 .2(1)° ) or terminal (178 .1 (1 )° ) are both very close to l i n e a r i t y . By analogy with the aluminium compound described above these magnesium derivat ives behaved as strong nucleophiles with the added advantage, however, that the cobalt and manganese species were soluble i n aromatic hydrocarbons. For example, Ph 3 SiCl was converted i n toluene to Ph 3 SiMn(C0) 5 in 34% y i e l d , a reaction which i s inaccessible using. NaMn!(C0)5 in THF. 52 Other workers have prepared almost ident i ca l magnesium der ivat ives using magnesium amalgam and metal carbonyl dimers i n the presence of a base. The compounds they obtained have the general formula BxMgfnl[B=py or THF; m=(h_5-Cp)Fe(C0)2, Mn(C0) 4L, (h 5 -Cp)NiC0, , Co(C0) 3L or Mo(C0)2L ( h 5 - C p ) , where L = CO, or a l k y l - or aryl -phosphine] ; the value of * i s two for a strongly nuc leophi l i c anion and four for a weak nucleophile . On the basis of molar conduct iv i t ies in THF they concluded that the magnesium-transition metal bond was not highly d i s soc ia ted , moreover infrared measurements of the benzene solutions of the iron and cobalt (L = Ph 3P) complexes showed no bands below 1820 cm" 1 . The absence of isocarbonyl bands contrasts s t r i k i n g l y with 50 the resul ts obtained by Burl itch...• Burl i t c h has extended his work on magnesium der ivat ives by employing 53 manganese metal i n exchange reactions with m2Hg to form compounds of the type m2Mnpytf[m = (h_5-Cp)M(C0) 3 , M = C r , Mo, and W]. Infrared measurements showed three strong carbonyl absorptions, one of which was around 1650 cm" 1 and at t r ibuted to an isocarbonyl l inkage. 15 In an attempt to form a covalent bond between aluminium and 54 tungsten, two apparently d i f f e r e n t bonding modes were obtained in the ethyl and methyl species of the react ion . R2A1H + HW(h 5-Cp)(C0) 3 + R2AlW(h.5-Cp) (C0) 3 + H 2 Cryoscopic molecular weight determinations showed both compounds to be dimeric in benzene. In the 'H n . m . r . , s ingle peaks for both the (h_5-Cp) r ing and the methyl groups (for R=Me) which did not broaden or s p l i t even at -65°C removed the p o s s i b i l i t i e s of r ing subst i tut ion and methyl bridges respect ive ly . The infrared spectrum showed considerably d i f f e r e n t absorptions: for R=Me in benzene strong bands at 2014 and 1926 cm" 1 only occurred, whereas the ethyl der ivat ive in methylcyclohexane displayed strong bands at 1986, 1692 and 1659 c m " 1 , typ ica l of isocarbonyl behaviour. The quandary was p a r t i a l l y resolved when an x-ray determination of the methyl compound was published 55 l a t e r . In the s o l i d state th i s compound also showed isocarbonyl behaviour as the compound consisted of a s l i g h t l y puckered twelve membered r ing contain-ing A1-0-C-W bridges. Me2 / A 1 V h 5-C P / \ CO 0 ( / \ / X h 5 - C p A i Me2 The tungsten i s roughly octahedrally coordinated w h i l s t aluminium i s four coordinated with a wide exocycl ic C-Al-C angle, found in other r ing systems (e .g. ( M e 2 A l C l ) 2 ) . 3 5 The A l - 0 distances are 1.83(3)A and 1.79(3)A, both very s i m i l a r to that found by B u r l i t c h i n Al [ (h 5 -Cp)W(C0) 3 ] 3 -3THF 5 0 16 (above). The same trend of longer W-C bonds i n terminal carbonyls compared to the W-C length i n the isocarbonyls i s also observed i n t h i s s t ructure . The molecular structure of the ethyl der iva t ive has not yet been reported; however, i f i t has a s o l i d state geometry s i m i l a r to that implied by i t s so lut ion infrared spectrum, i t could well be i sos t ruc tura l with the methyl analogue i n the s o l i d phase. The question remains as there appear to be no isocarbonyl l inkages , how does the methyl compound ex is t i n so lu-t ion as a dimer? If a W-Al bond ex is ts in s o l u t i o n , as shown i n the f igure below, then why does i t not pers i s t i n the s o l i d and why does not the ethyl complex adopt t h i s bonding scheme? Unfortunately, the so lut ion infrared spectra of the two complexes were not obtained i n the same solvent (above); i f th is dif ference were responsible for the change i n bonding a subtle lever would be provided for a f fec t ing the change of base s i t e s . (h5-Cp)(C0)yW-»-AlMe2 , I I Me 2Al + W(h 5-Cp)(C0) 3 I t has been known for some time that organic carbonyls , such as 56 ketones, in teract through oxygen with Lewis ac ids . I f the Lewis acid i s a lanthanide s h i f t reagent, s p e c i f i c s t ructura l information can often be gleaned fromcchanges i n the n.m.r . spectrum of the organic compound. It was reason-ab le , therefore , to use these paramagnetic s h i f t reagents i n an attempt to induce corresponding ef fects i n the n .m.r . spectrum of metal carbonyl con-ta in ing e n t i t i e s . Providing that the carbonyl oxygen was s u f f i c i e n t l y bas ic , n .m.r . s h i f t s were observed using Eu(fod) 3 (fod = 1 ,1 ,1 ,2 ,2 ,3 ,3-heptaf luoro-7,7-dimethyl-4,6-octanedionate) as the Lewis a c i d . 5 7 For example, (h_5-Cp)Fe(CO)2Me did not appear to i n t e r a c t , whereas the more basic bridging carbonyl of [(h. 5 -Cp)Fe(C0) 2 ] 2 d i d , r e s u l t i n g in a downfield s h i f t of up to 17 0.35 ppm of the (]^5-Cp) proton resonance. Marks et al_. reported one new example of a s h i f t in the proton resonances of a compound containing a terminal metal carbonyl as the base s i t e , (phen) (Ph 3 P) 2 Mo(C0) 2 . This compound has low infrared stretching frequencies (1800s and 1729s cm" 1) f o r the carbonyl groups, a property that appears to be associated with good b a s i c i t y . When the only base s i t e of an organometal!ic compound i s a terminal carbonyl group, „ the strength of the Lewis acid may be the l i m i t i n g factor for adduct formation. Acid strength may be increased, however, by a sui table choice of l igands . These restrict!*onsnmay l i m i t use of th i s n .m.r . technique. I t i s worth mentioning that the p .m.r . spectra of i n -organic compounds are rare ly as complicated as those of organic species , consequently th i s in teres t ing appl ica t ion possesses less potential for inorganic chemists. Nuclear magnetic resonance spectroscopy was also used by Chatt CO EQ and co-workers ' to determine the r e l a t i v e b a s i c i t y of some compounds containing carbonyl and the i soe lec t ronic dinitrogen as terminal l igands . Besides i s o l a t i n g 1:1 adducts with the Lewis acid M e 3 A l , they were able to measure the r e l a t i v e equi l ibr ium constant, K, for the competition react ion : [L-metal-(X=Y)] + Me3Al«Et20 s [L-metal-(X=Y)-AlMe 3] + Et 2 0 where X = C , Y = 0 o r X = Y = N and L = other l igands . The values of K were determined by r e l a t i n g the concentration of the species to t h e i r integrated peak heights in the proton n .m.r . spectrum of the equi l ibrated benzene solutions at 30°C. The values of K so obtained showed that a l l the M-X-Y species except one were more basic than die thyl ether. The exception had the highest in the set , whereas the most basic had the lowest v N M . Indeed, the most bas i c , trans(ReCl (N 2)(PMe 2Ph) 1 +) was the only 18 example for which there was the corresponding carbonyl avai lable for com-parison. Although the stretching frequency of the parent carbonyl is sub-s t a n t i a l l y lower, 1782 cm" 1 r e l a t i v e to the parent dinitrogen at 1923 c m " 1 , the dini trogen i s far more basic having K = 70 as opposed to K = 3.3 for the carbonyl complex. This r e s u l t i s r e a d i l y understood i n terms of the greater e f fec t ive nuclear charge of oxygen than nitrogen towards the lone pairs of e lec t rons , which are ava i lab le for adduct formation. By syste-mat ica l ly changing the nature of L 2 i n trans ReCl(N 2)L^ (L = a substituted phosphine), i t was shown that the most basic l i g a n d , L , caused the greatest lowering i n v N ? . "The impl icat ion of th i s i s that the more electron density the l i g a t i n g dinitrogen takes into i t s ir* o r b i t a l s from the metal d - o r b i t a l s , the more basic i s the terminal nitrogen atom", Chatt remarked. One of the e a r l i e s t reported examples of a Lewis acid complex with a metal carbonyl i s i C o 2 ( C 0 ) 8 - A l B r 3 . 6 0 The molecular structure of dicobal t octacarbonyl i n the s o l i d s t a t e 6 1 i s such that the two cobalt atoms and the two bridging CO groups are not in the same plane, the molecule having only C g symmetry. In s o l u t i o n , however, Co 2 (C0) 8 ex is ts in two isomeric forms which are related by a temperature dependent e q u i l i b r i u m . 6 ^ " 6 4 One isomer, ( I ) , corresponds to the c r y s t a l l i n e substance whereas the other, ( I I ) , has no bridging CO groups, but only a cobalt -cobalt bond, i . e . . 0 (0C) 3 Cq^-^Co(C0) 3 ?=== ( O C ^ C o — M C O ) , , 0 0 (I) (ID Chim" and E r c o l i suggested that the A l B r 3 adduct they formed involving Co 2 (C0) 8 was coordinated via_ a three-centre-two-electron bond to the formally 19 bent cobal t -cobal t bond of isomer ( I ) . Support for th is s t ructural proposal was the observation that the infrared spectrum (obtained with high reso lu-t ion opt ics) of the adduct was v i r t u a l l y ident i ca l with that of the parent carbonyl i n the CO stretching region. This r e s u l t in i t s e l f i s s u r p r i s i n g , since reasoning of the type out l ined previously leads one to expect that the carbonyl stretches of the adduct should occur at a somewhat higher frequency. Indeed, more recent work tends to refute the proposal that a three centre bond i s involved. During a study of the effects of high temp-erature and high pressure upon the reaction between CO and C o 2 ( C 0 ) 8 , Whyman66 found that C o 2 ( C 0 ) 8 - A l B r 3 , prepared according to Chi ni and E r c o l i , 6 0 65 displayed infrared bands i n the region 1800 - 2200 cm l , j u s t as claimed, but also he observed a strong absorption at 1600 cm" 1 . The lowest band was suggested to r e s u l t from possible isocarbonyl behaviour, although i t was not spec i f i ed whether t h i s was of the bridging or terminal type. Schmid and B a t z e l 6 , 7 were invest igat ing the c r i t e r i a for incorpor-ating hetero atoms, (X) , into the c lus ter u n i t , Co 3 (C0) 9 X, when they encoun-tered the C o 2 ( C 0 ) 8 - A l B r 3 adduct. S p e c i f i c a l l y , they had noticed that X could be sulphur, selenium, germanium or more commonly, carbon, but attempts to inser t boron or s i l i c o n gave only compounds of the formula Co 3 (C0) 9 C0SiR 3 or Co 3(C0) 9C0BH 2NR 3 (where R = a l k y l group). As a model f o r t h e i r studies they reacted Group 111b and IVb halides with C o 2 ( C 0 ) 8 , eg. 4CC14 + 9Co 2 (C0) 8 —*- 4Co 3 (C0) 9 CCl + 36C0 + 6CoCl 2 and 7Co 2 (C0) 8 + 4C13BNR3 ~+ 4Co 3(C0) 9C0BCl 2NR 3 + 16C0 + 2CoCl 2 From the reaction of A l B r 3 with Co 2 (C0) 8 they were able to i s o l a t e the adduct C o 2 ( C 0 ) 8 - A l B r 3 . Furthermore, they also noticed that e a r l i e r workers had over-looked a strong absorption at 1595 c m " 1 , which they assigned to a bridging 20 carbonyl bf isomer I acting as an isocarbonyl u n i t . Heating t h i s adduct i n benzene ul t imately led to a compound of the formula C o 3 ( C 0 ) 9 C 0 A l B r 2 - A l B r 3 , with the second A l B r 3 providing a bromine bridging uni t to coordinat ive ly saturate the aluminium atoms i n the structure given: (C0)3Co< \ / \ ^Al< y A l B r 2 o \y Co(C0)3 C'dIC0)3 The react ion- was described to proceed according to the fo l lowing equations: Co 2 (C0) 8 + A l B r 3 — C o 2 ( C 0 ) 8 . A l B r 3 C o 2 ( C 0 ) 8 - A l B r 3 —• 200(00)4 + AdiBr3 2 C o 2 ( C 0 ) 8 - A l B r 3 + 00(00)^ —* CoBr 2 + 4C0 + 2 C o 2 ( C 0 ) 8 ' A l B r 2 C o 2 ( C 0 ) 8 - A l B r 2 + 00(00)4 —* (C0) 9 Co 3 C0AlBr 2 + 200 (C0) 9 Co 3 C0.AlBr 2 + A l B r 3 —> ( C 0 ) 9 C o 3 C o A l B r 2 . A l B r 3 The proposal was made that Lewis acids i n i t i a l l y engage i n isocarbonyl bonding and hence are not incorporated into the cobalt c l u s t e r , but rather they remain oxygen bonded. CO In a l a t e r more deta i led study, i t was shown that the terminal carbonyl stretching frequencies of C o 2 ( C 0 ) 8 A l B r 3 have both increased i n frequency by about 30ccm _ 1 and i n number r e l a t i v e to the parent carbonyl , i n addit ion to the band at 1600 cm" 1 . These resul ts strongly favour the 21 oxygen bonded s tructure . (Attempts to form adducts employing aluminium 34 a l k y l s and Co 2 (C0) 8 were unsuccessful. The cobalt carbonyl appeared to be destroyed by the R3A1 species. The substituted carbonyl (Co(C0) 3 PPh 3 ) 2 showed no isocarbonyl behaviour e i t h e r , 3 4 although th i s compound ex is ts in so lut ion mostly as the non-bridged isomer.) In the same s t u d y , 6 8 Shriver presented evidence f o r 1:1 complex formation between [(h_ 5-Cp)Fe(C0) 2] 2 and BX 3 (X = F, Cl or B r ) . In the case of the BBr 3 adduct, the infrared mull spectra showed the bridging carbonyl bands to be at 1849 and 1437 cm" 1 . S i m i l a r l y both 1:1 and 1:2 adducts between C(jL 5-Cp)FeC0]i f and BX 3 (X = F, Cl or Br) could be obtained. The infrared mull spectra revealed one new strong band i n the 1300 - 1400 cm" 1 region for the 1:1 adduct and two new bands for the 1:2 compound i n the same region. Spectra in the 1650 - 1750 cm" 1 region for CH 2 C1 2 solutions of these adducts showed two bands for both the 1:1 and il,-:2 complexes. These resul ts are i n t e r -est ing i n that the weaker Lewis acids BX3 have caused a greater reduction i n carbonyl s tretching frequency than the aluminium a l k y l s which form 1:4 adducts with [(h^-CpjFeCO]^ . I t i s worth noting that in the cases where i t was possible to incrementally add Lewis acid i t was never found that an intermediate adduct had a lower carbonyl absorption than the f i n a l complex. An explanation of the greater s h i f t by BX 3 may be the more compatible s izes of the acceptor and donor o r b i t a l s of boron (2sp 3) and oxygen (2sp 3) respec t ive ly , as opposed to those for aluminium (3sp 3) and oxygen. Better overlap and a stronger B-0 bond would r e s u l t i n a lower C-0 s tretching frequency. Another consequence might be that the remaining uncomplexed CO ligands become much less basic be-cause of substantial charge withdrawal and so subsequent adduct formation i s reduced; hence BX 3 only forms up to 1:2 adducts with [(h^-CpjFeCO]^. 22 One paper has been published dealing with the e lec t ronic spectra 69 of an isocarbonyl compound. The species chosen for study was Mo(C0) 2(LL) (PPh 3 ) 2 (LL = phenanthroline or 5,6 dimethylphenanthroline), which has an intense charge transfer band centered at 693 nm ((molar a b s o r p t i v i t y : 7.7 x 10 3 mol " 1 c m " 1 ! ) . Upon formation of the adduct with t r ia lkyla luminium the colour changes from blue-green to deep red , resu l t ing from a s h i f t of the charge transfer (CT) band. The spectra of the complexes were described by a simple Hiickel M.O. model which treats the fragment (LL)Mo(C0) 2 . The charge transfer t r a n s i t i o n was assigned to the t r a n s i t i o n from a M.O. of bl symmetry mainly l o c a l i z e d on Mo(C0)2 to a b x symmetry M.O. which i s mostly phenanthrol ine ir* in character. The e f fect of a Lewis acid on the complex was introduced as a perturbation of the coulomb integral of the oxygen to which i t i s attached. A lowering of the Mo(C0)2 donor o r b i t a l was found to resu l t from thris perturbation and thus increased the energy of the CT t r a n s i - * t i o n . Shriver pointed out that no account was made of Lewis acid influence upon changes i n the a donation of CO toward molybdenum, which he regarded as less important than changes in TT bonding. Besides being able to calculate the energy of the band which agreed reasonably well with experiment (518 nm c a l c ; 530 ± 10 nm found), i t was shown that the charge shi f ted fromooxygen to the Lewis acid decreased i n the series E t 3 A l > ( i - B u ) 3 A l » Me3Ga; Aq = 0.085; 0.080 and 0.035 e, r e s p e c t i v e l y ) . The l a t t e r resul t agrees with the known Lewis a c i d i t y of these Group 11 lb a l k y l s , and although the numbers are not meant to be taken l i t e r a l l y they of fer an idea of the magnitude of electron transfer involved. From t h i s l i t e r a t u r e review i t can be seen that the establishment and invest igat ion of isocarbonyl linkages i s very recent. Indeed, almost a l l 23 the work was published w h i l s t our studies were i n progress. This research was undertaken for a var ie ty of reasons. At the s t a r t only aluminium a l k y l s were known to be s u f f i c i e n t l y strong Lewis acids to induce isocarbonyl bonding. E a r l i e r w o r k ^ ' 5 ^ had shown that Cp 3Ln complexes form thermally stable i l . : l adducts with conventional Lewis bases such as Ph 3 P, N H 3 , C 6 H 1 ; 1 NC and THF. Further, lanthanides are known to be hard ac ids . I t therefore seemed feas ib le to determine whether the metals in compounds of the type Cp 3Ln were s u f f i c i e n t l y hard to interact with the oxygen of atmetal carbonyl group. If th i s p o s s i b i l i t y proved to be the case, i t would then be desirable to determine whether the s h i f t of CO stretching frequency could be used as an approximate measure of a c i d i t y within the lanthanide ser ies .^ One would expect an increasing a c i d i t y as the metals contract i n s i z e , that i s with increasing atomic weight. Conversely, by using one Cp 3Ln compound i t might be possible to use the same parameter to compare various metal carbonyl-containing molecules and obtain some guide towards t h e i r b a s i c i t y . I f metal carbonyl anions could be used as potential bases, the question would ar ise whether isocarbonyl behaviour or a lanthanide- t rans i t ion element bond would p r e v a i l . Only one report of the l a t t e r p o s s i b i l i t y e x i s t e d , that was for the < species Er (Co (C0) l t ) 3 -3Ci + H e 0. This unique compound, prepared from Hg(Co(C0)i t) 2 and a 1% w/w Er/Hg mixture, has not been the subject of an x-ray study and * S t r i c t l y speaking, the change i n stretching frequency of the carbonyl or n i t r o s y l band i s not a measure of a c i d i t y , which i s better described by thermodynamic functions which take account of changes within the donor and acceptor molecules themselves (such as bond lengthening). The v a r i a t i o n i n infrared band posit ions i s more a r e f l e c t i o n of the donor-acceptor bond i n t e r a c t i o n . 24 the assumption that i t contains a metal-metal bond rests upon infrared data. Hence the reactions involving metal carbonyl anions were par t i cu lary i n t r i g u i n g . The experiments involving n i t r o s y l s were undertaken i n order to determine the general i ty of the phenomenon and=to compare NO with carbonyl ligands i n t h i s respect. Information of th i s type i s u s e f u l , because i t may lend ins ight into theirmechanism of reactions between metal carbonyls or n i t r o s y l s and strong Lewis ac ids . 25 2.2 EXPERIMENTAL A l l reagents used were of reagent grade or comparable p u r i t y . The fo l lowing chemicals were purchased from commercial suppl ie rs : C 6 H 7 Mn(C0) 3 , [ ( h - 5 - C 5 H 5 ) F e ( C 0) 2]2. (h . 5 -C 5 H 5 )Fe(C0) 2 I , and [(h_5-C 5H 5)NiC0] 2 from A l f a Inorganics and Co 2 (C0) 8 from Pressure Chemical Company. The fo l lowing chemicals were prepared according to l i t e r a t u r e procedures and t h e i r p u r i t y was ascertained by elemental analyses and/or melting point determinations: ( J i J t -C 8 H 8 )Fe(C0 )3, 7 2 ( h 5 - C 5 H 5 ) 2 W H 2 , 7 3 ( h 5 - C 5 H 5 ) C r ( N 0 ) 2 C l , 7 4 F e ( C 0 ) 2 ( N 0 ) 2 , 7 5 ( h 5 - C 5 H 5 ) F e ( C 0 ) 2 C l , 7 6 ( C 5 H 5 ) 3 L n , 1 3 ( C 6 H 7 ) 3 L n , 1 3 ( C 5 H 5 ) 2 L n C l , 7 7 (C G H 7 ) 2 LnCl , 7 7 N a [ ( h 5 - C 5 H 5 ) M ( C 0 ) 3 ] 7 8 (M = C r , Mo or W), [ ( C 6 H 5 ) 3 P ] 2 R u ( N 0 ) C l 3 7 9 and [ ( h 5 - C 5 H 5 ) M n ( C 0 ) N 0 ] 2 . 8 0 [ ( C 6 H 5 ) 2 ( C H 3 ) P ] 2 F e ( N 0 ) 2 was donated by Prof . W. Cul len . [(CH 3 ) 2 NCS 2 ] 2 FeN0 and (h 5 -C 5 H 5 )M(C0) 2 N0 (M = C r , Mo or W) were prepared by our own methods (below). Lanthanide halides were purchased as hexahydrates (99.9% pure) from the Rare Earth D i v i s i o n , American Potash and Chemical Corporation. Dehydration was achieved by the method outl ined below. A l l solvents were p u r i f i e d according to known 81 methods. Solvents were degassed j u s t p r i o r to use; two methods were employed: the freeze-thaw method and a modif icat ion of th is procedure by not freezing before each pumping and by f i n a l l y immersing a sintered glass tube into the solvent through which nitrogen was bubbled for a few minutes. L-grade nitrogen was used throughout a l l experiments and for f i l l i n g the dry box, which was a Vacuum Atmospheres Corporation Dri-Lab model He-43-2. Bench top operations were carr ied out under ni trogen. Infrared spectra were recorded on a Perkin Elmer 457 spectrophoto-meter, using a polystyrene f i l m as a c a l i b r a n t . Proton magnetic resonance spectra were recorded on a Varian Associates T-60 spectrometer with t e t r a -26 methylsilane being employed as an internal standard. Conductivity measure-ments were made using a Yellow Springs Instrument Company (YSI), Ohio, con-d u c t i v i t y bridge model 31 equipped with a YSI 3403 conductivi ty c e l l . The conduct ivi ty bridge was i n i t i a l l y ca l ibrated against known resistances and then against a standard solut ion (a 101.47 g so lut ion of d i s t i l l e d water containing 0.0756 g 'Analar ' KC1) at 25°C. The standard solut ion was used to determine the c e l l constant, K = 1.015. Elemental analyses were carr ied out by Galbraith Laboratories, K n o x v i l l e , Tennessee, and by Mr. P. Borda of th is department (see Appendix). Melting points of so l ids i n sealed glass c a p i l l a r y tubes were recorded on a Gallenkamp Melting Point Apparatus. Preparation of bis(dimethyldithiocarbamato)nitrosyl iron(I) In attempting to prepare der ivat ives of the formula R 2NCS 2Fe(N0) 2 (R = a l k y l ) from Fe(C0) 2 (N0) 2 and (Fe(N0) 2 X) 2 (X = Br or I) we discovered a f a c i l e , high y i e l d synthesis of [ (CH 3 ) 2 NCS 2 ] 2 FeN0. A so lut ion containing NaS 2CN(CH 3) 2 (0.60 g , 4.19 mmol) i n CH30H (50 ml) was added dropwise to CH30H (35 ml) containing (Fe(N0) 2 Br) 2 (1.09 g, 2.78 mmol). A prec ip i ta te formed immediately and was col lec ted by f i l t r a t i o n . The green s o l i d was r e c r y s t a l l i s e d from CH 2C1 2-hexanes. Y i e l d : 0.55 g (80%). A n a l y s i s , c a l c . for C 6 H 1 2 N 3 0S 1 + Fe : C, 22.1; H, 3.68; N, 12.9. Found: C, 22.1; H, 3.87; N, 12.4. Infrared (CH 2 C1 2 ) : V^Q , 1717 cm" 1 and v ^ , 1535 cm" 1 . Dehydration of lanthanide chloride hexahydrates In essence, th is method i s a v a r i a t i o n of one used to dehydrate t r a n s i t i o n metal hydrates. The reaction i s as fo l lows : 27 MC1 n .xH 20 + xSOCl 2 + MCl n + xS0 2 + 2xHCl The dehydration of SmCl 3 .6H 20 i s a typ ica l example of th is method. P u r i f i e d thionyl chloride (40 ml) was added to f i n e l y ground SmCl 3 .6H 2 0 (14.2 g) and refluxed with s t i r r i n g for 24 h. Excess S0C12 was removed by evaporation at reduced pressure and then by heating to 80°C for 3 h at 10~2 mm Hg. The s o l i d was transferred to a dry box, ground to a f ine powder and treated again with S0C12 (40 ml) at re f lux for 24 h. Af ter evaporation of S0C1 2 , the s o l i d was heated to 80°C for 16 h at 10" 2 mm Hg. Analysis for chlorine was sa t i s fac tory at th is point , e .g . c a l c . for SmCl 3 , C l , 41.5; found C l , 41.9. Preparation of the cyclopentadienyldicarbonylnitrosylmetal compounds The compounds (h_ 5-C 5H 5)M(C0) 2N0 (M = Cr , Mo or W) were a l l synthesized i n an ident i ca l manner. The method of preparation, using the chromium complex as a typica l example, was as fo l lows . Na[(Jx 5-C 5h5)Cr(C0) 3] (1.77 g, 7.90 mmol) was suspended i n die thyl ether (30 ml) and the mixture was cooled to 10°C. Then, N-methyl-N-nitroso-p-toluenesulphonamide (1.81 g, 8.5 mmol) dissolved i n diethyl ether (10 ml) was added slowly to the suspen-s i o n , whereupon the reaction mixture evolved gas and darkened in colour . Af ter the addit ion was complete, the reaction mixture was s t i r r e d for a further 15 min at 10°C. The solvent was removed under vacuum and the remain-ing residue was sublimed (50°C, 10" 2 mm Hg) on to an ice-cooled probe. Bright orange crys ta ls of the desired product were thus obtained i n at least 60% y i e l d . A n a l y s i s , c a l c . for C 7 H 5 C r N 0 3 : C, 41.4; H, 2.5; N, 6.9. Found: C, 41.3; H, 2.6; N, 6.8. 28 The molybdenum and tungsten compounds of comparable pur i ty were obtained in s i m i l a r y i e l d s . The physical and chemical properties of these complexes have been extensively d e s c r i b e d . 8 4 Preparation of the tris(methylcyclopentadienyl )samariumoctacarbon.yldicobalt  2:1 adduct (C 6 H 7 ) 3 Sm (1.59 g, 4.10 mmol) was dissolved in toluene (30 ml) to produce a bright orange s o l u t i o n , and Co 2 (C0) 8 (1.40 g, 4.10 mmol) was dissolved in toluene (15 ml) to produce a red-brown s o l u t i o n . The two solutions were f i l t e r e d into a common f lask and the mixture of f i l t r a t e s was rapidly s t i r r e d . Within 20 min the so lut ion became cloudy and a s o l i d began to p r e c i p i t a t e . After a further 2 h of s t i r r i n g at room temperature, the reaction was f i l t e r e d and the yellow orange s o l i d thus obtained was washed thoroughly with toluene (5 x 20 ml) u n t i l the washings were co lour less . The remaining bright yellow s o l i d was dried overnight i n vacuo. The y i e l d (2.10 g) was v i r t u a l l y quant i ta t ive . A n a l y s i s , c a l c . for C^H^CoC+Sm: C, 47.3; H, 3.8; Sm, 26.9. Found: C, 46.8; H, 4 .0 ; Sm, 26.3. The extremely a i r - and moisture-sensi t ive adduct i s rapid ly converted to a green s o l i d i f exposed to moist a i r and even decomposes over a period of several days in an atmosphere of prepuri f ied nitrogen. I t i s insoluble in non-polar organic solvents such as benzene, toluene and hexane. It dissolves i n polar solvents such as diethyl ether, tetrahydrofuran, dichloromethane, and a c e t o n i t r i l e only with attendant d i s s o c i a t i o n , as indicated by the appear-ance of the c h a r a c t e r i s t i c red-brown colour of Co 2 (C0) 8 and the diagnostic infrared bands of th is species i n the CO-stretching region. 29 Preparation of the tris(methylcyclopentadienyl)samarium-bis(h 5 -cyclopenta- dienyl )dicarbonyl i ron 2:1 adduct [ (h_ 5 -C 5 H 5 )Fe(C0 ) 2 ]2 (0.49 g, 1.38 mmol) dissolved in benzene (15 ml) was added dropwise at room temperature to a s t i r r e d solut ion of (C 6 H 7 ) 3 Sm (1.01 g, 2.68 mmol) dissolved i n benzene (50 ml) . Within a few minutes a s o l i d began to p r e c i p i t a t e . S t i r r i n g was continued for 2 h to ensure completion of the react ion. The reaction mixture was f i l t e r e d to obtain a bright red s o l i d which was washed thoroughly with pentane (3 x 10 ml) and dried i n vacuo. The s o l i d was r e c r y s t a l l i z e d from a dichloromethane-hexane so lut ion to obtain red microcrystals of the adduct i n 70% y i e l d . A n a l y s i s , c a l c . for C 2 5H 2 6 Fe0 2 Sm: C, 53.2; H, 4 .6 ; Fe, 9.9. Found: C, 51.1; H, 4 .6 ; Fe, 9.9. The low carbon value i s not unexpected 5 i n the analysis of organolanthanide complexes. The red compound i s rapid ly destroyed by a i r and moisture. I t i s thermally unstable above 120°C at 10" 1 mm Hg, and attempts at vacuum sublima-t ion of the complex give only low y i e l d s of [ ( J i 5-C5H5 )Fe(C0) 2 ] 2 . Further, the adduct i s v i r t u a l l y insoluble i n a l l common organic solvents except dichloromethane and tetrahydrofuran, but once i n so lut ion i t experiences a l -most complete d i ssoc ia t ion as shown by i t s infrared spectrum. Preparation of the complexes ( c p ) ? L n [ M ( h 5 - C s H s ) ( C O )3] , where cp = C S H S or  C K H 7 , Ln = Dy, Ho, Er , or Yb, and M = Mo or W. Method A: Reaction of (cp) ? LnCl with N a [ M ( h 5 - C s H Q ( C O ) , M = Mo or W. The preparation of (cp) 2Yb [W(h_5-C5H5) (CO) 3] t y p i f i e s the synthetic method employed to obtain a l l the (cp)2Ln [M(h_ 5 -C 5 H 5 ) (CO) 3 ] compounds. A so lut ion containing Na[(J i 5 -C5H5)M(C0) 3 ] (1.00 g, 2.81 mmol) i n THF (40 ml) was syringed into THF (40 ml) containing (C 5 H 5 ) 2 YbCl (0.956 g, 2.82 mmol). 30 A prec ip i ta te formed immediately, and the resultant brown mixture was s t i r r e d overnight at room temperature. The mixture was then centrifuged and the supernatant brown l i q u i d was decanted from the s o l i d matter. The solut ion was concentrated under reduced pressure u n t i l c rys ta ls began to form, at which point i t was cooled in an i c e - s a l t bath for 2 h. The golden c rys ta l s thus formed were col lec ted by f i l t r a t i o n , were washed with pentane ( 3 x 5 m l ) , and were dried in vacuo. The desired product was obtained i n 75% y i e l d . A n a l y t i c a l data for the complexes are summarised below. Complex Found Calc . (C5H5) 2Yb[W(h 5-C 5H5)(C0)3] C, 33.9; H, 2.4; Yb, 27.2 C 3 4 . 0 ; H, 2.4; Yb, 27.2 (C 5H5) 2 Er[W(h 5 -C 5H5)(C0)3] C, 34.2; H, 2.7; Er , 26.4 C, 34.3; H, 2.4; Er , 26.5 (C 6 H 7 ) 2 Ho[W(h 5 -C 5 H 5 ) (C0) 3 ] C, 36.1; H, 3.3 C, 36.6; H, 2.9 (C 6 H 7 ) 2 Dy[W(h 5 -C 5 H 5 ) (C0) 3 ] C, 36.2; H, .3.0 C, 36.7; H, 2.9 Method B: Reaction of Mo(C0)p; with (cp) 3 Ln The reaction of Mo(C0)6 with ( C 5 H 5 ) 3 E r i s typ ica l of th is route where i t i s successful . A suspension of ( C 5 H 5 ) 3 E r (1.025 g, 2.83 mmol) and Mo(C0)6 (0.75 g, 2.86 mmol) in THF (15 ml) was refluxed overnight, by which time i t had turned orange-red from pink. The solvent was removed under reduced pressure and any excess Mo(C0) 6 , usually n i l , was removed by sublima-t i o n at 10~ 2 mm Hg and 50°C for a few hours. A n a l y s i s , c a l c . for C 1 8 H 1 5 0 3 E r M o : C, 39.9; H, 2.79. Found: C, 41.1; H, 4.00. A l l of these complexes are a i r - and mois ture-sens i t ive , and decompose at temperatures greater than 220°C. They are readi ly soluble i n donor solvents such as THF and dimethylsulphoxide (DMSO), but are decomposed by halogenated solvents such as CH 2 C1 2 . 31 Preparation of the complex Er[ (h 5 -C s H s )Mo (C0)c i ]^.7H ? 0 A solut ion containing Na[(h_ 5 -C 5H 5 )Mo(C0) 3] (3.22 g, 12.0 mmol) i n water (120 ml) was f i l t e r e d into an aqueous solut ion (60 ml) containing E r C l 3 . 6H 2 0 (1.27 g , 3.30 mmol). A s o l i d immediately deposited. Af ter s t i r r i n g at room temperature for 15 min, the s o l i d was co l lec ted by f i l t r a -t ion and was. washed with water (3 x 15 ml ) . The resu l t ing tan s o l i d was dried over P t t 0 1 0 i n vacuo for one day. Y i e l d , 1.65 g (48%). A n a l y s i s , c a l c . for C 2 l + H 2 9 E r M o 3 0 1 6 : i C , 28.0; H, 2.8. Found: C, 28.0; H, 2.4. The product i s t o t a l l y insoluble in water or common organic solvents and i t does not sublime. Moreover, i t decomposes within minutes upon exposure to a i r to y i e l d detectable amounts of [ (h_ 5 -C 5 H5 )Mo(C0) 3] 2. 32 2.3 RESULTS AND DISCUSSION Evidence for the Lewis acid behaviour of the R^Ln complexes We have u t i l i z e d infrared and n.m.r . spectroscopy (as well as elemental analysis where possible) to provide evidence that R 3Ln (R = Cp or MeCp) species can generally function as Lewis acids towards a var ie ty of base s i t e s . The fo l lowing discussion i s divided into sections dealing with each type of the electron donors invest igated. The infrared data discussed are contained i n Table I . A) Terminal n i t r o s y l and carbonyl ligands Upon complexation of h_ 5-CpCr(N0) 2Cl and R 3Ln i n CH 2 C1 2 the i n f r a -red spectrum shows that both of the NO stretching frequencies are lowered (e .g . 1818 to 1786 and 1712 to 1686 cm" 1 for Cp 3 Yb), thereby implying a weakening of the N-0 bond as electrons are donated by the oxygen atom of the l igand to the lanthanide. In order to explain why both N-0 stretching frequencies are lowered upon adduct formation, several p o s s i b i l i t i e s must be considered: (a) R 3Ln has attached to both n i t r o s y l groups. This conclusion cannot be correct because for a 1:1 stoichiometry no parent n i t r o s y l bands were observed using Cp3Yb or Cp 3Er as acceptors, (b) Some sort of averaging i s occurring with the R 3Ln attaching and leaving each n i t r o s y l i n turn . The time scale of the infrared measurement compared to that for a Lewis acid-base equi l ibr ium, however, suggests that th i s explana-t i o n i s u n l i k e l y , (c) Coupling of the complexed and non-complexed N-0 v i b r a -t ions occurs. Local symmetry of the adduct i s so low, C l s that th is explana-t i o n i s quite feasible." The extent of complexation depends upon the lanthanide employed. TABLE I Infrared spectra in the carbonyl and n i t r o s y l s tretching region Lewis base Lewis acid CO and NO absorptions, crrf 1 (A) Terminal n i t rosy l and carbonyl ligands CH?C1? solutions (h 5 -Cp)Cr(NO) 2 Cl (h 5 -Cp)Cr(N0) 2 Cl (h 5 -Cp)Cr(N0) 2 Cl (h 5 -Cp)Cr(NO) 2 Cl (h 5 -Cp)Cr(N0) 2 Cl (h 5 -Cp)Cr(N0) 2 Cl Cs Hfi solutions (h 5 -Cp)Cr(N0) 2 Cl (h_5-Cp)Cr(N0)2Cl CH? C17 solutions (h5-MeCp)Mn(C0)3 (h?-MeCp)Mn(C0)3 (h5-MeCp)Mn(C0)3 (h5-MeCp)Mn(C0)3 (h_5-MeCp)Mn(C0)3 >2 (MeCp)3Sm Cp 3Er Cp3Yb Cp2YbCl (MeCp)2YbCl (MeCp).Sm 1 or 2 (MeCp)3Sm Cp3 Er or Cp3 Yb 2 (MeCp)3Nd Cp? YbCl 1818s, 1712s 1818sh, 1784s, 1712sh, 1684s 1786s, 1688s 1786s, 1686s 1818s, 1712s 1818*, 1784sh, 1712s, 1684sh 1820s, 1705s 1820m, 1775s, 1705m, 1680s 2018s, 1928br,s 2018s, 1928br,s, 1868m 2018s, 1928br,s, 1868m 2018s, 1928br,s, 1865m 2018s, 1928br,s Table I (continued) (h_5-Cp)Cr(C0)2(N0) (h 5-Cp)Cr(C0) 2(N0) (h 5-Cp)Cr(C0) 2(N0) (h_5-Cp)Cr(C0)2(N0) (h5-Cp)Mo(C0)2(N0) (h 5-Cp)Mo(C0) 2(N0) (h5-Cp)W(C0)2(N0) (MeCp)3Sm (MeCp)3Ho >2 (MeCp)3Ho up to 3 Cp<3Yb (h5-Cp)W(C0)2(N0) (MeCp)3Er (B) Bridging carbonyl ligands (a) Nujol mulls [ (h 5 -Cp)Fe(C0) 2 ] 2 [ (h 5 -Cp)Fe(C0) 2 ] 2 2 (MeCp)3Sm [ (h 5 -Cp)Fe(C0) 2 ] 2 2 Cp3Sm Co 2 (C0) 8 Co 2 (C0) 8 2 (MeCp)3Sm 2018s, 1945s, 1692s 2038s, 2018s, 1973s, 2038s, 2018s, 1973s, 2038s, 2018s, 1973s, 2016s, 1938s, 1662s 2035s, 2016s, 1968s, 2000s, 1923s, 1658s 2030s, 2000s, 1955s, 1945s, 1692s, 1635s 1945s, 1692s, 1635s 1945s, 1692s, 1635s 1938s, 1662s, 1586s 1923s, 1658s, 1580s 1955s, 1938s, 1752s 2024s, 1980br,s, 1700br,s 2020s, 1985br,s, 1700br,s 2035sh, 2015br,s, 1846sh, 1830br,s 2025vs, 1941br,s, 1841br,s, 1781br,s Table I (continued) (b) CH2C12 solutions [ (h 5 -Cp)Fe(C0) 2 ] 2 [ (h 5 -Cp)Fe(C0) 2 ] 2 2 (MeCp)3Sm (h 5 -Cp)Ni(C0) 2 Fe(h 5 -Cp)(C0) (h 5-Cp)Ni(C0) 2Fe(h 5-Cp)(CO) (h 5-Cp)Ni(C0) 2Fe(h 5-Cp)(CO) >2 (MeCp)3Ho >2 (MeCp)3Gd [ (h 5 -Cp)Ni(C0) ] 2 [ (h 5 -Cp)Ni(C0) ] 2 C ( h 5 - C p ) N i ( C 0 ) ] 2 (MeCp)3Er 2 (MeCp)3Sm [(h 5-Cp)Mn(C0)(N0)] 2 [(h 5-Cp)Mn(C0)(N0)] 2 [(h 5-Cp)Mn(C0)(N0)] 2 [(h 5-Cp)Mn(C0)(N0)] 2 2 Cp 3Er 3 Cp3Dy 4 Cp 3Er 1994s, 1953s, 1774s 1993s, 1952m, 1772s, 1700w 1995s, 1810s 2010s, 1738s 2010s, 1738s 1886w, 1824s 1888m, 1824br,s, 1780m 1888m, 1842br,s, 1780m 1962s, 1785s, 1708s, 1510s 1985br,s, 1790m, 1734s, 1708s, 1680s, 1525s 1990s, 1790m, 1735s, 1685s, 1525s 1990s, 1790vw, 1735s, 1685s, 1525s 36 For example, even though (MeCp)3Sm i s present i n excess, the spectral data indicate the presence of a small amount of the uncomplexed n i t r o s y l compound, whereas for Cp 3Er and £p3Yb complete complexation i s observed as expected. Moreover, i f an a l ternat ive solvent such as benzene i s used, complete adduct formation with (MeCp)3Sm s t i l l does not occur. The choice of solvents for these studies i s l i m i t e d by the s o l u b i l i t y of the R 3Ln species to benzene, toluene and dichloromethane. Strongly coordinating solvents , such as THF and DMSO, prevent isocarbonyl formation and are obviously unsuitable. In f a c t , the CH 2 C1 2 solutions shows signs of decomposition within a couple of hours for the least stable cases, that i s , where R = Cp i n R 3 Ln. When R = MeCp the organolanthanides are less susceptible to deter iorat ion in ch lor -inated solvents and solutions can be stored up to two days in a nitrogen at-mosphere. I t should also be noted that the magnitude of the s h i f t to lower wave numbers of the NO frequencies i s r e l a t i v e l y constant for a l l the lanthanide complexes which form i s o n i t r o s y l l inkages. Further, Cp2YbCl and (MeCp)2YbCl show no or very l i t t l e adduct formation respect ive ly . The s l i g h t complexation of the (MeCp)2YbCl did not increase upon standing twenty-four hours nor upon adddition of more Lewis acid up to a r a t i o of 1.5:1 for (MeCp) 2YbCl: (h^ 5-Cp)Cr(NO) 2Cl. This observation i s further supporting evidence for the existence of R 2LnCl as chlorine bridged dimers i n non-donor s o l v e n t s . 7 7 The interac t ion of the organolanthanides with the oxygen atom of a terminal carbonyl groups i s weaker than that previously reported for var-44 46 50 ious aluminium systems. ' Consistent with group theoret ical predic-tions of lowered symmetry, three bands are observed in the carbonyl stretching region of the infrared spectrum when (h5-MeCp)Mn(C0)3 and R 3Ln are allowed to 37 react i n a 1:1 r a t i o . Two of the bands are unchanged from those of the parent compound, but a t h i r d weaker band appears in a l l cases around 1868 cm" 1 . (When aluminium i s used as the a c i d i c s i t e i n cpylidMo(C0) 3 the - i 44 N lowest absorption i s around 1660 cm 1 . ) No change i n the spectrum occurs when the r e l a t i v e quantity of R 3Ln to (h_5-MeCp)Mn(C0) 3 i s increased beyond 1:1. Once again R 2LnCl exh ib i t no behaviour a t t r ibutab le to the formation of an isocarbonyl l inkage. The fact that the parent' absorptions are unshifted and strong in the adduct whi l s t the new band i s weaker may suggest that complexation i s not complete up to the stoichiometry of 1:1. Upon complexation there i s a loss of local symmetry of the Mn(C0)3 group. Consequently, the stretching frequencies of the non-complexed carbonyls in the adduct may be coincident with those of the parent molecule, p a r t i c u l a r l y i f the in terac t ion i s weak. In both the terminal n i t r o s y l and carbonyl cases, d i l u t i o n of the adduct so lut ion below ca. 10~3 M causes substantial d i s s o c i a t i o n . For th i s reason the infrared spectra were recorded at a concentration close to 5 x 10" 2 M. The compounds (h_5-Cp)M(C0)2N0 (M = Cr , Mo or W) possess both terminal carbonyl and n i t r o s y l l igands . Our spectral data indicate that the n i t r o s y l l igand i n these complexes i s a better Lewis base than the carbonyl l igands , i r respect ive of the group Via metal or the lanthanide involved. For example, (Ji 5-Cp)Cr(C0) 2N0, in the presence of an equimolar amount of (MeCp)3Sm, exhibi ts new absorptions at 2038, 1973, and 1635 cm" 1 i n addit ion to those normally observed for the parent compound, The lowest band i n th i s region of the infrared spectrum c l e a r l y v e r i f i e s the existence of an i s o n i t r o s y l l inkage. The observations of two raised terminal carbonyl frequencies i s consistent with the explanation offered e a r l i e r . (See section 38 2-1.) Again, an increase i n the r e l a t i v e amount of R 3Ln beyond the 1:1 stoichiometry does not cause complete complexation with any of the Group Via compounds, and a l l of the adducts detected are exc lus ive ly of the i s o -n i t r o s y l type. B) Bridging carbonyl ligands When [(h_ 5-Cp)Fe(C0) 2] 2 and (MeCp)3Sm are reacted i n a 1:2 r a t i o i n benzene at room temperature, a bright red a i r - s e n s i t i v e s o l i d rapidly p r e c i p i t a t e s . The elemental analyses of the r e c r y s t a l l i s e d s o l i d are con-s i s tent with the complex being formulated as [(h_ 5-Cp)Fe(C0) 2] 2.2Sm(MeCp) 3. The infrared spectrum of th i s s o l i d as a Nujol mull indicates that i s o -carbonyl linkages have been formed by the samarium atoms at each of the bridging carbonyl groups. The s h i f t to higher frequencies of the terminal CO stretches and the s h i f t to lower frequency of the bridging CO stretch r e l a t i v e to those observed for the uncomplexed iron compound are spectral features also exhibited by the known [ ( h _ 5 - C p ) F e ( C 0 ) 2 ] 2 . 2 A l E t 3 . 2 6 These s h i f t s are again smaller in magnitude for the samarium adduct than for the aluminium compound, but they do indicate that the two complexes are quite probably i s o s t r u c t u r a l . The x-ray study of the E t 3 A l adduct, which was discussed e a r l i e r (section 2-1), has conclusively shown the ac id ic aluminium atoms to be coordinated to the oxygen ends of the bridging car-bonyl groups. In donor (THF) or polar (CH 2C1 2) solvents , the infrared spectrum shows [(h_ 5-Cp)Fe(C0) 2] 2.2Sm(MeCp) 3 almost completely d issoc ia ted ; however, i n the case of CH 2 C1 2 the s o l u b i l i t y i s only s l i g h t . I f (h_ 5-Cp)Ni(C0) 2Fe (h_5-Cp)(CG) i s employed as the Lewis base, only bridging isocarbonyl behaviour i s observed for CH 2 C1 2 solutions containing greater than a twofold excess of 39 R 3 Ln. The mixed n i c k e l - i r o n complex thus forms adducts which are less susceptible to d i s soc ia t ion and appears to be a better donor of electrons than the i soe lec t ronic i ron compound, although these effects may accrue from increased s o l u b i l i t y . As mentioned e a r l i e r (section 2-1), Co2 (C0) 8 exis ts in so lut ion in two isomeric forms: one containing bridging carbonyl groups whi l s t the other has only a metal-metal bond j o i n i n g two 00(00)1+ u n i t s . When C o 2 ( C 0 ) 8 and (MeCp)3Sm are reacted in a 1:1 r a t i o i n toluene at room temperature, only a 1:2 adduct readi ly prec ipi tates as an extremely a i r - and moisture-sensit ive yellow s o l i d . Its low resolut ion infrared spectrum (as a Nujol mull) i s q u a l i t a t i v e l y d i f f e r e n t from that reported for the A l B r 3 adduct investigated by Cotton and Monchamp,,65 the only l i t e r a t u r e report at the time. That i s to say, the spectrum exhibi ts a band which i s c h a r a c t e r i s t i c of a bridging isocarbonyl l inkage at 1781 c m " 1 , plus an e n t i r e l y d i f f e r e n t contour of higher frequency terminal carbonyl bands r e l a t i v e to C o 2 ( C 0 ) 8 . The presence of (MeCp)3Sm displaces the C o 2 ( C 0 ) 8 equi l ibr ium completely to the form involving bridging CO l igands , the preferred s i tes of Lewis b a s i c i t y . The >C0-Sm(MeCp)3 bonds are established presumably because the hard lanthanide ac id favours the harder oxygen of the CO groups over the sof ter metal-metal bond, although s t e r i c factors may also be operative during the formation of t h i s compound. Regrettably, the adduct i s not amenable to study in so lut ion since i t rapid ly dissociates i n those solvents in which i t i s so luble . Recently, th is work has been substantiated by several invest igators using other Group III Lewis a c i d s . 6 6 - 6 8 Once again a comparison of the magnitude of the s h i f t s of carbonyl stretching frequencies shows that those of the aluminium system are greater for complexed and uncomplexed carbonyl groups than those of the organolanthanide adduct. 40 In l i g h t of the resul ts with Co 2 (C0) 8 i t was decided to use the s t r u c t u r a l l y s i m i l a r [(h_ 5-Cp)NiC0] 2 as i t offers both bridging carbonyls and a bent metal-metal bond as potential Lewis base s i tes i n s o l u t i o n . In CH 2 C1 2 when [ ( j i 5 -Cp)NiC0] 2 i s treated with various R 3Ln complexes, a new band appears i n the infrared spectra at 1780 cm" 1 and the band at 1886 cm" 1 becomes s l i g h t l y more intense. The spectra are invar iant when the lanthanide concentration i s increased beyond the 1:1 stoichiometry. In subsequent work 34 Shriver made the claim that the increased in tens i ty of the higher frequency band upon complexation arises from the lack of a centre of symmetry i n the 1:1 adduct. He pointed out that although the i n i t i a l (NiC0). 2 group is not planar, the deviat ion i s probably small since the dipole moment in benzene i s not large (0.38D) and th i s would account for the weak symmetric stretch i n the parent carbonyl . These spectral features strongly suggest, therefore, the formation of a 1:1 adduct and they confirm the preference of the l a n -thanides for the bridging carbonyl groups. Attempts to i s o l a t e the adducts from toluene solutions meet with f a i l u r e . The molecular structure of [(h_5-Cp)Mn(C0)N0]2 in the s o l i d state 85 i s not yet known with c e r t a i n t y . In so lut ion the compound i s thought to ex i s t as a mixture of dissymmetric c i s and trans isomers. Recent 1H n .m.r . 86 studies indicate that rapid intramolecular posi t ional exchange of CO and NO groups occurs. The CH 2 C1 2 so lut ion infrared spectrum of the compound exhibi ts bands at 1962s (terminal v (C0) ) , 1785s (bridging v (C0) ) , 1708s (terminal v(N0)) and 1510s cm 1 (bridging v (N0)) . As progressively increasing amounts of R 3Ln species are added to th i s s o l u t i o n , new bands appear and grow i n in tens i ty in the infrared spectrum while the absorptions of the parent carbonyl -n i t rosy l gradually diminish in i n t e n s i t y . The l i m i t i n g case i s reached (Table I) at a mole r a t i o of base:acid = 1:4, when the fol lowing bands 41 are observed: 1990s, 1790vw, 1735s, 1680s, and 1525s cm" 1 . This spectrum indicates that with [(h5-C5H5)Mn(C0)N0]2 bridging isocarbonyl and terminal i s o n i t r o s y l bonds have been p r e f e r e n t i a l l y formed by the lanthanide element. This r e s u l t i s unexpected s ince , arguing by analogy with the purely carbonyl systems, one would have predicted the Lewis acid to add to the bridging n i t r o s y l l igand rather than the terminal one. Not a l l carbonyl or n i t r o s y l ligands are s u f f i c i e n t l y basic to form isocarbonyl Or i s o n i t r o s y l linkages (as determined by infrared spectros-copy) with the acids under i n v e s t i g a t i o n . For example, Fe(C0) 2 (N0) 2 , (Ph 2 MeP) 2 Fe(N0) 2 , (Me 2NCS 2) 2Fe(N0), (h 5 -Cp)Fe(C0) 2 I , (h t t -C 8 H 8 )Fe (C0) 3 , (Ph 3 P) 2 RuN0Cl 3 , and [(Ji 5-Cp)Mo(C0) 3] 2 exhib i t no changes i n t h e i r customary carbonyl and/or n i t r o s y l s tretching frequencies when i n solutions also con-ta in ing various R 3Ln complexes. The [(h_5-Cp)Mo(C0) 3 ] 2 compound i s p a r t i c u l a r l y 37 in teres t ing because, as mentioned e a r l i e r (section 2-1) , C 1 3 n .m.r . spec-troscopy reveals no carbonyl transfer v i a a bridging intermediate, yet addi -t ion of ( i - B u ) 3 A l i n heptane 3 4 (at concentrations that are unattainable for the R 3Ln complexes) forms an equi l ibr ium product whose infrared spectrum i s consistent with two R 3A1-coordinated bridging carbonyls and four terminal carbonyls. Upon standing th i s equi l ibr ium product i s converted into a k i n e t i c product with d i f f e r e n t carbonyl absorptions i n the range 1990 - 1620 cm" 1 , however, unl ike the equi l ibr ium product, addit ion of tr iethylamine does not regenerate [ ( J i 5 -Cp)Mo(C0) 3 ] 2 . The k ine t i c product i s under further invest igat ion according to Shriver who does not speculate upon i t s nature. I t i s worthwhile noting the s i m i l a r i t y between the infrared spectrum of Al [ ( J i 5 -Cp)W(C0) 3 ] 3 . 3THF 5 0 (section 2-1) and Shr iver ' s k i n e t i c product as well as our own work (see below and Table I I I ) with anions of the type [(J l 5 -Cp)M(C0) 3 ]" (M = Mo or W). The k i n e t i c product may involve cleavage of 42 the Mo-Mo bond with the equi l ibr ium product acting as an intermediate. This type of cleavage i s not without precedent, for BC13 performs i n th i s C O way with [(h_ 5-Cp)Fe(C0) 2] 2 a n d almost c e r t a i n l y so does ytterbium metal with [(h_ 5-Cp)Mo(C0) 3] 2 (see Chapter I I I ) . Regardless of the true nature of these der ivat ives of [ (h. 5 -Cp)Mo(C0) 3 ] 2 , th i s i s another example of organo-aluminiums being more react ive towards carbonyl-containing e n t i t i e s than R 3 Ln. Considering the accumulated infrared data (Table I) of R 3Ln i n t e r -actions with the various carbonyl or n i t r o s y l l igands , i t should be noted that the magnitude of the s h i f t to lower wave numbers of the CO or NO frequency concerned i s r e l a t i v e l y constant for a l l the organolanthanides which form an isocarbonyl or i s o n i t r o s y l l i n k . Consequently th is spectral technique can only provide q u a l i t a t i v e information about the degree of complexation and i t cannot be used to es tabl i sh a quant i tat ive scale of a c i d i t i e s for the R 3Ln species. On the other hand our resul ts do not imply that a l l R 3Ln are of comparable a c i d i t y , neither do our resul ts suggest that infrared spectroscopy i s insens i t ive to the differences i n Lewis a c i d i t y of organolanthanides. For example, although (MeCp)3Sm, (Cp) 3 Er and (Cp) 3Yb a l l produce the same s ize s h i f t s in the n i t r o s y l absorptions of (h_5-Cp)Cr(NO)2Cl i n C H 2 C 1 2 , in the case of (MeCp)3Sm parent n i t r o s y l bands are also present, indica t ing incomplete complexation, even up to a r a t i o of base:acid = 1:2.15. Supposedly i t i s possible to determine at which point along the lanthanide series complete complexation occurs for a 1:1 stoichiometry; th i s was not pursued. In teres t ing ly , i f other Lewis acids besides R 3Ln are u t i l i z e d , s h i f t s of varying magnitude are obtained for the same base. Unfortunately, a s t r i c t comparison i s not a v a i l a b l e ; however, a consideration of the data 43 involving [ ( J l 5 -Cp)Fe(C0)2]2 with B B r 3 , R3A1 (R = Et or i-Bu) and (MeCp)3Sm is i l l u s t r a t i v e (Table I I ) . Although (MeCp)3Sm did not form a 1:1 adduct with the designated base the isocarbonyl band appears not to vary for 1:1 or 1:2 stoichiometry i n the case of A1R 3 . If th is i s generally t rue , a comparison of the s h i f t s generated by the three acids upon complex formation can be made. In order of increasing s h i f t they are 52 cm" 1 ((MeCp)3Sm), 112 cm" 1 (R3A1) and 345 cm" 1 (BBr 3 ) . C lear ly more deta i led studies (such as s i m i l a r ligands for a l l the a c i d s ) , are necessary before conclusions can be drawn, however, the increasing s ize of the s h i f t with decreasing atomic radius of the Lewis acid central atom i s s t r i k i n g . On th is bas is , a s i m i l a r trend might have been expected across the lanthanide s e r i e s , but th i s appears not to be the case. Moreover, the decrease of the carbonyl s h i f t with the increase of the molecular weight of the acid v e r i f i e s that the lower v i b r a -t ional frequency of the coordinated carbonyl i s not caused so le ly by a mass e f f ec t . 69 Another general corre la t ion that has been suggested i s the con-nection between low V^Q and b a s i c i t y of the carbonyl l igands , a r e f l e c t i o n of the idea that a low V^Q corresponds to high electron density on the carbonyl . Attempts to plot V^Q (unshifted) versus AV^Q for a l l reported cases showed no simple c o r r e l a t i o n , although th i s rule i s worth bearing i n mind as a general guide to those metal carbonyls l i k e l y to display b a s i c i t y . As an example of the d i f f i c u l t y involved i n predict ing a base s i t e + 45 (cpylid)Mo(C0) 3 forms complexes with H and BF 3 bound to the metal , where-44 + as Me 3Al attaches to the oxygen of a carbonyl group. S i m i l a r l y , H adds to the metal i n [ (h_ 5 -Cp)Fe(C0 )2]2> 4 5 but B B r 3 , 6 8 R 3 A 1 , 2 6 and R 3Ln attach to the bridging carbonyl. S ter ic factors may well be a determining factor i n the choice of the base s i t e . 44 TABLE II Infrared spectra of adducts formed from [(h_ 5-Cp)Fe(C0) 2] 2 and some Lewis acids Lewis base Lewis acid Stoichiometry CO Absorptions,cm" 1 Ref. [ (h 5 -Cp)Fe(C0) 2 ] 2 B B r 3 c 1:1 ^2020, a 1849, 1437 68 II R 3 A l d 1:1 2026s, shm 1993b ~. 1983s, 1828m, 1682s 4 II R 3 A l d I f f 2042s, 2009 b, 2004s, ~. 1682s M (MeCp)3Smc 1:2 2024s,*1980br,s, 1700br,s a: The communication implies the terminal bands are more complex than a s ingle absorption. b: Probably a r i s i n g from the trans form of the 1:2 adduct and i n some spectra 34 are not resolved - Shr iver 1 s footnote, c: Nujol mull spectrum, d: Heptane s o l u t i o n . 45 C) Other base s i tes The interact ions of the R 3Ln complexes with Lewis bases can also be monitored by proton magnetic resonance as shown by the representative data displayed in Table I I I . This physical technique need not be hampered by the presence of an excess of Lewis base, and furthermore, i t i s quite s e n s i t i v e ; i t can detect adduct formation for some complexes (e .g . (h_ 4-C 8H 8)Fe(C0) 3) whose infrared spectra do not change i n the presence of R 3 Ln. It i s far more d i f f i c u l t , and in some cases not poss ib le , to use n.m.r . as an independent determination of which base s i t e i s being u t i l i z e d in a compound possessing a number of base s i t e s . For several compounds such a d i s t i n c t i o n does not have to be made. For example, the XH n.m.r . spectrum of (h_5-Cp)2WH2 i n a benzene solut ion also containing (MeCp)3Nd ( p a r t i c u l a r l y chosen because i t does not broaden the observed resonances) shows u p f i e l d s h i f t s of the cyclopentadienyl and hydride resonances. This e f fec t can readi ly be ascribed to a paramagnetic s h i f t of the signals r e s u l t -ing from a Lewis adduct i n which the lone pair of electrons si tuated on tungsten forms a coordinate bond to neodymium. S imi lar compounds are known 45 87 88 i n which R 3A1 ' ' functions as the electron pair acceptor. The observation that (MeCp)3Nd induces an u p f i e l d s h i f t i n the acetylenic proton resonance of phenyl acetylene i s quite interes t ing i n that T s u t s u i ^ has reported that R3Sm catalyzes the t r i m e r i z a t i o n of the alkyne to 1 ,3 ,5-tr iphenylbenzene. Our r e s u l t substantiates the role of the lanthani as a coordination s i t e for ir-bases in such processes, Further, the (h_5-Cp)Fe(C0)2X (X = Cl or I) compounds react with (MeCp)3Nd i n benzene, although the iodide does so s u f f i c i e n t l y slowly for an n .m.r . spectrum to be recorded. This spectrum shows an u p f i e l d s h i f t of the cyclopentadienyl reasonance. Coordination of the Lewis acid can, in p r i n c i p l e , occur at 46 TABLE III Changes induced by (MeCp)3Nd i n the proton magnetic resonance spectra of various Lewis bases Lewis base Concentration, M " f a , ppm Upfie ld sh i f t * 3 , ppm (h5-MeCp)Mn(C0)3 10 ' 1 5.74 C ; 8.34 d 0.36 c ; 0.26 d (h 5 -Cp)Cr(C0) 2 (N0) 10" 1 5.67 1.76 (h*-C 8 H 8 )Fe(C0) 3 10" 1 5.14 0.28 [ (h 5 -Cp)Ni (C0) ] 2 10" 1 4.78 1.10 (h 5 -Cp)Fe(C0) 2 I 10" 1 5.96 1.09 (h 5 -Cp) 2 WH 2 10" 1 5.68 C ; 23 .2 e 0.36 C ; 2 . 3 e C6H5C=CH 10° 7 .20 f o.n f A l l samples, except the one involving C6H5C=CH, were prepared with a saturated benzene so lut ion of (MeCp)3Nd such that the molar r a t i o of neodymium:base was 1:2. a Spectrum of pure compound r e l a t i v e to TMS (T = 10). b S h i f t in resonances af ter adding (MeCp)3Nd. c Cyclopentadienyl protons. d Methyl protons. e Hydride protons. ^ Acetylenic proton. 47 ei ther the CO or the I l i g a n d , but i n view of the infrared resul ts (which showed no isocarbonyl formation), the neodymium i s probably attached at the halide s i t e . Strong supporting evidence for th is assignment comes from 89a the work of Pankowski et a l_ . , who investigated the ef fects of Lewis acids such as A1X3 (X = C l , Br or I ) , F e C l 3 , and SbCl 5 upon the metal carbonyl halides Mn(C0) 5X, (h 5 -Cp)Fe(C0) 2 X, (Me3P).2Fe(C0)2X2, and (Me 3 P) 2 Ni(CO)I 2 (X = C l , Br or I) i n CH 2 C1 2 s o l u t i o n . They concluded on the basis of infrared measurements that Lewis acid-base adduct formation occurred v ia the halogen of the metal carbonyl . They observed no lower carbonyl bands a t t r ibutab le to isocarbonyl l inkages ; i n f a c t , the number of carbonyl bands was conserved i n the complexes suggesting a retention of local symmetry. Some r a i s i n g of the ex i s t ing V^Q was observed upon complexation and th i s corroborates addit ion at the halogen s i t e . Some recent studies by Cullen et a l 8 9 * 3 are also relevant to th i s d iscuss ion. They found that [SbCl 3(Fe(C0) 2(h_ 5-Cp)Cl ) 2 ] 2 could be prepared e i ther by mixing SbCl 3 and (h 5 -Cp)Fe(C0) 2 Cl in ether or by the reaction of Sb*Cl3 with [ (h 5 -Cp)Fe(C0) 2 ] 2 i n CH2C12- A crys ta l structure determination of the i so la ted dimer revealed adduct formation v i a the chlorine attached to i ron and no isocarbonyl l inkages. 57 In recent work, Marks ejt aj_. have shown that Eu(fod) 3 can be used to s h i f t the lti n .m.r . s ignals of appropriate compounds, since i t also coordinates at various base s i t e s , including carbonyl ligands (section 2-1). A d i r e c t comparison of the Eu(fod) 3 and (MeCp)3Nd s h i f t reagents as chemical and s t ructural probes i s not poss ib le , because of the d i f f e r e n t conditions ex i s t ing i n both sets of experiments. We would emphasize, however, that i n spi te of the substantial s h i f t s induced by (MeCp)3Nd (Table I I I ) , i t i s not sui table for general use as a s h i f t reagent because of (a) i t s low s o l u b i l i t y i n common organic solvents ; (b) i t s s e n s i t i v i t y to a i r and moisture; (c) i t s 48 noticeable decomposition i n chlorinated or a c i d i c solvents , and (d) i t s r e a c t i v i t y with some substrates of i n t e r e s t . (A good i l l u s t r a t i o n of the l a s t handicap i s the fact that (MeCp)3Nd reacts rap id ly with (Cp) 2Sn i n benzene to y i e l d a prec ip i ta te which leaves the supernatant solut ion devoid of proton resonances.) Remark (a) may be exemplified by the fact that a saturated solut ion of (MeCp)3Nd i n benzene i s ca. 5 x 10" 2 M. The corresponding R3Eu compound, which would be a more desirable s h i f t reagent because of less broadening, i s very d i f f i c u l t to desolvate reproducibly. The (Cp) 3Eu.THF adduct dissolved i n benzene shows the THF proton resonances -29 ppm and at -63 ppm r e l a t i v e to TMS, ind ica t ing the a b i l i t y of europium to generate large s h i f t s . Nevertheless, (MeCp)3Nd can be employed to detect solut ion i n t e r -actions which are d i f f i c u l t to determine otherwise, even at stoichiometrics less than 1:1; in f a c t , nearly a l l the spectra in Table III were recorded at a base:acid r a t i o of 2:1. Although interact ions with the stronger bases appear to produce the larger s h i f t s , no accurate comment can be made w i t h -out the l i m i t i n g spectrum when the degree of complexation would be known with c e r t a i n t y . Studies involving the incremental addit ion of Lewis acid would also be useful i n that the equi l ibr ium constant for the complex formation could be determined. Reactions involv ing metal carbonyl anions The proton magnetic resonance evidence for the existence of a tungsten-neodymium bond i n benzene solutions containing (h_5-Cp)2WH2 and (MeCp)3Nd encouraged the invest igat ion of the products of the reactions between R 2LnCl compounds and various metal carbonyl anions. As mentioned 49 e a r l i e r (section 2-1) two types of products were thought poss ible : those containing a discrete t r a n s i t i o n metal-lanthanide bond or those containing the lanthanide coordinated to the carbonyl fragment via an isocarbonyl 1inkage. When R 2LnCl and Na[(h_5-Cp)W(C0)3] are reacted in a 1:1 r a t i o i n THF (Method A) the a n a l y t i c a l l y pure R 2Ln[W(h 5-Cp)(CO) 3] (Ln = Dy, Ho, Er or Yb) complexes can be i so la ted i n good y i e l d s . Other carbonyl anions, however, do not produce analogous d e r i v a t i v e s . For example, [00(00)^]" 90 91 i s not s u f f i c i e n t l y nucleophi l ic ' to react with the organolanthanide compound. Indeed, s t i r r i n g Na[Co(C0)i+] with (MeCp)2YbCl i n THF for twenty hours at room temperature and three hours at re f lux produces no change. On the other hand, use of the stronger nucleophiles [(h[ 5-Cp)Fe(C0) 2]~ and [(j^ 5-Cp)Mo(C0) 3]~ resul ts i n the contamination of the products with t h e i r corresponding neutral dimers. The lanthanide-tungsten compounds are obtained as solvent-free c r y s t a l s , which are thermally stable up to 220°C, soluble only in strongly donor solvents such as THF and DMSO, and decomposed by various halogenated solvents. The poor s o l u b i l i t y of these compounds prevents obtaining an n .m.r . spectrum: s u f f i c i e n t l y concentrated d6-DMS0 solutions have the consistency of a glassy syrup and display only very broad, weak absorptions. Spectra could probably be obtained i n d 8 -THF, however, th is i s a very expensive solvent. Conductivity values for d i l u t e THF solutions of the holmium and ytterbium compounds plus a few R 3Ln (Ln = L a , Gd, Ho and Yb) and (MeCp)2HoCl are l i s t e d in Table IV. The R 3Ln species have conduct iv i -t i e s only s l i g h t l y greater than that of the solvent , an observation made e a r l i e r by Birmingham and Wilkinson. The W-Ho and W-Yb compounds have 50 conduct iv i t ies which are l a r g e r , suggesting that these compounds dissoc iate more than the almost e n t i r e l y associated R 3Ln and (MeCp)2HoCl i n THF s o l u -t i o n . The THF s o l u t i o n i i n f r a r e d spectra bear out th i s i n t e r p r e t a t i o n : the strong absorptions in the carbonyl stretching region for the tungsten der ivat ives and for Na[(h_5-Cp)W(C0)3] are i d e n t i c a l . A comparison of the mull spectra (Table V) of R 2 Ln[W(h 5 -Cp)(C0) 3 ] and Na[(h_ 5-Cp)W(C0) 3], however, reveals that the former complexes exh ib i t addit ional bands at 2010m and ca. 1940s cm" 1 as well as a lowering of the lowest carbonyl stretching frequency to ca. 1565 cm" 1 . These features can be interpreted as indica t ing the formation of some isocarbonyl bonds, and the large number of C0-stretching absorptions suggests a polymeric structure for these compounds. The existence of bridging carbonyl groups in th i s structure seems u n l i k e l y i n view of the covalent r a d i i of the metals involved. Moreover, such bridges would involve lanthanide-carbon sigma bonds which would have to cleave i n s o l u t i o n . General ly , organolanthanide compounds thought to involve such a-bonds, are very i n s o l u b l e , polymeric mater ia l s , which are 5 d i f f i c u l t to obtain i n a pure s tate . In a d d i t i o n , i t might be noted that only very recently lanthanide (neodymium and ytterbium) carbonyls have 12 been detected by infrared spectroscopy in argon matrices at 10°K; th i s i s the only report of lanthanide carbonyls. A structure which i s consis-tent with the experimental observations i s one i n which the tungsten units are l inked v ia a bridging framework of the type shown below: Cp Cp Cp Cp Cp Cp I I I ! I I -Ln-0C-W-C0-Ln-0C-W-C0-Ln-0C-W-C0 I ! I I j ! Cp CO Cp CO Cp CO TABLE IV CONDUCTIVITY MEASUREMENTS IN THF COMPOUND CONCENTRATION (10"2M) RESISTANCE (10 5 ohm) MOLAR CONDUCTIVITY (10" 2cm 2 ohms"1 mole" 1 ) TEMPERATURE (°C) Cp 3La 1.62 0.76 11.1 23.9 5.65 5.60 26.2 26.2 Cp3Gd 1.50 0.71 10.6 >25 6.38 <2.51 26.4 26.4 (MeCp)3Ho 1.62 >25 <2.51 27.0 Cp3Yb 1.60 12.6 5.03 25.6 \ (MeCp)2HoCl 1.61 14.7 42.8 25.0 (MeCp)2Ho[(h_5-Cp)W(C0)3] 0.90 0.69 163 25.8 (Cp) 2Yb[(h5-Cp)W(C0) 3] 1.52 0.29 234 25.8 COMPOUND TABLE V Infrared spectra of R 2LnM(]^ 5-Cp)(C0) 3 species v c o , cm Other absorptions, cm" 1 Cp 2Yb[W(h 5-Cp)(C0) 3] i n THF as Nujol mull (MeCp)2Ho[W(Jh5-Cp)(C0)3] in THF as Nujol mull Cp 2Dy[W(h5-Cp)(C0) 3] as Nujol mull Cp 2 Er[W(h5-Cp)(C0) 3 ] as Nujol mull Cp 2Er[Mo(h.5-Cp)(C0) 3] a as Nujol mull C p 2 E r M o(h5 - C p ) ( C 0 ) 3 : b as Nujol mull 2010m, 1936s, 1890s, 1790s, 1740s 2010m, 1984s, 1934s, 1890br,s, 1790br,s 1045m, 1012m, 892w, 790m, 662w, 584m 1565br,s 505m, 484m 1040m, 1010m, 780ms, 580m, 505m, 484m 1045w, 1012ms, 890w, 780ms, 690ms, 662mw, 580m, 505m, 484m 2010m, 1936s, 1890s, 1790s, 1740s 2010m, 1976s, 1930br,s, 1880br,s 1750br,s, 1560br,s 2010m, 1980s, 1934br,s, 1890br,s 1770br,s, 1507br,s 2010m, 1980m, 1910br,s, 1780br,s, 1560br, s 2020m, 1940s, 1885br,s, 1770br,s 1690br,s 1045w, 10<T0m, 780m, 505w 2020m-s, 1940s, 1900br,m, 1790br,m, 1660br,m Table V (continued) Er[h_ 5-CpMo(C0) 3]3.7H 20 as Nujol mull Na[hj5-CpW(CO)3] in THF as Nujol mull Na[h 5-CpMo(C0) 3] in THF as Nujol mull 2020s, 1982s, 1940br,s, 1780br,w 1690br,ww 1936m, 1892br,s, 1790s, 1740s 1980s, 1890s, 1772s, 1690s 1940m, 1895s, 1790s, 1740s 1940w, 1895s, 1775s, 1690s 3600-3400br,m, lOlOw, 810ms, 655m, 570m, 515w, 485m, 455w 805m, 585m, 515m, 505m Prepared by Method A. Prepared by Method B. 54 Such an arrangement s a t i s f i e s the requirement that there be one or more isocarbonyls and one or more terminal carbonyls present. The compounds [(Jx 5-Cp)Mo(C0)3] 2Mg(pyridine) £ t 5^ and Al[W(C0)3(Jh5-Cp)]«3THF,5^ by comparison are both shown to be monomers in the s o l i d state from x-ray analysis (section 2-1). Both compounds are heavily solvated by good donor bases and manage to achieve octahedral coordination about the central metal. The lanthanide-tungsten complexes are not solvated and given the propensity of the rare-earths to high coordination number i t i s u n l i k e l y that a s i m i l a r structure obtains. The central lanthanide element probably re l ieves i t s high pos i t ive charge i n a polymeric s t ruc ture , possibly of the type suggested above. Certa inly the number of strong infrared bands in the carbonyl region cannot be conclusive since the above mentioned magnesium and aluminium compounds have three and s ix respec t ive ly , whi l s t the lanthanide species have f i v e . Unlike Na[(h_5-Cp)W(C0)3] which r a p i d l y smoulders i n a i r , the R2Ln[W(h_5-Cp)(CO)3] s a l t s turn deep red i n dry a i r over a period of two days. Sublimation of t h i s red s o l i d y i e l d s pure [(h. 5-Cp)W(C0) 3] 2 which can be i d e n t i f i e d by i t s elemental a n a l y s i s , melting point and infrared spectrum. Indeed attempts to prepare the R2Ln[Mo(h_5-Cp) (C0) 3 ] ail.ways y i e l d some [( iL 5 Cp)Mo(C0) 3 ] 2 ^. W i l k i n s o n 7 8 has shown that (h.5-Cp)W(C0)3H also oxidises ^The contamination by the molybdenum dimer could r e f l e c t the extremely a i r -sens i t ive character of the complexes or i t could resul t from the presence of a Lewis acid s i t e which promotes dimerization i n the absence of a i r . I t has been shown 1 5 1 that the sa l t s [Ph 1 +As][(Ji 5-Cp)M(C0) 3] (M = Mo or W) react with BF 3 i n dichloromethane or dimethoxyethane to produce a mixture of (h_5-Cp)M(C0)3H and [(h_ 5 -Cp)M(C0) 3 ] 2 , even when a l l manipulations are made using a vacuum l i n e . Presumably the lanthanide could act i n a s i m i l a r ro le to that of B F 3 . 55 i n a i r to the dimer. This s i m i l a r i t y in behaviour of the anion and the hydride suggested that i t should be possible to n i t r o s y l a t e the anion d i r e c t l y , thereby e l iminat ing the intermediate conversion of the anion to the hydride during the preparation of (h_5-Cp)M(C0)2N0 (M = C r , Mo or W.) In f a c t , good y i e l d s {>60% based on the anion) of the desired n i t r o s y l s are obtained i f the metal carbonyl anions are treated i n die thyl ether with N-methyl-N-nitroso-p-toluenesulphonamide (Diazald) . Only one other instance i s known i n which t h i s n i t r o s y l a t i n g agent converts a carbonyl anion to a n i t r o s y l . ; , qq i e . [Mn(C0) 5] to Mn(C0)tfN0. In a d d i t i o n , the lanthanide-Group Via complexes can also be n i t rosy la ted i n THF with Diazald to produce (h. 5-Cp)M(C0) 2N0. The s i m i l a r i t y in chemical properties of Na[(h_ 5- Cp)M(C0) 3] and R 2Ln[M(Jx 5-Cp)(CO) 3] sa l t s suggested another synthetic route to the preparation of the l a t t e r . The d i rec t reaction of M(C0) 6 (M = C r , Mo or W) with NaCp in re f lux ing THF produces Na[(h_ 5-Cp)M(C0) 3], hence the analogous reaction (Method B) was attempted: R 3Ln + M(C0) 6 ^ 1 u x » R 2 Ln(h 5 -Cp)M(C0) 3 + 3C0 R = Cp or MeCp: M = C r , Mo or W. The route i s of interest because R 2LnCl compounds only ex i s t for lanthanides heavier than neodymium, a p e c u l i a r i t y - a t t r i b u t e d to the lanthanide contrac-t i o n . 7 7 The previously described method (Method A) of preparing R 2Ln[(h_ 5-Cp)M(C0) 3] i s not accessible therefore , for a l l the rare earths. Of the Group Via carbonyls only Mo(C0)6 reacts with R ?Ln under the conditions s tudied, namely, i n the absence and presence of l i g h t and up to f o r t y - e i g h t hours re f lux ing i n THF. The reactions were monitored by infrared spectroscopy, 56 and a l l the reaction mixtures were taken to dryness and checked for possible new products by sublimation. In the case of Mo(C0)6 gas i s evolved as expected and products can be i so la ted for Ln = L a , Er and Ho. The a n a l y t i c a l data for these products are a l l consis tent ly h igh , suggesting incomplete conversion. The mull infrared spectra (Table V) exh ib i t the charac ter i s t i c isocarbonyl absorptions also observed f o r the tungsten analogues prepared by the anion route. Further, the behaviour in a i r i s ident ica l with the other lanthanide-Group Via d e r i v a t i v e s , namely a slow darkening to a deep red s o l i d , from which [(h_ 5-Cp)M6(C0) 3] 2 can be sublimed. F i n a l l y , an attempt was made to prepare the t o t a l l y substituted d e r i v a t i v e s , L n A 3 , (A = (h_5-Gp)M(C0)3; M = C r , Mo or W). A related compound, 94 Tl[(h_ 5 -Cp)Mo(C0) 3 ] 3 had been prepared by King. The infrared spectrum of t h i s complex showed three bands i n the terminal carbonyl region and none a t t r ibutab le to isocarbonyls . This compound was said to have a red-green dichro ic appearance resembling cer ta in sixteen electron square planar complexes of rhodium(I) (e .g. C 5 H 7 0 2 R h ( C 0)2) 9 5 and platinum(II) (e .g. P t ( N H 3 ) i t ( P t C l 1 + ) 9 6 j , which can form "stacked" polmeric c rys ta l structures with metal-metal bonding on an axis perpendicular to the coordination square. Subsequently a s ingle c rys ta l x-ray structure determination was undertaken because of the p o t e n t i a l l y high e l e c t r i c a l conduct ivi ty that t h i s compound 97 might d i s p l a y . The resul ts have shown that the proposed "stacked" arrange-ment with T l - T l interact ions i s not confirmed. The molecule was found to have t r igonal pyramidal coordination at tha l l ium with an average Tl-Mo bond o length of 2.965 A , an average Mo-Tl-Mo angle of 119 .7° , and the thal l ium atom o displaced out of the plane of the three molybdenum atoms by 0.586 A. At the time of our experiments, the x-ray resu l t was not published, however, the other spectral and chemical evidence suggested a metal-metal bonded s tructure . 57 Consequently, an analogous reaction was undertaken to prepare a completely substituted lanthanide by employing the metathetical react ions: L n C l 3 + 3NaA * - L n A 3 + 3NaCl A l l of the Group Via anions react rap id ly in the expected manner, but i n every case the product formed i s an intrac table a i r - s e n s i t i v e s o l i d , which i s t o t a l l y insoluble in water or common organic solvents^ and does not sublime. Only one such complex i s reasonably pure, namely, Er[(Jx 5 -Gp)Mo(C0) 3 ] 3 ' 7H 20. This compound decomposes r a p i d l y in a i r to produce the corresponding metal carbonyl dimer, as evidenced by infrared spectroscopy, an e f fect also noticed for the R 2 Ln[(h 5 -Cp)W(C0) 3 ] and Er[Co(C0) t t ] 3 -3THF 7 1 complexes. The infrared spectrum of the Er-Mo compound (Table V) v e r i f i e s the presence of water and shows strong terminal CO bands and only weak lower frequency CO absorptions. I t appears that Er[(h_ 5-Cp)Mo(C0) 3] 3«7H 20 i s some-what s i m i l a r to Er[Co(C0) l t] 3-3THF and Tl [(h_ 5-Cp)Mo(C0) 3] 3 ; In both of the erbium compounds i n the s o l i d state there may be some E r - t r a n s i t i o n metal bonding made possible by the substantial solvat ion of the lanthanide element. The strong carbonyl frequencies above the values observed for the corresponding sodium sa l t s and the lack of strong isocarbonyl bands are consistent with th i s view. With the l i m i t e d evidence avai lable the assignment of a metal-metal bond i s at best tentat ive and any analogies to the thal l ium compound are weak. X X X X The evidence presented i n th i s thesis (and the work of others) ^For example, i n the case of YbCl 3 + Na[(h_5-Cp)W(C0)3] the product shows absolutely no extract ion into acetone during a twenty four hour hot extract ion using a Soxhlet apparatus. 58 documents isocarbonyl linkages as a v a l i d bonding mode, however, the question remains as to what i s occurring at an e lec t ronic l e v e l . Q u a l i t a t i v e l y there appears to be a donation of electrons from an oxygen lone pair o r b i t a l , i f P u r c e l l ' s explanation i s accepted, (section 2-1) to the lanthanide involved. The ro le of the lanthanide must depend i n large part upon the type of bonding in the i n i t i a l Lewis acid complex. The organolanthanides whose bonding has been most in tens ive ly studied are the der ivat ives Cp 3 Ln. For example, work 1 g by Nugent et al_ has shown the covalent character of the lanthanide-r ing bonding to be notgreater than 5% based upon the nephelauxetic parameter determined from absorption spectra. Further support for low covalency comes from another technique; the mass spectra of i o n i c a l l y bound dicyclopentadienyl metal sandwich compounds have been characterised as having the peak correspond--98 ing to the loss of one cyclopentadienyl group as the most intense. On the other hand r e l a t i v e abundancies in covalently bound complexes consis tent ly showed the parent ion as the most intense peak. The lanthanide data are con-19 s i s tent with an ionic model. General chemical properties such as attack by halogenated solvents , s o l u b i l i t y only in donor solvents and the instantaneously 3 quant i tat ive reaction with F e C l 2 to produce ferrocene confirm the highly ion ic nature of these complexes. Unfortunately no x-ray crysta l structure determinations have been carr ied out upon Cp 3Ln-Lewis base systems. Moreover, only a s ingle report of 99 a structure of a Cp 3Ln compound e x i s t s . Apparently Cp3Sm does not e x i s t as a discrete molecule i n the s o l i d phase, but as i n f i n i t e chains with a d is tor ted tetrahedral geometry about Sm. This structure could be interpreted as Cp3Sm act ing as a Lewis acid and i n so doing increasing i t s coordination by a t t r a c t -ing part of a Cp r ing of an adjacent molecule. I f one assumes a C 3 y or even 59 D 3 n symmetry for the R 3Ln part of a Lewis acid-base adduct the most l i k e l y vacant o r b i t a l on the lanthanide i s the 5d z 2 , assuming the 4f o r b i t a l s have 19 i n s u f f i c i e n t radia l extension to enter the bonding. Whatever the o r b i t a l , the in terac t ion between carbonyl oxygen and Cp 3Ln i s not as strong as f o r aluminium and boron complexes as judged by the s h i f t in infrared absorptions (see above). Possibly the dif ference i n s ize of the lanthanide and oxygen o r b i t a l s contributes to th is observation. Hence thellanthanide-oxygen bond may be less covalent than the corresponding aluminium- and boron-oxygen bonds. In the case of the species m2Mgpyit (section 2-1) Burl i t c h 5 1 believes that the Mg^isocarbonyl bond i s predominantly i o n i c . The high r e a c t i v i t y of the m" anions and the fact that the average carbonyl s tretching frequency i s the same in these complexes and i n the corresponding sodium s a l t lead him to th is conclusion. Chemically the organolanthanides resemble Group I la der ivat ives more c lose ly than Group I l i a and so the resul ts of the magnesium systems may well be relevant to our compounds. In a l l p r o b a b i l i t y the bonding between the lanthanide and the oxygen atom of a metal carbonyl or n i t r o s y l l igand i s e s s e n t i a l l y i o n i c . 60 CHAPTER II I REACTIONS INVOLVING ELEMENTAL METALS 3.1 INTRODUCTION The interac t ion of an elemental metal with an organic substrate i s one of the oldest and most d i r e c t methods of preparing an organometal1ic compound. Indeed, the pioneefcing contr ibution of Frankland using zinc and the s igni f icance of Grignard's work f i f t y years l a t e r are well recognized. E l e c t r o - p o s i t i v e metals are most often employed i n t h i s type of reaction e i ther alone or i n an a l l o y . For example, one of the most common organo-m e t a l l i c compounds, Et^Pb, i s prepared i n th i s way. 4(Na/Pb) + 4EtCl - E t ^ P b + 3Pb + 4NaCl Instead of forming an a l l o y , metals can often be made to react more readi ly by reducing t h e i r p a r t i c l e s i z e . Recently, Timrns 1 ^ and other workers 1 *^ have found that low temperature cordeposition of metal atoms with organic species permits the formation of a var ie ty of in teres t ing compounds. Metals such as platinum or palladium (which are known for t h e i r "noble" character) can be used in th i s way i n addit ion to more common t r a n s i t i o n e le -ments. In th i s case chemical discoveries followed as a r e s u l t of the develop-ment of a physical technique. S i m i l a r l y the rare earth chemist has benefitted from the improved separation methods made avai lable i n the l as t twenty years. Consequently, highly p u r i f i e d lanthanide metal powders can now be obtained economically. Even so , the l i t e r a t u r e reports of organometal!ic compounds prepared d i r e c t l y from lanthanide metals are few. 102 The f i r s t attempt at such a reaction involved lanthanum metal , of unspecified s ize and q u a l i t y , heated at 135°C with biphenylmercury i n a sealed tube under nitrogen for 100 days. Af ter treatment of the black l i q u i d 61 product with carbon dioxide and extract ion into benzene, the only i so lab le product was biphenyl in 15% y i e l d . Whether the organic product was the r e s u l t of the thermal decomposition of Ph2Hg or was generated v i a in terac t ion 102 with lanthanum was not determined. In the same study Gilman found that iodobenzene and lanthanum metal i n diethyl ether or benzene at room tempera-ture did not react , even af ter four months. The f i r s t substantial r e s u l t using rare earth metals employed the fact that ytterbium and europium dissolve i n l i q u i d ammonia to form stable 1Q3 divalent cat ions . Fischer treated these solutions with cyclopentadiene to produce Cp2Eu and Cp 2 Yb, which were i so la ted by sublimation of the dried reaction mixture. Cp2Eu i s a ye l low, paramagnetic (v>eff = 7.6 BM) s o l i d whose Debye-Scherrer pattern i s i d e n t i c a l to that of Cp 2 Sr. The evidence supports an ion ic bonding mode for europium. There i s some controversy, however, regarding the species Cp 2Yb. The red diamagnetic s o l i d which sublimed about 103 13 400°C, was said to be unsolvated Cp 2Yb. Calderazzo prepared Cp2Yb v i a the l i q u i d ammonia route and by the reduction of a Y b 3 + species. He cil.aimed Cp2Yb to be emerald green when desolvated and that the red sublimate obtained by Fischer was not exc lus ive ly Cp 2 Yb, but contained some Y b 3 + as evidenced 104 by u l t r a v i o l e t spectroscopy. Later work by Hayes and Thomas on the l i q u i d ammonia reaction revealed three sublimable products: at 150°C, a deep green s o l i d assigned to be Cp 3 Yb; at 170°C, a yellow s o l i d formulated as a mixture of C p 3 Y b 2 ( N H 2 ) 2 ; and >.Cpi tYb 2(NH 2) 2; and at 360°C, a red s o l i d which was claimed to be Cp 2Yb. A l l formulations by Hayes and Thomas were based on detai led mass spectral and n.m.r . evidence. One component, [Cp 2 YbNH 2 ] 2 , of the ytterbium 1 nc mixture subliming at 170°C, has also been observed by M u l l e r . L iquid ammonia solutions of europium and ytterbium also react with 62 cyclooctatetraene to form COTEu and COTYb. Once again, the europium complex was shown to be paramagnetic and to display an e . s . r . signal consistent with the state E u 2 + . The diamagnetic ytterbium complex was characterised by metal ana lys i s . Both compounds were readi ly solvated by strongly donor solvents . Passage of propyne 1^ through solutions of europium or ytterbium i n l i q u i d ammonia produced (CH 3-CEC) EU and the mixture (CH 3 C = C ) 2 Y b / Y b(NH 2) 2 according to elemental ana lys i s . No properties other than the reaction with water to regenerate the alkyne were recorded. Work, more c lose ly related to our own, by Evans and c o w o r k e r s 9 , 1 ^ describes reactions of lanthanide metals with a l k y l or aryl iodides i n THF. I t was found that ytterbium powder and europiumiingots reacted readi ly at -15°C to give greater than 70% conversion to the corresponding Grignard-type species, RM**I (M = Eu or Yb). Samarium powder required heating to 30°C and the conversion was never more than 50%. These reagents behaved towards aldehydes, ketones, R 3 S i C l (R = a l k y l or aryl group), and water as typica l Grignard reagents. Magnetic s u s c e p t i b i l i t i e s as well as iodimetr ic and acidometric t i t r a t i o n s helped to confirm the formulation as RM**I. On the other hand gadolinium and erbium did not react even under vigorous condi t ions , whereas cerium ingots and lanthanum wire did show some r e a c t i v i t y towards a l k y l and ary l iodides . On the basis of a Ce:I r a t i o of 1 :1 .5 , these so lu-tions were t e n t a t i v e l y described as an equimolar mixture of R2CeI and RCeI 2 , or possibly a l a b i l e equi l ibr ium of R3Ce arid C e l 3 . Less data were reported for the lanthanum compounds except that much p r e c i p i t a t i o n of L a l 3 occurred and the conversion was only about 25% compared to about 45% for the cerium species. 6 3 In summary then, the reactions used to prepare organolanthanides using elemental metals have been of the fol lowing types: M + nRH - M R n + n/2H2 ( i ) M + R' *- MR' ( i i ) M + R"X »-R"MX ( i i i ) (R = Cp or CH 3CEC ; R' = COT and R" = a l k y l or aryl group). We were interested i n reaction type ( i i i ) where R i s not an organic group, but rather a t r a n s i t i o n metal organometallic e n t i t y . That i s to say, we were considering the formation of an organometallic Grignard-type reagent based upon the lanthanides. At the time the work was undertaken there were only two reports of s i m i l a r species , both involving magnesium. Since then no further papers on t h i s topic have appeared i n the l i t e r a t u r e . The f i r s t report of a t r a n s i t i o n metal-Grignard reagent was by G l o c k i n g . 1 0 7 Two s i m i l a r complexes were prepared by s t i r r i n g (h^-Cp)(C0)3MGeR3 (M = Mo or W, Me or Et) with anhydrous MgBr2 i n THF for four days; the solvent was removed i n vacuo and, f i n a l l y , the remaining s o l i d was extracted and c r y s t a l l i z e d from benzene. The elemental analyses for both der ivat ives d id not f i t one formulation s a t i s f a c t o r i l y as there appeared to be some non-stoichiometr ic solvat ion by THF; the general formula (h_5-Cp)M(C0)3MgBr-xTHF (x = 0V|) could be assigned to these complexes. A molecular weight determina-t ion of the tungsten compound showed i t to be monomeric i n benzene. In terms of r e a c t i v i t y both molybdenum and tungsten complexes were exothermally hydrolysed to the corresponding hydride and the tungsten reagent could not be converted into the a-phenyl tungsten complex with bromobenzene. No conditions were spec i f ied for these conversion react ions. 64 The only use of elemental magnesium with a t r a n s i t i o n metal organo-1 QO meta l l i c substrate was by Burl i t c h and Ulmer. They used three routes to prepare the Grignard-type reagents: ( i ) Magnesium powder reacted in THF at room temperature with Mn(C0)5Br or (h 5 -Cp)Fe(C0) 2 Cl to produce highly coloured, a i r - s e n s i t i v e solutions which reacted with Ph 3SnCl to give the corresponding Mn-Sn or Fe-Sn compound i n greater than 70% y i e l d . Using these react ive solutions with Mel , i t was shown by infrared measurements that Mn(C0)5Me was obtained i n 90% y i e l d ; ( i i ) An a l ternat ive method involved cleavage of a metal-metal bond in the presence of magnesium powder and (CH 2 Br) 2 i n re f lux ing THF for 20 hours. The resul ts are summarized below:)(yields are in brackets) L-L + (CH 2 Br) 2 + Mg ^"L-Mg-Br" (a) (CF 3 C0) 2 0 " (b) 4gCl 2 (c) Ph3SnCl CF3C0Mn(C0)5 (h 5-Cp)Mo(C0) 3HgCl Ph3Sn-R (79%) (>80%) (>70%) L = Mn(C0) 5 , Fe(C0) 2 (h 5 -Cp) , or Mo(CO) 3(h_ 5-Cp), R = Mn(C0) 5 , Fe(C0) 2 (h 5 -Cp) . This method i s best regarded as a reaction of very react ive MgBr 2 , prepared i n  s i t u by o l e f i n e l iminat ion from (CH 2 Br) 2 and magnesium, with the metal dimers. The presence of halide was found to be essential for the formation of the Grignard-type species. The cleavage of Mn 2 (CO) 1 0 by magnesium only did not occur. From the equations above i t can be seen that besides using M n 2 ( C 0 ) 1 0 and (h_ 5 -Cp) 2 Fe 2 (C0) l t , the Mo-Mo bond i n [(h_ 5-Cp)Mo(C0) 3] 2 was cleaved. I t was found that Ph 3SnCl did not react with the molybdenum-Grignard and t h i s was a t t r ibuted to the weakly nucleophi l i c character of the carbonyl containing group. Burl i t c h remarked that the lack of formation of the Sn-Mo bonded 6 5 compound was hardly surpr i s ing i n so far as the compound iso la ted by G l o o k i n g 1 ^ 7 was produced by a process which was e s s e n t i a l l y the reverse of reaction (c) . Furthermore, the metal-metal bond i n Co 2 (C0) 8 could not be cleaved with magnesium and ( C H 2 B r ) ^ , , since the carbonyl dimer rap id ly decomposed i n the presence of MgBr 2 ; ( i i i ) The f i n a l method was based upon mercury displacement from RHgCl species. For example, when a THF solut ion of (C0)5MnHgBr was s t i r r e d with an excess of magnesium for two hours at room temperature a l i g h t ye l low, highly a i r - s e n s i t i v e so lut ion was formed. Reaction of the so lut ion with Ph 3SnCl gave a 93% y i e l d of Ph 3SnMn(C0) 5. This method was considered the cleanest route to "(C0) 5MnMgBr", however, the colour of the so lut ion prepared i n th i s way was d i f f e r e n t from the deep burgundy so lut ion prepared using ( C H 2 B r ) 2 . The analogous "Co(C0)i+MgBr" could be prepared s i m i l a r l y , but onee^again complete reaction with Ph 3SnCl was not achieved, presumably for the same reason as in the molybdenum case above. No spectral or other physical data were reported for the Grignard-type rea-gents, although i t was noted that preliminary resul ts suggested considerable covalent character i n the magnesium-transition metal bond. This comment i s in teres t ing in the l i g h t of Burl i t c h ' s l a t e r work of preparing der ivat ives of the type m2Mgpy4. (m = metal carbonyl anion, py = p y r i d i n e , section 2-1) , where he found the bonding to be almost e n t i r e l y ionic i n nature. The only reaction involving an elemental lanthanide with an organo-meta l l i c complex i s the reaction of a 1% w/w Er-Hg mixture with [ (C0)i f Co] 2 Hg to produce E r [ C o ( C 0)tJ 3 - 4 T H F , 7 1 which was discussed i n de ta i l e a r l i e r (section 2-1) . o Our attempts to form lanthanide Grignard-type reagents were based 9 II upon the observations that Evans had prepared compounds of the type RM 1 and 66 that magnesium had been shown to produce react ive so lu t ions , t e n t a t i v e l y described as "L-Mg-X" , with t r a n s i t i o n metal organometallic substrates in the presence of haHide. I t seemed reasonable to study the reactions of lan-thanides which can exis t i n a +2 oxidation s ta te , with sui table metal carbonyl hal ides . This approach may be regarded as another route to compounds contain-ing t r a n s i t i o n metal-lanthanide bonds. In addit ion isocarbonyl bonding could be involved once more. Furthermore, resul ts from our l a b o r a t o r y 1 1 0 had shown that " R M 1 1 ! " , formed from a l k y l or a ry l iodides according to Evans1 method, did not always react with organic funct ional groups i n a fashion s i m i l a r to t h e i r magnesium analogues. Although most of th i s work involved Mn(C0)5Br and (h_ 3 -C 3 H 5 )Fe(C0) 3 I , prel iminary experiments were carr ied out using [(h_ 5-Cp)Mo(C0) 3] 2 and (h_ 5-Cp)Cr(C0) 3HgCl. Cleavage of the Mo-Mo bond i n the absence of halogen was also attempted. The d i r e c t cleavage of a metal-metal bond by a metal other than a Group la element i s also a f i e l d which has received l i t t l e at tention to date. For example, Robinson and coworkers 1 1 1 have shown that tha l l ium reacts with Co 2 (C0) 8 in benzene at room temperature to generate Tl[Co(C0)i+]. Zinc and 112 cadmium are known to inser t into the metal-metal bond of Mn 2 (C0) 1 0 when 52 ref luxed i n dig.l'yme for 10 hours. McVicker and Matyas have shown that magnesium amalgam can ef fect the reductive cleavage of dimeric t r a n s i t i o n metal carbonyl complexes i n the presence of a strong Lewis base such as pyridine or THF (see section 2-1 for preparative d e t a i l s ) . M-M + Mg(ng) b a s e . R v M g ( M ) 2 A The species that r e s u l t from metal carbonyl dimer cleavage are reported to have useful synthetic a p p l i c a b i l i t y . The T l [ C o ( C 0 ) i J formed above involved 67 a f a c i l e , high y i e l d preparation, which avoided the use of large quanti t ies of mercury and the formation of unwanted Hg[Co(C0)i t ] 2 , and produced a s a l t which i s soluble in non-basic solvents such as CH 2C1 2 and benzene. Further, Tl[00(00)4] has s u f f i c i e n t s t a b i l i t y towards the a i r to be weighed rap id ly without the use of an i n e r t atmosphere. S i m i l a r l y , the reagents B Mg(M)2 A have excel lent hydrocarbon s o l u b i l i t y , unl ike the sodium sa l t s of metal carbonyl anions. Aside from attempting to determine whether a lanthanide- t rans i t ion metal bond could be formed, these studies using elemental rare earths held the p o s s i b i l i t y of generating synthe t i ca l ly useful reagents. 68 3.2 EXPERIMENTAL The lanthanide metals used were of 99.9% p u r i t y and were purchased from Research Organic/Inorganic Chemical Corporation or the Research Chemicals D i v i s i o n of Nucor Corporation. Manganese, t i n and iron were of ' C e r t i f i e d ' pur i ty from Fischer S c i e n t i f i c Company. Mn'(C0)5Br and [ ( h 5 - C 5 H 5 ) M o ( C 0 ) 3 ] 2 were purchased from Strem Chemicals Incorporated, Mass. M n 2 ( C 0 ) 1 0 , Fe(C0) 5 and Mo(C0)6 were purchased from Pressure Chemical Company, Pi t tsburgh. N 0 C 1 , 1 1 3 M o ( N 0 ) 2 C l 2 , 8 5 [ ( h 5 - C 5 H 5 ) C r ( C 0 ) 3 ] H g C l , 1 1 4 ( h 3 - C 3 H 5 ) F e ( C 0 ) 3 I 1 2 4 and M n 2 ( C 0 ) 9 H 2 1 1 5 were prepared according to l i t e r a t u r e methods. For the reactions involving deactivated alumina, the deactivation was achieved by exposing Fisher adsorption alumina to the atmosphere for at least f o r t y eight hours. For chromatography purposes Woelm neutral grade 1 alumina was used. The pur i ty of reagents was ascertained by elemental analysis and/or melting point determinations. A l l solvents were dried ( i f necessary), d i s t i l l e d , and degassed jus t p r i o r to use, and a l l manipulations were performed i n an atmosphere of prepur i f ied nitrogen. Infrared spectra were recorded on Perkin Elmer 457 and 710 spectro-photometers, and proton magnetic resonance spectra were recorded on a Varian Associates T-60 spectrometer with tetramethylsilane or the solvent being used as internal standards. Electron spin resonance spectra were obtained with a Varian Associates E-3 , X-band spectrometer, and v i s i b l e and u l t r a v i o l e t spectra were obtained with a Cary 14 recording spectrophotometer. Conductance measure-ments were made with a Yellow Springs Instrument Company (YSI) model 31 con-d u c t i v i t y bridge using a ca l ibrated YSI 3403 conduct ivi ty c e l l . Elemental analyses were carr ied out by Mr. P. Borda of th is department. 69 3.2a REACTIONS INVOLVING MANGANESE CARBONYLS  Preparation of the red solut ion  Method A An excess of manganese powder (0.32 g , 5.8 mmol) was vigorously s t i r r e d i n a THF solut ion (30 ml) of Mn(C0)5Br (1.15 g , 4.2 mmol) overnight at room temperature. The metal was allowed to se t t l e and the supernatant l i q u i d was decanted and f i l t e r e d back onto the manganese. S t i r r i n g was continued for a further 2 days; the mixture was f i l t e r e d to y i e l d a deep red s o l u t i o n . Other metals reacted s i m i l a r l y . For the more react ive elements such as ytterbium and samarium, the solut ion generally turned red within 12 h whereas the less react ive metals required up to 5 days of s t i r r i n g i n extreme cases. Method B Manganese powder (0.54 g, 9.9 mmol) was s t i r r e d i n a re f lux ing THF solut ion (40 ml) of M n 2 ( C 0 ) 1 0 (0.80 g , 2.1 mmol) and (CH 2 Br) 2 (2 m l , 23 mmol) for 24 h . The resultant red solut ion was f i l t e r e d to remove excess metal. Ytterbium reacted in an ident i ca l manner. Method C Anhydrous MnCl 2 (0.26 g , 2.1 mmol) was s t i r r e d in a re f lux ing THF solut ion (40 ml) of Mn 2 (C0) 1 0 (0.81 g , 2.1 mmol) for 3 days, by which time the solut ion had turned red. The solut ion was f i l t e r e d before use. Anhydrous YbCl3 d id not react under s i m i l a r conditions a f ter 4 days of r e f l u x i n g . 70 Method D NaMn(C0)5 (0.88 g , 4.0 mmol) [prepared by reduction of M n 2 ( C 0 ) 1 0 with a 1% Na-Hg amalgam in THF] and MnCl 2 (0.50 g , 4;0 mmol) were s t i r r e d overnight in THF (40 ml) at room temperature. The solut ion had turned red and was f i l t e r e d . In order to obtain a solut ion whose infrared spectrum was q u a l i t a t i v e l y s i m i l a r to the spectra exhibited by solutions prepared by other methods, th i s red solut ion was allowed to s t i r for several days or was d i lu ted with THF. A l l the red s o l u t i o n s , whatever the synthetic method employed, were monitored by infrared spectroscopy during t h e i r preparation. In the reactions described below, the actual preparative method of the red solut ion i s generally of no consequence as f a r as the eventual success of the described typical- syntheses i s concerned. The or ig ins of the red solutions are indicated in each descr ipt ion for the sake of completeness. Reaction of the red solut ion with tr iphenyl t i n chloride S o l i d (C 6 H 5 ) 3 SnCl (0.18 g , 0.47 mmol) was added to a s t i r r e d red solut ion prepared from Yb (0.17 g , 0.99 mmol) and Mn(C0)5Br (0.13 g , 0.46 mmol) i n THF (7 ml) according to Method A. The mixture was warmed to 50°C for 1 h , then hydrolyzed at 25°C with saturated aqueous NH^Cl solut ion (10 m l ) , and washed with saturated aqueous NaCl so lut ion (3 x 10 ml) . The remaining brown THF solut ion was dried over anhydrous MgSO^ and taken to dryness. The s o l i d residue was extracted with hot hexane (3 x 15 m l ) , the extracts were combined and f i l t e r e d , and the solvent was removed from the f i l t r a t e under reduced pressure to y i e l d a pale yellow s o l i d . M n 2 ( C 0 ) 1 0 was removed from the l a t t e r s o l i d by sublimation (25°C, IO" 3 mm Hg), and the 71 remaining white s o l i d was r e c r y s t a l l i z e d from hexane to obtain (C 6 H 5 ) 3 SnMn(C0) 5 . Y i e l d 0.15 g , 59%. The i d e n t i t y of the product was confirmed by i t s infrared spectrum, melting point and elemental a n a l y s i s . 1 1 6 Anal . Calc for C 2 3 H 1 5 0 5 S n M n : C, 50.69; H, 2.75. Found: C, 50.99; H, 2.77. Reaction of the red solut ion with methyl iodide A THF solut ion (10 ml) of CH 3I (0.5 mil:, 8.0 mmol) was added to a red solut ion prepared according to Method A from Sm (0.24 g , 1.6 mmol) and Mn(C0)5Br (0.27 g , 0.98 mmol) i n THF (8 ml ) . The reaction mixture was refluxed with s t i r r i n g for 24 h , whereupon H 20 (10 ml) was added and the volume of the so lut ion was reduced to 15 ml . The mixture was extracted with C C I 4 (3 x 7 ml) and the combined extracts were concentrated. Proton magnetic resonance and infrared s p e c t r a 1 1 7 revealed CH3Mn(C0)5 to be present in the extracts . Y i e l d ^ 10%. Reaction of the red solut ion with a l l y l chlor ide A l l y l chloride (7.2 m l , 88.3 mmol) was syringed into a s t i r r e d red solut ion prepared according to Method D from MnCl 2 (0.97 g , 7.7 mmol) and NaMn(C0)5 (1.68 g , 7.7 mmol) i n THF (20 ml ) . The solut ion was s t i r r e d at room temperature overnight, by which time a prec ip i ta te had formed and the solut ion had turned yel low. The products were i so la ted by t rap- to- t rap d i s t i l l a t i o n (0°C, -35°C and -78°C) of the reaction mixture under vacuum. A small amount of M n 2 ( C 0 ) 1 0 co l lec ted i n the 0°C t r a p , but the major carbonyl-containing product was a yellow l i q u i d (1 ml) which col lec ted i n the -35°C t rap . Infrared and n .m.r . spectra showed the l i q u i d to consist only of a 72 mixture of h_ 1-C 3H 5Mn(C0) 5 and h_ 3-C 3H 5Mn(C0) l t i n a r a t i o of 2 .4 :1 . The yellow l i q u i d was heated at 80 - 90°C for one hour and the n .m.r . spectrum of the resu l t ing yellow s o l i d revealed complete conversion to h_3-C3H5Mn(C0) as e x p e c t e d . 1 1 8 The n .m.r . spectrum (C 6 D 6 ) of h_ 3 -C 3 H 5 Mn(C0) l t 1 1 9 was T 5.73 (m, 1 H) , 7.72 (d , J = 7 Hz, 2 H) , 8.62 (d , J = 12 Hz, 2 H). Reaction of the red so lut ion with benzoyl chlor ide Benzoyl chloride (1.0 m l , 8.6 mmol) i n THF (20 ml) was added dropwise to a red solut ion prepared according to Method A from Yb (0.90 g , 5.2 mmol) and Mn(C0)5Br (1.11 g , 4.0 mmol) i n THF (32 ml) . No change was apparent a f ter s t i r r i n g for 2 h at room temperature, and the mixture was warmed to 65°C for 2 h, thereby causing a brown colorat ion to appear. The solvent was removed under reduced pressure at 25°C to y i e l d an o i l -covered orange s o l i d which was extracted with l i g h t petroleum ether (3 x 45 ml) . The combined extracts were reduced i n volume and were transferred to an alumina column which was eluted with a l i g h t petroleum ether-diethyl ether solvent gradient. The two major products obtained by chromatography were M n 2 ( C 0 ) 1 0 (0.43 g , 55% y i e l d ) and C 6 H 5 C0 2 (CH 2 ) 4 C1 (0.41 g , 48% y i e l d ) . The y i e l d s were based on Mn(C0) 5Br. The ester was i d e n t i f i e d spectroscopical ly infrared ( f i l m ) , v m g x 1720, 1280, 1120, 720 cm" 1 [compare with Sadtler standard infrared spectrum 25038]; n .m.r . ( C C l J , T 8.1 (m, 4 H) , 6.43 (m, 2 H) , 5.70 (m, 2 H) , 2.81 - 2.40 (m, 3 H) , 1.98 (m, 2 H) [compare with Sadtler standard n .m.r . spectrum 98444 M]; mass spectrum, m/e 214 ( M + , C l 3 7 212 ( M + , C l 3 5 ) , 77 (base peak). 73 Reaction of the red so lut ion with sodium acetylacetonate Sodium acetylacetonate (0.37 g , 3.0 mmol) was added to a red solut ion prepared according to Method A from Ho (0.71 g , 4.3 mmol) and Mn(C0)5Br (0.66 g , 2.4 mmol) in THF (30 ml) . The mixture was s t i r r e d at room temperature for 2 days and f i l t e r e d to y i e l d a golden brown solut ion which was taken to dryness under reduced pressure. The golden brown residue was extracted i n a Soxhlet apparatus with re f lux ing die thyl ether (100 ml) for 4 h. The ether solut ion was reduced i n volume to y i e l d l i g h t brown crys ta l s which were col lected by f i l t r a t i o n and washed with a small volume of pentane. Y i e l d 0.12 g , 14%; infrared (Nujo l ) , v m 3 V 2060(s), 2010(vs), 1975(s), 1925(s), 1910(s), 1620(sh), 1595(s), 1520(s), 1380(s), 1265(m), 1020(m), 930(w), 800(w) and 650(s) c m " 1 ; infrared (CH 2 C1 2 ) , v m a v 2055(s), Mia X 2005(s), 1970(s), 1925(s), 1620(sh) and 1595(s) c m " 1 . A n a l . Calc for C 2 3 H 2 7 0 l l t H o M n : C, 34.42; H, 3.37. Found: C, 34.42; H, 2.15. The acetylacetonate complex [m.p. 78°C (decomp)] turns yel low upon exposure to a i r f o r a few hours, and M n 2 ( C 0 ) 1 0 can be sublimed from the o x i -dized product. Reaction of the red solut ion with triphenylphosphine A solut ion of ( C 6 H 5 ) 3 P (3.9 g , 14.9 mmol) i n THF (20 ml) was added dropwise at room temperature to a s t i r r e d red so lut ion prepared according to Method D from NaMn(C0)5B(0.88 g , 4.0 mmol) and MnCl 2 (0.51 g , 4.0 mmol) i n THF (15 ml) and d i lu ted with THF (40 ml) . Af ter s t i r r i n g from 20 minutes, the reaction mixture was taken to dryness under reduced pressure. The r e s i -74 dual s o l i d was extracted with hexane (5 x 20 m l ) , the combined extracts were f i l t e r e d , and the resu l t ing c lear orange solut ion was reduced in volume under reduced pressure to 50 ml whereupon c r y s t a l l i z a t i o n occurred. The orange crys ta l s were col lec ted by f i l t r a t i o n , washed with a small volume of hexane, and dried i n vacuo (10~3 mm Hg). The melting point and infrared spectrum of the c rys ta l s were found to be ident ica l with those of M n 2 ( C 0 ) 8 [ P ( C 6 H 5 ) 3 ] 2 . 1 2 0 A n a l . Calc for C 1 + t t H 3 o 0 8 P 2 M n z : C, 61.56; H, 3.52. Found: C, 61.65; H, 3.79. Further evidence for the dimeric (rather than the monomeric) nature of the product i s i t s n .m.r . spectrum in CH 2 C1 2 which exhibi ts no broadening or s h i f t i n g of the solvent peak, behaviour i n d i c a t i v e of a diamagnetic species. Reaction of the red solut ion with 2 , 2 ' - b i p y r i d y l A solut ion of 2 , 2 ' - b i p y r i d y l (2.00 g , 12.8 mmol) in THF (10 ml) was added to a s t i r r e d red solut ion prepared according to Method D from MnCl 2 (0.50 g , 4.0 mmol) and NaMn(C0)5 (0.88 g , 4.0 mmol) in THF (50 ml ) . This addit ion and subsequent manipulations of the reaction mixture were carr ied out i n the absence of d i r e c t l i g h t . If Method A or B with Mn as the metal was used to prepare the red s o l u t i o n , MnBr 2-2 bipy ( i d e n t i f i e d by elemental analys is ) prec ipi ta ted i n s t a n t l y upon the addit ion of the 2 , 2 ' -b i p y r i d y l . If Method D was employed, however, no prec ip i ta te formed. Iin a l l three cases, the so lut ion became deep purple a f ter s t i r r i n g at room temperature overnight. The purple so lut ion was f i l t e r e d , i f necessary, and the solvent was removed under reduced pressure. The residue was extracted 75 with benzene (6 x 15 m l ) , and the combined extracts were f i l t e r e d and con-centrated u n t i l c r y s t a l l i z a t i o n occurred. The purple-red crys ta ls were col lec ted by f i l t r a t i o n and were washed with hexane (3 x 10 ml ) . Y i e l d 0.43 g , 44%. The infrared spectrum and elemental analysis of the i so la ted product confirmed the formulation M n 2 ( C 0 ) 8 ( b i p y ) J 1 5 A n a l . Calc for C 1 8 H 8 0 8 N 2 M n 2 : C, 44.11; H, 1.64; N, 5.71. Found: C, 44.84; R, 1.63; N , 5.48. Reaction of the red solut ion with N-meth,yl-N-nitroso-p-toluenesulphonamide  (Diazald) A solut ion of Diazald (0.40 g , 0.20 mmol) in THF (4 ml) was added dropwise at 0°C to a s t i r r e d red solut ion which was prepared according to Method A from Sm (0.50 g , 0.30 mmol) and Mn(C0)5Br (0.06 g , 0.20 mmol) i n THF (5 ml) . The solut ion turned brown wi th in 10 minutes and infrared spec-troscopy revealed the appearance and growth of an absorption at 1752 c m " 1 , 121 charac ter i s t i c of Mn(C0)itN0. A comparison of i n t e n s i t i e s of the i n i t i a l and f i n a l Diazald bands i n the infrared spectrum suggested a y i e l d of ca. 40% of Mn(C0KN0. Reaction of the red solut ion with iodine S o l i d iodine (0.25 g , 2.0 mmol) was added at room temperature to a s t i r r e d red so lut ion prepared according to Method A from Mn(C0)5Br (0.57 g , 2.0 mmol) and Sm (0.45 g , 3.0 mmol) i n THF (12 ml) . An immediate color change to brown occurred and a prec ip i ta te formed. The mixture was s t i r r e d overnight and the solvent was removed under reduced pressure. The residual s o l i d was extracted with benzene (3 x 20 m l ) , and the f i l t e r e d extracts were 76 taken to dryness. The s o l i d thus obtained was f r a c t i o n a l l y sublimed: at room temperature and 10" 3 mm Hg, yellow and orange crysta ls formed; at 90°C and 10" 3 mm Hg, deep red-brown crys ta ls formed. The former sublimate was shown to be a mixture of Mn 2 (C0) 1 0 and a small quantity of Mn(C0.)5I by 12g infrared spectroscopy and a pos i t ive P d C l 2 spot tes t for iodine . The higher temperature sublimate had an infrared spectrum i d e n t i c a l with that 122 of M n 2 ( C 0 ) 8 I 2 and also gave a pos i t ive iodine spot t e s t . Y i e l d 40 mg, 6%. Attempts to react the red solut ion with 1-butyne e i ther by conden-o sing the alkyne onto the cooled (-30°C) red solut ion and al lowing the mixture to warm to room temperature, or by bubbling the gas through a red solut ion at 25°C were unsuccessful. S i m i l a r l y , no reaction was apparent when CO or H 2 were bubbled through the red solut ion at room temperature; neither did cyclopentadiene give any detectable change i n the red solut ion when heated to 60°C. A l l of these attempted reactions were monitored by infrared spec-troscopy. If the red solut ion i s taken to dryness and N,N-dimethylacetamide i s introduced as the solvent , gas evolution can be detected, p a r t i c u l a r l y upon warming the solut ion to 70°C. 77 3.2b REACTIONS INVOLVING ALLYTRICARBONYLIRON IODIDE Preparation of the mauve solut ion from (h 3-C3Hs)Fe(C0)3I and a metal Yb powder (0.47 g , 2.7 mmol) was vigorously s t i r r e d i n a THF (17 ml) so lut ion of (h^-C 3 H 5 )Fe(C0) 3 I (0.62 g , 2.0 mmol) at room temperature u n t i l the solut ion turned mauve, normally about 1.5 - 2 h. The mauve so lu-t ion was f i l t e r e d and used immediately. I f the solut ion was l e f t to s t i r a f ter i t had achieved the mauve c o l o r , i t turned brown within 1 h and af ter about 12 h i t had turned black. The black so lut ion was f i l t e r e d to y i e l d an intrac table black s o l i d and a c lear black solut ion which, when taken to dryness by solvent evaporation under reduced pressure, did not redissolve i n common organic solvents . Infrared (black s o l i d from black solut ion taken to dryness) , (Nujo l ) , v m a v : 2010(sh), 1965(br,s) , 1030(s) and 875(s) c m ' 1 . Samarium, manganese and y t t r i u m , reacted s i m i l a r l y . The reaction of a l l y l t r i c a r b o n y l i r o n iodide with deactivated alumina (a) In benzene: (Jh 3 -C 3 H 5 )Fe(C0) 3 I (0.21 g , 0.7 mmol) was s t i r r e d with deac-t ivated alumina (10 g) i n benzene (10 ml) for 1 min, by which time the color was mauve. An infrared spectrum was recorded of t h i s s o l u t i o n , the remain-ing mauve solut ion was d i lu ted (benzene, 10 ml) and the infrared spectrum was recorded once more. A further d i l u t i o n (benzene, 20 ml) was performed and the mixture was allowed to stand u n t i l i t had turned green (35 mins); a f i n a l in f rared spectrum was recorded (see Figure 3-3). (b) i n THF: the above procedure was repeated using THF as the solvent. The same color changes were obtained wi th in the same time periods; the reaction 78 was monitored by infrared spectroscopy (see Figure 3-3). A mauve s o l u t i o n , prepared from (h_ 3-C 3H 5)Fe(C0) 3 I (0.41 g , 1.3 mmol) and deactivated alumina 20 g i n THF (30 ml) was treated with dropwise addit ion of Diazald (0.29 g , 1.3 mmol) i n THF (6 ml) . Immediately the so lu-t ion turned brown and an infrared spectrum of the reaction mixture showed the fo l lowing bands, v m a v 2040(s), 1977(s), and 1756(s) cm" 1 . Reaction of the mauve solut ion with Diazald Diazald (0.40 g , 1.9 mmol) i n THF (8 ml) was added dropwise at 0°C, to a s t i r r e d , l ight -protec ted mauve solut ion which was prepared accord-ing to the above procedure from Yb (0.45 g , 2.6 mmol), and (h_ 3-C 3H 5)Fe(C0) 3I (0.61 g , 2.0 rnrnol) in THF (16 ml ) . Af ter s t i r r i n g 5 min, the solut ion was concentrated to ca_. 5 ml by solvent removal under reduced pressure. The remaining red solut ion was transferred to a neutral alumina chromatography column, which was eluted with pentane. The f i r s t red f r a c t i o n was col lec ted and separated by trap to trap d i s t i l l a t i o n (15°C, -35°C, -78°C) , under vacuum. A red solut ion co l lec ted in the -35°C t r a p , upon warming to room 12 temperature the s o l i d melted to a red o i l . Infrared and n.m.r . measurements showed the o i l to be (h_ 3-C 3H 5)Fe(C0) 2N0. N.m.r . ( C 6 H 6 ) , T 6.88 (d , J = 11 Hz, 2 H) , 6.02 (d , J = 5.5 Hz, 2 H) and 5.62 (m, 1 H). Infrared (pentane), v m a x 2040, 1977 and 1756 cm" 1 i n the CO and NO region. Y i e l d : 0.16 g (37%). Reaction of the mauve so lut ion with iodine . S o l i d iodine (0.14 g, 0.6 mmol) was added to a 5 ml al iquote of mauve solut ion prepared from Yb (0.47 g , 2.7 mmol) and (h_ 3-C 3H 5)Fe(C0) 3 I 79 (0.62 g , 2.0 mmol) i n THF (17 ml) . Immediately the so lut ion turned brown. The infrared spectrum i n the carbonyl region displayed only the bands at 2090s, 2040vs, and 2020sh c m " 1 , c h a r a c t e r i s t i c of ( h 3 - C 3 H 5 ) F e ( C 0 ) 3 I . 1 2 4 Reaction of the mauve solut ion with triphenylphosphine Ph 3P (2.60 g , 9.9 mmol) i n THF (12 ml) was added dropwise to a s t i r r e d mauve solut ion prepared from Yb (0.53 g , 3.1 mmol) and (Jx 3-C 3H 5)Fe(C0) 3I (0.60 g , 1.9 mmol) i n THF (17 ml) by the method described above. Immediately the solut ion turned green-brown; a f ter s t i r r i n g at room temperature for 15 mins the solut ion was f i l t e r e d . The solvent was removed from the f i l t r a t e under reduced pressure to y i e l d a green s o l i d . The s o l i d was r e c r y s t a l l i z e d from THF-hexane and then from toluene at -78°C. The green crys ta l s were washed with hexane (5 ml) and were dried at room temperature under vacuum (10" 2 mm Hg). Y i e l d 0.73 g. The a i r - s e n s i t i v e green s o l i d i s s l i g h t l y soluble i n CCl^ and CH 3CN, but f a i r l y soluble i n benzene and C H 2 C 1 2 ; infrared (CH 2 C1 2 ) , v m 3 V 1960s and 1885s c m " 1 ; (Nujol) v m a v 1955s, 1875s(br,vs), 1440s, 1095m, 750m, max 710m, 700s, and 640m. N.m.r. ( C 6 D 6 ) , x 2.82(m).m.pt.: 89°C (dec). 5 7 F e Mossbauer (Figure 3-4): ( s o l i d at 77°K): Q . S . ^ , 2.64 mms"1; I .S . l l t , -0.09 mms"1; Q . S . 2 3 , 0.59 mms"1; I . S . 2 3 , +0.12 mms"1. I .S . quoted r e l a t i v e to Fe f o i l . Anal : C, 60.13; H, 4.16; I , 11.27. Reactions using cyclopentadienyldicarbonyliron halide (h 5 -Cp)Fe(C0) 2 I (1.54 g , 5.1 mmol) and Yb (1.05 g , 6.1 mmol) i n THF (70 ml) were s t i r r e d at -5°C for 1 h , then allowed to warm to room temper-80 ature. So l id Ph 3SnCl (1.8 g , 4.7 mmol) was added to the f i l t e r e d s o l u t i o n , and s t i r r i n g was continued for 1.5 h . The mixture was hydrolyzed with 50 ml aqueous NH^Cl so lut ion and was washed with aqueous NaCl solut ion (2 x 25 ml) . The THF layer was dried over anhydrous MgSOit and the solvent was removed under reduced pressure. The remaining s o l i d was extracted with CH 2 C1 2 (3 x 25 ml ) , f i l t e r e d and hexane added to induce c r y s t a l l i z a t i o n . The dark mauve crys ta l s were r e c r y s t a l l i z e d from CH 2 Cl 2 /hexane. The i n f r a -red spectrum and m.pt, (191 - 192, l i t 194°C) were ident i ca l to those of [ ( h . 5 - C p ) F e ( C 0 ) 2 ] 2 . 1 2 5 Y i e l d 0.61 g (68%). S imi lar resul ts were obtained using (h[ 5-Cp)Fe(C0) 2Cl and ytterbium. Other attempted reactions N i t r o s y l a t i o n of the mauve solut ion using a dropwise addit ion of N0C1 i n THF was unsuccessful. Although the solut ion turned brown, the i n f r a -red spectrum revealed only the presence of (h^-C 3 H 5 )Fe(C0) 3 I . Several attempts were made to react Ph 3SnCl with the mauve s o l u t i o n ; i n each case a small quantity of a brown tarry s o l i d was obtained, [infrared (hexane) v m a x 2050s, 1978s, and 1960s c m " 1 ; n .m.r . (CCl^ ) , T 2.57 (m, 15 H) , 7.31 (d , J =5 .0 Hz, 2 H)-and 7.90 (d , J = 12.2 Hz, 2 H)]and much unreacted Ph 3 SnCl. 81 3.2c REACTIONS USING CYCLOPENTADIENYLTRICARBONYLMOLYBDENUM DIMER AND  YTTERBIUM METAL (a) In the presence of 1,2-dibromoethane Yb (0.42 g , 2.4 mmol) was s t i r r e d i n a re f lux ing THF (30 ml) so lut ion of [(h_ 5-Cp)Mo(C0) 3] 2 (0.700 g , 1.4 mmol) and (CH 2 Br) 2 (0.25 m l , 2.7 mmol) for 26 h. The solut ion turned red-brown and an infrared spectrum of the f i l t e r e d solut ion was recorded. The red-brown solut ion was treated with the dropwise addit ion of Diazald (1.0 g, 4.7 mmol) i n THF (10 ml) at room temperature. Af ter s t i r -r ing for 30 mins, the solvent was removed under reduced pressure and the remaining s o l i d was sublimed at 50°C (10" 2 mm Hg). Orange crys ta l s c o l l e c -ted . The infrared spectrum and m.pt. (86 - 87°Ci, l i t 85.2 - 85.7°C) con-firmed the formulation (h. 5 -Cp)Mo(C0) 2 N0. 7 8 Y i e l d 0.26 g (37%). A s i m i l a r l y prepared solut ion from Yb, (CH 2 Br) 2 and [(h_ 5-Cp)Mo(C0) 3] 2 in THF did not react with Ph 3SnCl at 60°C a f ter 1 h , as evidenced by infrared spectroscopy. (b) In the absence of 1,2-dibromoethane Y/b (0.54 g , 3.1 mmol) was s t i r r e d i n a re f lux ing THF (30 ml) solut ion of [(h_ 5-Cp)Mo(C0) 3] 2 (0.960 g , 2.0 mmol) for 20 h. The solut ion turned brown, i t was f i l t e r e d and the infrared spectrum was recorded. The brown solut ion was treated as fo l lows : ( i ) addit ion of iodine S o l i d iodine (0.25 g , 1.0 mmol) was added to 5 ml of the above s o l u t i o n . After s t i r r i n g at room temperature for 20 mins the solvent was 82 removed under reduced pressure. The remaining s o l i d was twice r e c r y s t a l l i z e d from C H 2 C l 2 / C 6 H 1 i t to y i e l d a ruby red s o l i d , whose infrared and n.m.r . 70 spectra confirmed the assignment (h_5-Cp)Mo(C0)3 I . Y i e l d 0.28 g (26%). ( i i ) addit ion of H 20 Introduction of H 20 (0.1 m l , 5.6 mmol) to the remainder of the above solut ion (8 ml) caused a dark brown j e l l y to form immediately. The solvent was removed under reduced pressure and the remaining s o l i d was sublimed at 60°C (10~2 mm Hg). Light yellow crys ta l s c o l l e c t e d . The infrared and n.m.r . spectra and m.pt. (51 - 52°C, l i t 50 - 52°C) were found to be ident i ca l to those of (h 5 -Cp)Mo(C0) 3 H. 7 8 Y ie ld 0.21 g (36%). 3.2d REACTIONS INVOLVING MERCURY COMPOUNDS AND YTTERBIUM METAL 83 ( i ) (Cyclopentadienyltricarbonylchromium)mercury chloride and ytterbium [ (h 5 -Cp)Cr(C0) 3 ] 2 Hg (0.51 g , 0.8 mmol) and HgCl 2 (0.22 g , 0.8 mmol) were dissolved i n THF (20 ml) and s t i r r e d for 30 min before the addi-t i o n of Yb powder (0.39 g , 2.3 mmol)., Immediately the solut ion turned green; wi th in 10 min i t was pink-brown and a grey deposit had formed. The mixture was f i l t e r e d and the solvent was removed under reduced pressure to y i e l d 0.6 g of pink-green s o l i d . The f i l t e r e d s o l i d contained v i s i b l e 126 globules of a meta l l i c l i q u i d and gave a pos i t ive mercury spot test a f ter digest ion i n c .HN0 3 . A portion of the dried f i l t r a t e was sublimed at 110°C 10" 2 mm Hg to y i e l d a small quantity of green-blue s o l i d , whose infrared spectrum and m.pt. (164 dec, l i t 163-168°C dec) were i d e n t i c a l to those of 127 [ ( J l 5 -Cp)Cr (C0) 3 ] 2 . The remaining dried f i l t r a t e was redissolved i n THF (15 ml) and was treated with 0.37 g (1.0 mmol) s o l i d Ph 3 SnCl . The mixture was heated to 55°C for 1 h ; the solvent was removed under reduced pressure to y i e l d a s o l i d which was extracted with CCl^ (5 x 10 ml) and f i l t e r e d . Crystals formed as the solvent was removed slowly under reduced pressure. The m.pt. (215 - 217°C, l i t 219 - 221°C) and infrared and n.m.r . spectra 14? confirmed the assignment (h_ 5-Cp)Cr(C0) 3SnPh 3 . Y i e l d 0.16 g . ( i i ) Phenylmercuric chloride and ytterbium PhHgCl (0.65 g , 2.1 mmol) and Yb (0.52 g , 3.0 mmol) were s t i r r e d i n THF for 2 days at room temperature. An a i r - s t a b l e grey s o l i d was deposited and the mixture was f i l t e r e d . The solvent was removed under reduced pressure from the f i l t r a t e to y i e l d an a i r - s t a b l e white s o l i d . The grey s o l i d gave 84 pos i t ive mercury and chlor ide spot tests and could not be extracted with C H 2 C 1 2 . The white s o l i d was extracted with HCC13 (2 x 20 m l ) , these extracts were washed with water (2 x 20 ml ) , and the HCC13 was allowed to evaporate. Thin layer chromatography ( s i l i c a gel) developed with 10% benzene-hexane) showed the resu l t ing white s o l i d to contain only one component. The white s o l i d was sublimed at 120°C 1 0 _ 1 mm Hg. The white sublimate was found to 129 have an m.pt. (123 - 4, l i t 125°C), infrared and n .m.r . spectra consistent with the formulation Ph 2Hg. Furthermore, the sublimate showed a pos i t ive mercury spot test and a negative halide tes t . Y i e l d 0.3 g (41%). 85 3.3 RESULTS 3.3a REACTIONS INVOLVING MANGANESE CARBONYLS (1) Methods of preparing and factors a f fec t ing formation of the red solut ion  Method A: Metal + Mn(C0)sBr We have found that cer ta in lanthanide metals (Y, P r , Sm, Dy, Ho, Er and Yb) and manganese react with Mn(C0)5Br i n THF under ambient conditions to produce red , a i r - and moisture-sensi t ive so lut ions . Those lanthanides which are known to possess a stable +2 oxidation s ta te , ytterbium and samarium, react more rap id ly than the other metals. [In a l l p r o b a b i l i t y , europium would react s i m i l a r l y or faster than ytterbium, but th i s metal i s rather expensive and was not used.] Common to a l l the metals, however, i s the increased r e a c t i v i t y associated with a smaller p a r t i c l e s ize and the need for good agi ta t ion of the reaction mixture. For the less react ive metals, careful f i l t r a t i o n of the solut ion (as described in the experimental sec t ion , 3.2a) a f ter one day's s t i r r i n g i s b e n e f i c i a l . Some dark, a i r -unstable s o l i d i s removed by th i s process, the metal surface becomes more reac t ive , and the reaction proceeds smoothly. Iodine, when added i n trace amounts, i s not an e f fec t ive i n i t i a t i n g agent, but concentration of the solutions by solvent removal under reduced pressure, i s often helpful to avoid a sluggish react ion. Other than the metals mentioned previous ly , lanthanum, gadolinium, t i n and i ron show some reaction with Mn(C0) 5Br i n THF at room temperature. Although the red solut ion does not develop, an infrared spectrum of the carbonyl-stretching region i s ident i ca l for a l l four metals. In addit ion to bands which can be a t t r ibuted to Mn(C0) 5Br, the spectrum exhibi ts two lower 86 frequency absorptions at 1964m and 1930m cm 1 . The lack of conversion to the red solut ion may be a t t r ibuted to p a r t i c l e s ize i n the cases of lanthanum and gadolinium which were only avai lable as ingots . Despite the experimental d i f f i c u l t i e s associated with th i s synthetic method, i t i s reproducible for the indicated metals, and since the reaction conditions involved are quite m i l d , i t i s the method of choice for a general preparation of the red s o l u t i o n . The other routes to t h i s react ive solut ion have been less thoroughly invest igated; they are described below. Method B: Metal + Mn2(CO)io + (CH 2 Br) 2 Both manganese and ytterbium produce the red so lut ion when s t i r r e d i n a re f lux ing THF solut ion of M n 2 ( C 0 ) 1 0 and (CH 2 Br) 2 for at least 16 h. Method C: MnC12 + Mn2(CO)io The red solut ion can be obtained by re f lux ing a THF solut ion of MnCl 2 and M n 2 ( C 0 ) 1 0 for three days. The extent of conversion, however, i s not great. In contrast , Y b C l 3 does not react with M n 2 ( C 0 ) 1 0 , as evidenced by infrared spectroscopy, even a f ter four days of s t i r r i n g in a re f lux ing THF s o l u t i o n . Method D: MnC12 + NaMn(C0)s Af ter s t i r r i n g a 1:1 mixture of MnCl 2 and NaMn(C0)5 i n THF over-n ight , a s o l i d precipi tates and the solut ion turns red. This solut ion exhibi ts the c h a r a c t e r i s t i c bands in the carbonyl region of i t s infrared spectrum (see below). However, th is spectrum also reveals that the so lut ion generated i n th is manner contains a greater proportion of anionic (e .g . [Mn(C0) 5]~) 87 components than those prepared by the other synthetic methods. S t i r r i n g the so lut ion for a longer period of time causes an in tens i ty decrease of the infrared bands a t t r ibutable to the anions and an increase of some of the higher frequency absorptions. I f , on the other hand, the o r i g i n a l l y precip-i ta ted s o l i d i s removed, the remaining red solut ion deposits more s o l i d wi th in a matter of hours and the higher frequency bands increase i n i n t e n s i t y at the expense of the lower frequency absorptions i n the infrared spectrum. S imi lar spectral changes occur when the o r i g i n a l red solut ion i s d i l u t e d (Figure 3-1). (2) Properties of the red solut ion (a) Infrared spectrum The fo l lowing seven bands are observed i n the carbonyl region of the infrared spectrum regardless of the mode of preparation of the red s o l u -t i o n : ('I) 2040, (II) 2005, ( I I I ) 1965, (IV) 1930, (V) 1885, (VI) 1850 and (VII) 1820 cm" 1 . The r e l a t i v e i n t e n s i t i e s of the seven bands r e f l e c t the preparative route employed. For example, bands (V) and ( V I ) , which are also observed i n the infrared spectrum of NaMh'(C0)5, are i n i t i a l l y quite intense for a red so lut ion prepared by Method D, but they diminish i n i n t e n s i t y with time and d i l u t i o n (see Figure 3-1) as the simultaneous growth i n bands ( I ) , ( I I I ) and (IV) occurs. Band (II) i s generally s trong, whereas band (VII) i s usual ly of medium i n t e n s i t y . (b) U l t r a v i o l e t spectrum A concentrated red so lut ion prepared with Yb shows a weak band at 13 10,250 cm" 1 whdich can be a t t r ibuted to a 2 F 7 y 2 ^ 2 f : 5 / 2 t r a n s i t i o n and i s 88 i n d i c a t i v e of at least a part of the lanthanide being in a +3 oxidation s tate . (c) Ionic d issoc ia t ion i n so lut ion E l e c t r i c a l conduct ivi ty measurements i n THF at 25°C indicate that the red solutions are e l e c t r i c a l l y conducting. Tables VI and VII show that the conduct ivi ty of the red s o l u t i o n , whatever i t s mode of preparation, i s subs tant ia l ly greater than Mn(C0)5Br or M n 2 ( C 0 ) 1 0 . The extent of ion iza t ion i s apparently not great since attempts to i so la te any carbonyl anions with (Ph 3 P) 2 NCl are unsuccessful. (d) Electron spin resonance spectrum Preliminary e . s . r . data of the red solut ion derived from manganese are consistent with the presence of M n 2 + ( i e . the expected s i x - l i n e spectrum is observed). However, as out l ined above, d i l u t i o n appears to a f fec t the composition of the s o l u t i o n , and consequently care must be taken i n i n t e r -preting these r e s u l t s . (e) The s o l i d remaining a f te r removal of THF A red-brown, a i r - and moisture-sensi t ive s o l i d can be obtained by taking the red solut ion to dryness under reduced pressure and allowing the mixture to cool i n the process. I f Method A i s used for the preparation of the red s o l u t i o n , the weight of the s o l i d thus i so la ted i s greater than that of the o r i g i n a l Mn(C0)5Br (eg. by a factor of ^ 1.4 for samarium). Sublimation of the red-brown s o l i d at 10" 3 mm Hg and room temperature causes some M n 2 ( C 0 ) 1 0 to c o l l e c t on the water-cooled probe. The s o l i d which does not sublime exhibi ts a pos i t ive halide test and upon exposure to a i r i t slowly Figure 3-1 Dilution of the red solution. 2000 1800 i , i 2000 1800 — I , I f (a) The red solution prepared by method D after 2 days (b) Diluted 1 0 0 % 2000 1800 -// 1 • I r •(c) Diluted a further 50 % VA 2000 18,00 2000 1800(crrfi) 1 / / L _ ^ _ J (d) Diluted a further 67 % =solution A r \ (e) Solution A left 24 hours The solutions were stirred for a few hours after dilution of stages (b),(c) and (d), before the i.r. spectra were recorded. 00 WD TABLE VI Conductivity Results for Manganese Carbonyl Reactions METHOD OF PREPARATION3 METAL USED INITIAL MOLAR (x lO" 2 ) CONCN. "Mn(C0) 5" MEASURED RESISTANCE R, a x 10 3 CALCULATED CONDUCTIVITY K=K/R , ohm" 1 cm" 1 xlO- 5 b TEMPERATURE °C A Mn 4.4 34.0 2.98 24 A Sm 9.2 2.88 35.2 23 A Sm 3.1 9.62 10.5 23 A Sm 1.0 22.0 4.62 23 A Sm 0.3 49.2 2.06 23 B Mn 4.6 23.0 4.42 24 D - 4.5 2.90 35.0 25 a: see section!3.2a b: K = 1.015 cm" 1 TABLE VII Conductivity Measurements of Start ing Reagents and Solvent MOLAR ( x l O - 2 ) CONCN. MEASURED RESISTANCE CALCULATED CONDUCTIVITY TEMPERATURE OF "Mn(C0) 5" R, ohm x J O 3 K=K/R , ohnwlerrfix IO" 5 °C Mn(C0)5Br 10.0 5.0 X 10 2 2.03 X IO" 1 25 4.1 6.0 X IO 2 1.69 X IO" 1 25 1.7 11.2 X IO 2 9.07 X IO* 2 25 Mn 2 (CO) 1 0 19.8 > 25 X IO 2 « 4 X I O ' 2 25 7.8 > 25 X IO 2 < 4 X I O " 2 25 NaMn(C0)5 9.9 7.58 X IO" 1 1.34 X IO 2 24 6.6 12.25 X IO" 1 8.28 X 10 25 4.4 20.01 X IO" 1 5.0 X 10 24 1.5 7.28 1.40 X 10 25 0.5 2.47 X 10 4.11 24 THF Neat > 25 X IO 2 4 X IO" 2 24 92 turns yel low. More M n 2 ( C 0 ) 1 0 can then be sublimed from th i s oxidized species. Metal analyses of the red-brown s o l i d derived from red solut ions prepared from lanthanide metals and Mn(C0)5Br have so far been obstructed by mutual cross- interference of the metals during the a n a l y t i c a l process. At room temperature, the red-brown s o l i d i s soluble in acetone, THF and C H 2 C 1 2 , although i t decomposes within an hour i n the l a s t solvent . In N,N-dimethyl-acetamide, however, the s o l i d dissolves upon warming with concomitant gas evolut ion . (3) Reactions of the red solut ion (a) Halide displacement reactions Treatment of the red solut ion with Ph 3SnCl at 50°C for one h affords Ph3SnMn(C0)5 i n up to 60% y i e l d . In a s i m i l a r manner, a l l y l chlor ide reacts smoothly with the red solut ion at room temperature^overnight to produce a mixture of (M-C 3H5)Mn(C0)5 and (h. 3 -C 3 H 5 )Mn(C0)i t i n a r a t i o of 2 .4 :1 . >•; Methylation of the red solut ion by Mel , on the other hand, proceeds slowly and incompletely to give a low y i e l d 10%) of MeMn(C0)5 even af ter re f lux ing for 24 h . A der ivat ive containing the lanthanide metal may be isolated i f the red s o l u t i o n , prepared using holmium, i s treated with sodium acetylacetonate for two days at room temperature. Af ter appropriate work up of the reaction mixture, a brown, a i r - s e n s i t i v e s o l i d , whose elemental analysis i s consistent with the formulation [Mn(CO) 5 ] 2 Ho(C 5 H 7 0 2 ) -2Et 2 0 i s obtained. The mull i n f r a -red spectrum of th i s s o l i d exhibi ts both metal carbonyl absorptions and the bands charac ter i s t i c of the acetylacetonate l i g a n d . 93 (b) Carbonyl subst i tut ion reactions Addit ion of good Lewis bases such as 2 , 2 ' - b i p y r i d y l and t r i p h e n y l -phosphine to the red solut ion at room temperature and in the absence of l i g h t produces f a i r y i e l d s of the substituted manganese carbonyls [Mn(C0)j+L]2 (L = PPh 3 or l / 2 ( b i p y ) ) . [ I f manganese i s used to prepare the red solut ion v i a Method A or B, MnBr 2-2 bipy can be i so la ted as a by-product of the bipy reac t ion . ] (c) Other reactions The red solut ion reacts smoothly with Diazald at 0°C to produce Mn(C0)itN0 i n f a i r y i e l d as evidenced by infared spectroscopy. Reaction with benzoyl chlor ide at 65°C for two h does not lead to the expected acyl der iva t ive of manganese pentacarbonyl, but cleavage of the THF solvent occurs instead. The major organic product of t h i s l a t t e r reaction i s C 6 H 5 C0 2 ( C H 2 r ) i + C l (48% y i e l d ) and the pr inc ipa l metal carbonyl ent i ty i s M n 2 ( C 0 ) 1 0 (55% y i e l d ) . Iodine reacts with the red solut ion to give low y i e l d s (^ 5%) of Mn(C0) 5I and [Mn (C0 ) t + I ] 2 . The red solut ion does not react with CO, H 2 S C 5 H 6 or simple o l e f i n s under ambient condit ions . 94 3.3b THE MAUVE SOLUTION FROM ALLYTRICARBONYLIRON IODIDE AND A METAL (1) Preparation An a i r - and moisture-sensit ive mauve solut ion can be readi ly pro-duced at room temperature wi th in two hours by vigorously s t i r r i n g ytterbium, y t t r i u m , samarium, or manganese with (h_ 3-C 3H 5)Fe(C0) 3I in THF. Other metals have not been invest igated. The mauve solut ion must be used as soon as the infrared spectrum indicates complete disappearance of the parent carbonyl (see below). I f s t i r r i n g i s continued, the solut ion turns black eventually and a large quantity of prec ip i ta te forms. The black supernatant l i q u i d s t i l l contains some metal carbonyl e n t i t i e s . However, upon solvent removal the small amount of s o l i d that remains does not redissolve completely i n common organic solvents . (2) Infrared measurements The formation of the mauve solut ion canbbe monitored by infrared spectroscopy (Figure 3-2). A THF solut ion of the parent carbonyl , ( h . 3 - c 3 H 5 ) F e ( C 0 ) 3 I , displays only two bands i n the carbonyl stretching region of the infrared spectrum: 2085s and 2030br,s cm" 1 . As the reaction proceeds the higher band diminishes, u l t imate ly to zero absorbance when the solut ion i s mauve and ready to be used. The lower band at 2030 cm" 1 broadens and develops a shoulder at 1995 cm" 1 . Separation of these absorptions f i n a l l y occurs to give two bands at 2050s and 1995br,s c m " 1 . Meanwhile, a new peak develops at 1950 c m " 1 . The f i n a l mauve solut ion has the fol lowing bands: 2050s, 1995br,s and 1950vs cm" 1 . Upon d i l u t i o n these peaks change as follows (Figure 3*3): the 1950 band weakens; the 1995 absorption s p l i t s into one at Figure 3 - 2 F o r m a t i o n o f t h e m a u v e s o l u t i o n 2 0 0 0 1800 i . i ^ r (a) 2 0 0 0 1800 i . i (b) (h -C, H ) Fe (COLI Solution 1 + Yb in T H F = solution 1 a f t e r 5 0 mins. 2 0 0 0 1800 I (c) after 7 8 mins. 2 0 0 0 1800 (cm" 1 ) i i ^ r after 136 mins. (mauve) Figure 3 - 3 Dilution of the mauve solution 2000 1800 l . i VA 2000 1800 l i i VA I. Prepared with Ytterbium in THF 2000 1800 (cm-') J L . (a) Mauve solution (b) Mauve solution diluted 1 0 0 % (c) Diluted a further 40 % 2000 1800 2000 1800 i L _ J ^ I I I I. Prepared with deactivated Alumina in Benzene 2000 1800 2000 1800 (cm~i) - / / • — ' / / ' 1 ' ^ r \ (a) (h 3-C 3H 5)Fe(CO) 3I in Benzene=solution 1 (b) Solution 1 + Alumina after 5 mins, mauve (cJ (d) Mauve solution Mauve solution diluted I00%(5min) diluted 100%(70min) 2000 1800 1 1 1 //-2000 1800 i i i III. Prepared from deactivated Alumina in THF 2000 1800 (cm-i) i , I (a) (h-C 3 H 5 )Fe(CO) 3 I + Alumina after 5 min= Solution 1 r ~ (b) Solution 1 diluted 100% (10 min) (c) Diluted a further 50 % , (50 min! 97 1995, which increases in i n t e n s i t y r e l a t i v e l y , and one at 2018 c m " 1 ; the 2050 peak remains constant as the weakest band. The f i n a l so lut ion has the f o l l o w -ing bands: 2050m, 2015m-s, 1995s and 1950s cm" 1 . The formation of the radical-dimer e q u i l i b r i u m , 2 ( h 3 - C 3 H 5 ) F e ( C 0 ) 3 = = = [ ( h . 3 - C 3 H 5 ) F e ( C 0 ) 3 ] 2 by dehalogenation of (Jh 3-C 3H 5)Fe(C0) 3I using deactivated alumina has been 130 described previously . Using THF and benzene as solvents and the method 130 above i t i s possible to study the e f fec t of d i l u t i o n on the infrared spectrum as out l ined e a r l i e r (Figure 3-3). Regardless of the solvent , the c h a r a c t e r i s t i c mauve colour develops within 5 mins and the infrared spectrum has three bands at 2050s, 1995br,s and 1950vs cm" 1 . D i l u t i o n causes s i m i l a r behaviour to that noted for the mauve solut ion prepared using ytterbium (Figure 3-3) , namely diminishing in tens i ty of the 1950 cm" 1 band and s p l i t t i n g of that at 1995 cm" 1 . I f the d i l u t e so lut ion i s allowed to stand over the alumina for over one hour, the only remaining bands are those at 1995vs and 2015 cm" 1 and the solut ion i s a pale green colour. Further, addit ion of Diazald to the mauve so lut ion prepared from alumina and (h_ 3 -C 3 H 5 )Fe(C0) 3 I causes an immediate colour change; the resul tant so lut ion has an infrared spectrum ident i ca l to that of (h_ 3-C 3H 5)Fe(C0)2N0. (3) Reactions The mauve solut ion reacts smoothly under ambient conditions with Diazald to form (h, 3 -C 3 H5)Fe(C0) 2 N0 i n 37% y i e l d . Another feas ib le route to t h i s compound, the action of N0C1 upon the mauve s o l u t i o n , did not give any of the expected product according to the infrared spectrum. 98 The reaction with s o l i d iodine to produce (]i3-C3H5)Fe(C0) 3 I i s instantaneous and judging from t h e i i n f r a r e d spectrum, probably quant i ta t ive . After the addit ion of triphenylphosphine to the mauve s o l u t i o n , an a i r - and moisture-sensit ive green s o l i d can be i s o l a t e d . To date t h i s green s o l i d has been incompletely characterised by i n f r a r e d , 5 7 F e Mossbauer, and n.m.r . spectroscopy and elemental analysis of the carbon, hydrogen and iodine con-tent . The n .m.r . data show only phenyl proton resonances, however, other protons present may not have been detected because the s o l i d ds not very soluble in d 6-benzene. The 5 7 F e Mossbauer spectrum (Figure 3-4) i s complex with four peaks; th is resu l t indicates that there are at least two iron s i t e s , which are equally populated. The broadness of peak 3 may be caused by decom-pos i t ion of the sample or a l t e r n a t i v e l y i t may resu l t from a t h i r d i ron s i t e i n the compound. Repeated attempts to i s o l a t e a product from the reaction of the mauve solut ion with Ph 3SnCl y i e l d only a small quantity of a ta r ry brown s o l i d , whose infrared spectrum shows three bands i n the metal carbonyl region and whose n .m.r . spectrum displays phenyl proton and two higher f i e l d doublet resonances. By comparison, the ha l ides , (h.5-Cp)Fe(C0)2X (X = Cl or I ) , react with ytterbium to produce [(]^ 5-Cp)Fe(C0) 2] 2 i n £§_• 70% y i e l d . Addit ion of Ph 3SnCl to the reaction solut ion does not produce any (J^ 5-Cp)(C0) 2Fe-SnPh 3. Figure 3-4 Mossbauer Spectrum of green solid from rnauve solution and Triphenylphosphine. 100 3.3c CYCLOPENTADIENYLTRICARBONYLMOLYBDENUM DIMER REACTIONS Ytterbium reacts with [(h_ 5-Cp)Mo(CO) 3] 2 i n THF ei ther i n the presence or absence of (CH 2 Br) 2 to produce two react ive s o l u t i o n s , whose infrared spectra d i f f e r i n the studied range 2100 - 1500 cm" 1 . The solut ion prepared using (CH 2 Br) 2 shows bands a t t r ibutab le to the parent carbonyl plus the f o l l o w i n g : v m a Y 2025shf* 1925sh, 1875sh, 1850m, and 1585m cm" 1 . I f (CH 2 Br) 2 i s not employed the r e s u l t i n g brown solut ion has a completely new infrared spectrum i n comparison to [ (h 5 -Cp)Mo(C0) 3 ] 2 , namely: v m a Y 2025m, 1925sh, 1910s, 1815vs, iTlQ A 1790sh, 1745m, 1675br,s, and 1580m cm" 1 . The species prepared i n the presence of (CH 2 Br) 2 reacts with Diazald to form the corresponding n i t r o s y l (h_5-Cp)Mo(C0)2N0 (37% y i e l d ) , but attempts to form a Sn-Mo bond with Ph 3SnCl are not successful . On the other hand, the brown solut ion prepared i n the absence of (CH 2 Br) 2 reacts smoothly with iodine and water to form (h_5-Cp)Mo(C0)3I and (h_5-Cp)Mo(C0) 3 H respec t ive ly , i n f a i r y i e l d s . 101 3.3d REACTIONS WITH MERCURY COMPOUNDS Ytterbium displaces mercury from (h_5-Cp)Cr(C0)3HgCl to generate a reactive mixture, one of whose minor components can be i d e n t i f i e d as [(h_ 5 -Cp)Cr(C0) 3 ] 2 . An infrared spectrum (Nujol mull) of the reactive s o l u -t i o n , a f te r solvent removal, displays the fo l lowing peaks: v m a x 2010s, 1935sh, 1925s, 1905s, 1808s, 1678s, 1664s, 1010m, 862m, 710m, 656m, and 620m cm" 1 . The react ive solut ion combines smoothly at 55°C with Ph 3SnCl to y i e l d Ph 3Sn-Cr(h_ 5-Cp)(CO) 3 . The analogous reaction using PhHgCl with ytterbium produces only Ph2Hg in 41% y i e l d . 102 3.4 DISCUSSION 3.4a REACTIONS INVOLVING MANGANESE CARBONYLS The red so lut ion r e s u l t i n g from the a c t i v a t i o n of Mn(C0)5Br by elemental metals i n THF (as well as from other synthetic routes) exhib i ts a var ie ty of in teres t ing physical and chemical propert ies . I t undergoes halide displacement reactions to produce the der ivat ives RMn(C0)5 (R = Ph 3 Sn, Me or C 3 H 5 ) , but i t does so less e f f i c i e n t l y than the sodium s a l t , Na[Mn(C0)5] in terms of y i e l d s and reaction c o n d i t i o n s . 1 1 6 ' 1 1 8 131 I t i s worth noting that according to Dessy et_ al_. the y i e l d s of reactions to form metal-metal bonds from anions appear to be dependent upon the counter c a t i o n . S p e c i f i c a l l y , i n the formation of (C0) 5Mn-SnPh 3 from Na[Mn(C0)5] an 81% y i e l d i s o b t a i n e d , 1 1 6 whereas using (Ci tH 9)i t N [ M n ( C 0 ) 5 ] the y i e l d i s only 30 - 40%. 1 3 1 On the other hand, the red solut ion permits the preparation of the carbonyl-substi tuted compounds [Mn(C0) 1 +L] 2 (L = PP\ or 1 / 2 ( 2 , 2 ' - b i p y r i d y l ) ) under much milder experimental conditions than those previously employed. For example, i t has been reported that [Mn(C0) H PPh 3 ] 2 can be prepared from M n 2 ( C 0 ) 1 0 and Ph 3P by e i ther u l t r a v i o l e t i r r a d i a t i o n of a cyclohexane so lu-t ion for twelve hours or by heating both in a sealed tube for four hours 120 at 130°C. S imi lar forc ing conditions are necessary for the preparation of 115 the 2 , 2 ' - b i p y r i d y l - s u b s t i t u t e d product. The unexpected product, C 6 H 5 C 0 2 ( C H 2)i t C l , and the unusually high y i e l d of M n 2 ( C 0 ) 1 0 i so la ted from the reaction with benzoyl chlor ide are i n -t r i g u i n g i n that they give an"!indication of species possibly present i n the red s o l u t i o n . Two mechanisms have been proposed for s i m i l a r react ions. The 103 f i r s t proposal involves the cleavage of the solvent (diethyl ether) by •C6H5MgBr' i n the presence of benzoyl chloride and cobaltous chloride to 13? produce many products including the es ter , ethyl benzoate. This r e s u l t can be r a t i o n a l i z e d i n terms of the free radica l mechanism summarized below: C6H5MgBr + MCI —> C 6 H 5 M C l v , + MgBrCl A A™" l 2 C 6 H 5 MC1 X _ 1 —> C 6 H 5 - C 6 H 5 + 2 - M C l ^ . , C6H5C0C1 + .MCI , —> C 6 H 5 C0- + MC1V A - I /\ C 6 H 5 C0- * ( C 2 H 5 ) 2 0 —> C 6 H 5 C 0 2 C 2 H 5 + C 2 H 5 . [M = Co and x = 2 in the p a r t i c u l a r reaction being considered.] By comparison, ' C 6 H 5 Ybr reacts with C6H5C0C1 i n THF i n the absence of other metal halides to produce n-propylbenzoate as a b y - p r o d u c t . 1 1 ^ I t i s , there-f o r e , conceivable that a free radica l mechanism involving a lower halide of the lanthanide plays a ro le i n the reaction of the red solut ion with benzoyl c h l o r i d e . Possible supporting evidence for th i s view i s the substantial y i e l d of M n 2 ( C 0 ) 1 0 which could resul t from radica l coupling of -Mn(C0) 5 u n i t s . Ethers can also be cleaved by acyl chlorides in the presence of a 133 Lewis ac id to produce esters . The suggested mechanism i s an ionic one, involv ing the benzoyl cation as shown below: R'COCl + LA — » R'C0 + + LAC!" R-0 + R ' C 0 + — * R-y-j j-R' R R 0 R-O-C-R' + LAC1" —> RC1 + LA + R-O-C-R' [LA = Lewis a c i d . ] 104 If such a mechanism were operative during the reaction of the red solut ion with benzoyl c h l o r i d e , then cleavage of the c y c l i c ether (THF) would lead to C 6 H 5 C02(CH 2 ) i t Cl, the i d e n t i f i e d organic product. Moreover, the Lewis a c i d i t y of the lanthanide elements i n an organic environment has been established (eg. see Chapter I I ) . Obviously, the exact nature of-the l a n -thanide i n the red solut ion i s presently not known with c e r t a i n t y ; however, formation of the p o t e n t i a l l y a c i d i c halides LnX 2 or LnX 3 from Mn(C0)5X and Ln could also explain the production of M n 2 ( C 0 ) 1 0 . Some other work which may be relevant to th i s ether cleavage reac-t ion involves the species R3MMgX (R = aryl or a l k y l group; M = Si or Ge; X = Cl or Br ) . Although these compounds have never been i so la ted there i s 145 146 147 146 strong evidence from coupl ing , h y d r o l y s i s , ' and carbonation reactions to confirm t h e i r postulated formula. They also undergo cleavage reactions with THF to produce R 3 M ( C H 2 ) l f 0 H . 1 4 6 ' 1 4 8 This alcohol i s not the analogous product to our benzoyl chlor ide reactions but i t does i l l u s t r a t e the a b i l i t y of th is general c lass of reagent to cleave THF. However, the c a p a b i l i t y of ether cleavage may be more related to the coordinative power of magnesium (or ytterbium) than to the nature of the R group i n "RMgX". Some support for th i s l a s t viewpoint i s that Ph 3GeLi only affords a 23% y i e l d of Ph 3 Ge(CH 2 K0H, whereas "Ph 3GeMgCl" produces an 83% y i e l d of the same com-146 pound under s i m i l a r condit ions . The red solut ion i s r e a d i l y n i t rosy la ted with Diazald to give Mn(C0) t fN0. Diazald has previously been employed for the conversion of [ M n ( C 0 ) 5 ] " 9 3 or H M n ( C 0 ) 5 1 3 4 to the i soe lec t ronic c a r b o n y l n i t r o s y l . The i n f r a -red spectrum of the red solut ion exhibi ts strong absorptions charac ter i s t i c of [Mn(C0) 5]~ ( i e . 1885 and 1850 cm" 1) but not of HMn(C0)5. N i t r o s y l a t i o n of 105 the red s o l u t i o n , therefore, presumably occurs v ia the anionic species. The infrared spectrum of the red THF solut ion i n the carbonyl-stretching region shows seven bands of which s ix are strong at one concentra-t i o n or another. The large number of bands, t h e i r variance i n i n t e n s i t y with d i l u t i o n and the coincidence of some of the absorptions with those exhibited by [Mn(C0) 5]" suggest that a number of carbonyl-containing species are present i n the red s o l u t i o n . I t can be noted that red solutions of manganese carbonyl complexes in THF have been prepared by u t i l i z i n g i r r a d i a -t i o n techniques, but the e n t i t i e s present i n these solut ion have been assigned 135 -136 the compositions -MnCCO^ and [ H 2 M n 3 ( C 0 ) 1 2 ] " , neither of which has an infrared spectrum s i m i l a r to the one prepared here. Another possible explanation concerning the nature of the red solut ion i s a manganese carbonyl anion-hydride e q u i l i b r i u m , with the hydride 115 formed perhaps by proton abstract ion from the solvent . Hieber ejt al_. has shown that M n ^ C O ) ^ can react with chlorinated solvents under u l t r a v i o l e t i r r a d i a t i o n to produce Mn(C0) 5Cl and [Mn(C0-)i t Cl 2 ] ' . , and under unspecified conditions with 2 , 2 ' - b i p y r i d y l to form Mn 2 (C0) 8 (b ipy) . Moreover, the prepara-t ion of [ M n 2 ( C 0 ) 9 ] 2 " from M n 2 ( C 0 ) 1 0 and NaBH^ also generates [Mn(C0) 5]" in approximately equal amounts. We have repeated Hieber 's work and found that the mixture of anions i s red , but hasnno infrared absorptions above 1900 c m " 1 . Those peaks below 1900 cm" 1 are ident i ca l to bandsrobserved for the red s o l u t i o n . A c i d i f i c a t i o n of Hieber 's anionic mixture permits ultimate i s o l a t i o n of Mn 2 (C0) 9 H 2 whose infrared spectrum i n THF shows bands at 2050s, 2035ms, 1995s, and 1960m-s c m " 1 . The posit ions and r e l a t i v e i n t e n s i t i e s of these bands do not match those of the red solut ion at the equivalent or any other 138 concentrat ion, nor does the infrared spectrum of Mn(C0)5H give a better f i t . 106 The pertinent question i s whether or not a metal carbonyl hydride can ex i s t i n the presence of lanthanide metals without metal lat ion occurr ing . The answer i s not yet known, but i t seems u n l i k e l y that our so lut ion contains a hydride, although the presence of both anions, [Mn(C0) 5]~ and [ M n 2 ( C 0 ) 9 ] 2 " , i s a p o s s i b i l i t y . The u l t r a v i o l e t spectrum of the ytterbium-containing red so lut ion indicates that some of the lanthanide i s present in the t r i v a l e n t s tate . However, "RYbl 1 solutions i n which the metal i s formally divalent display magnetic properties i n d i c a t i v e of the presence of some Y b 3 + . 9 Consequently, the existence of Y b 2 + in the red solut ion cannot be ruled out. A compound containing the t r i v a l e n t lanthanide can be i so la ted a f te r treatment of the holmium-containing red solut ion with sodium acetylacetonate. This compound, whose elemental analysis i s consistent with the formulation [Mn(C0)5]2Ho(C5H702)* 2 E t 2 0 exhibi ts an infrared spectrum in the carbonyl region unl ike that of other complexes which are known to possess metal-Mn(C0) 5 interact ions e i ther 51 116 137 through d i r e c t metal-metal bonds or isocarbonyl l inkages. * ' I f manganese i s used as the ac t iva t ing metal , the e . s . r . spectrum of the red so lut ion and the products of i t s reaction with 2>v2 ,-bipyridyl (eg. MnBr 2*2 bipy) strongly suggest the presence of M n 2 + . In summary, the fact that the red solut ion can be prepared i n a var ie ty of ways, the fac t that i t undergoes a number of diverse react ions , and the fact that i t displays an infrared spectrum r i c h i n carbonyl bands suggest that the red solut ion consists of a mixture of complexes. The physical properties and some of the reactions undergone by the red solut ion suggest the presence of an anionic component, possibly [Mn(C0) 5]~. Moreover, other reac-t i o n products such as [Mn(C0)i +PPh 3] 2 imply the existence of a binuclear mangan-ese carbonyl en t i ty which r e a d i l y undergoes subst i tu t ion of the carbonyl 107 ligands and i s , therefore , u n l i k e l y to be anionic i n nature. Such a binuclear species, which could give r i s e to the higher frequency absorptions i n the infrared spectrum, may be a radica l although reaction of the red s o l u -t i o n with iodine gives only low y i e l d s of M n 2 ( C 0 ) 8 I 2 . 108 3.4b REACTIONS INVOLVING ALLYLTRICARBONYLIRON IODIDE The mauve solut ion prepared from (h^-C 3 H 5 )Fe(C0) 3 I and any of the metals, yt terbium, samarium, yt tr ium or manganese appears to contain the equi l ibr ium mixture: 2 (h3 -C 3 H 5 )Fe (C0) 3 [(h3 - C 3 H 5 ) F e ( C 0 ) 3 ] 2 The reasons for a r r i v i n g at t h i s conclusion are based upon comparative i n f r a -red measurements on an authentic sample of the monomer-dimer equi l ibr ium mix-ture and the mauve s o l u t i o n ; the s i m i l a r reaction of t h i s sample and the mauve solut ion with Diaza ld ; the short l i v e d nature of the mauve solut ion under an iner t atmosphere; the e f fect of iodine upon the mauve s o l u t i o n ; and a comparison with the products obtained from the reaction of the analogous (h_5-Cp)Fe(C0)2X (X = Cl or I) compounds with ytterbium. Each of these points w i l l be considered i n order. The mauve solut ion prepared using ytterbium has the same three carbonyl absorptions in the infrared spectrum as the solut ion prepared using alumina according to Murdoch (Figure 3-3) , namely, 2050, 1995, and 1950 cm" 1 . The behaviour upon d i l u t i o n i s s i m i l a r (Figure 3-3) , that i s , appearance and growth of a band at 2015 c m " 1 , whi l s t the 1950 cm" 1 absorption diminishes and a concomitant increase in i n t e n s i t y of the 1995 cm" 1 peak occurs. The bands due to the monomeric radica l are presumably observed at 2015 and 1995 c m " 1 ; i t then leaves those at 2050 and 1950 cm" 1 to be assigned to the dimer. The resul ts of the infrared d i l u t i o n studies of the mauve so lut ion are i d e n t i c a l , regard-less of the solvent or method of preparation used. Although these resul ts agree q u a l i t a t i v e l y with those of Murdoch, there i s some discrepancy regarding exact band posit ions and assignments. Murdoch's work was carr ied out in an unspecified hydrocarbon solvent and t h i s may account for the difference i n the 109 r e s u l t s . (h_ 3-C 3H 5)Fe(C0) 2N0 can be obtained from the mauve solut ion prepared from alumina or ytterbium and treated with Diazald. This r e s u l t i s in teres t ing 7ft because Diazald i s known to react with metal carbonyl hydrides and metal 93 carbonyl anions (section 2-2) to produce the corresponding i soe lec t ronic n i t r o s y l s . I t now appears that t h i s reagent also reacts with metal carbonyl r a d i c a l s . The mauve solut ion does not react with N0C1, although NOC1 has been observed to form n i t r o s y l s with metal carbonyl anions such as [ M n ( C 0 ) 5 ] " . 1 4 ^ This i n d i c a t i o n , plus the fac t that the anion [ (h > 3 -C 3 H 5 )Fe(C0 )3]~ gives a d i f f e r e n t infrared spectrum 1 4 1 i n THF solut ion ( i e . 1910 and 1855 cm"1) from the mauve s o l u t i o n , strongly suggests that there i s no anion present i n the mauve s o l u t i o n . The mauve solut ion prepared from ytterbium does;:not re ta in i t s colour beyond three hours, a f ter which time i t darkens with accompanying s o l i d deposi t ion. This behaviour i s consistent with the presence of a free radica l which may be undergoing reaction with the solvent or other species i n s o l u -t i on. According to the infrared spectrum, the addit ion of iodine to the mauve solut ion quant i ta t ive ly regenerates (h_ 3-C 3H 5)Fe(C0) 3 I , again t h i s i s ident i ca l to the behaviour Murdoch observed for the (hf -C 3 H 5 )Fe(C0) 3 r a d i c a l . F i n a l l y , the corresponding i ron halides (h_5-Cp)Fe(C0)2X (X = Cl or I) react with ytterbium to form the dimer [(h_ 5-Cp)Fe(C0) 2] 2 for which the analogous monomer dimer equi l ibr ium presumably i s not known because the r a d i c a l , i f i t ex.ists~in s o l u t i o n , i s very short l i v e d . Hence the a l l y l system i s anomalous since a stable free radica l can be formed in th i s case. The function of ytterbium i s the same i n both instances, to abstract halogens. no On the basis of th i s accumulated evidence the act ive component i n the mauve solut ion i s assigned to the free radica l ( h 3 - C 3 H 5 ) F e ( C 0 ) 3 . In reactions involving Ph 3SnCl and the mauve solut ion l i m i t e d evidence for a new product consists of the infrared and n.m.r . spectra, recorded on the very small quantity of impure material obtained. By compari-son with the changes i n the carbonyl region of the infrared spectrum for the conversion of (h_ 5-Cp)Mo(C0) 3I 7 8 to (h. 5 -Cp)Mo(C0) 3 SnPh 3 1 4 2 the spectrum obtained from the mauve so lut ion reaction i s q u a l i t a t i v e l y s i m i l a r : a s h i f t to lower wave numbers of the three observed bands by about 40 cm" 1 i s observed i n both instances. The n .m.r . spectrum displays a strong signal i n the phenyl proton region and two doublets at higher f i e l d . Normally a ir-bonded a l l y l group shows three resonances i n the n .m.r . spectrum, two doublets , each of i n t e n s i t y equivalent to two protons, and a mul t ip le t at lower f i e l d , equiva-lent i n i n t e n s i t y to one proton. I t i s conceivable that the lower f i e l d m u l t i p l e t , a r i s i n g from the sole proton of the central carbon atom of the a l l y l group, was not observed because the sample was not s u f f i c i e n t l y concentrated. The two observed higher f i e l d doublets have coupling constants 103 of the same order of magnitude as those reported for (h_ 3-C 3H 5)Fe(C0) 2N0 104 and (J^ 3 -C 3 H 5 )Fe(C0) 3 I . These two physical measurements are consistent with the formulation (h_ 3 -C 3 H 5 )(C0) 3 Fe-SnPh 3 , however, the assignment and existence of t h i s compound are only t e n t a t i v e l y suggested. In teres t ing ly , the reaction of the anion [(.h3-C3H5)Fe(C0) 3]~ with Ph 3SnCl does not produce 141 the Sn-Fe bonded compound, but instead Ph 6 Sn 2 i s formed in 87% y i e l d . Other attempts to replace bromine in (h^-C 3 H 5 )Fe(C0) 3 Br by t rans i t ion-meta l -containing anions, such as [Mn(C0) 5]~ and [(h. 5 -Cp)Fe(C0) 2 ]~, have only led to the establishment of the afore mentioned e q u i l i b r i u m , n i ( i e . 2 (h 3 -C 3 H 5 )Fe (C0) 3 [ (h_ 3 -C 3 H 5 )Fe(C0) 3 ] 2 ) and not to the desired metal-metal bonded products. The product obtained from the action of Ph 3P upon the mauve so lu-t ion i s somewhat puzzl ing . For the analogous reaction of Ph 3P with the • I O f t equi l ibr ium mixture of (h_ 3-C 3H 5)Fe(C0) 3 and i t s dimer, Murdoch obtained a compound for which he reported only infrared and e . s . r . spectral data: the benzene solut ion infrared spectrum of the carbonyl region had two bands at 1956 and 1893 c m " 1 ; the e . s . r . signal recorded for the Nujol so lut ion of the compound displayed a doublet (17.1 gauss separation) a t t r ibuted to 3 1 P hyperfine s p l i t t i n g . Murdoch interpreted the e . s . r . r e s u l t to mean that only one Ph 3P group was present i n the radica l and taken i n conjunction with the infrared spectrum that there was no doubt the species was the radica l (h_ 3 -C 3 H 5 )Fe(C0) 2 PPh 3 . In contrast , the compound prepared here has s i m i l a r infrared spectral values for the carbonyl s tretching frequencies (1960 and 1885 cm" 1 (CH 2 C1 2 ) , 1955 and 1875 cm" 1 (Nujo l ) , but elemental analyses reveal a substantial presence of iodine and an n.m.r . spectrum displays no para-magnetic broadening of i t s only resonance s i tuated i n the usual pos i t ion for phenyl protons. Even i f the so lut ion used to record the n.m.r . spectrum was too d i l u t e to reveal the a l l y ! group protons, the species i s not (h_ 3 -C 3 H 5 )Fe(C0) 2 PPh 3 I , which i s a d i f f e r e n t co lour , brown, and has d i f f e r e n t 143 infrared carbonyl bands at 1965 and 2015 cm (HCC13 as so lvent ) . The elemental analyses can best be f i t t e d , by the formulation F e 2 ( C 0 ) 5 ( P P h 3 ) 3 I (calc C, 60.7; H, 3.90; I , 10.90. Found: C, 60213; H, 4.16, I , 11.27). Mossbauer ( 5 7 Fe) measurements are s i m i l a r to quadrupole s p l i t t i n g and isomer s h i f t values for monomeric i ron carbonyls containing P h 3 P 1 4 9 (eg. Fe(C0KPPh 3 , Q . S . : 2.54 mms"1, I . S . F e : -0.088 mms"1, and F e ( C 0 ) 3 ( P P h 3 ) 2 s Q .S . : 2.76 mms"1, . L S . p T -0.098 mms'1) or i o d i n e 1 4 9 (eg. Fe(C0) 1 + I 2 , Q .S . : 0.38 mms"1, I - S . p e : +0.14 mms"1)- The quadrupole s p l i t t i n g s for monomeric i ron carbonyls contain-112 ing an a l l y ! group are generally quite d i f f e r e n t from those of the compound iso la ted here (eg. (h_ 3 -C 3 H 5 )Fe(C0) 2 PPh 3 Br, Q .S . : 1.60 mms"1, I - S . ^ : +0.19 m m s " l l 5 ° and ( h 3 - C 3 H 5 ) F e ( C 0 ) 3 B r , Q.S . : 1.50 mms"1, I . S . p e : +0.10 m m s " * 1 5 0 ) . Amongst dimeric i ron carbonyls, (0C)i tFe(PMe 2) 2Fe(C0)i + and (0C) 3 Fe(PMe 2 ) 2 Fe(C0) 3 have s i m i l a r parameters (Q.S. : 2.58 mms"1, I . S . p : -0.032 m m s ~ l 1 4 9 and Q.S . : 0.685 mms' 1 , I . S . F : -0.043 m m s " 1 1 4 9 respect ively) to . those-of the newly prepared species, whereas I(0C) 3 Fe(PMe 2 ) 2 Fe(C0) 3 I (Q.S. : 0.99 mms"1, I .S .p : -0.001 m m s " 1 1 4 9 ) has somewhat d i f f e r e n t values. The compound pre-pared herenrmay have one i ron s i t e which i s f i v e coordinate because the quad-rupole s p l i t t i n g of l ines 1 and 4 i s of the r i g h t magnitude, that i s ca. - i 149 2.5 mms L . Furthermore, the n.m.r . evidence taken in conjunction with the Mossbauer spectrum suggests the absence of an a l l y l group. F i n a l l y , the infrared spectrum indicates only terminal carbonyl absorptions, which leaves one to speculate upon the existence of an i r o n - i r o n bond or iodine bridges, since there are at l eas t two iron s i tes in the compound. 1.13 3.4c REACTIONS INVOLVING CYCLOPENTADIENYLTRICARBONYLMOLYBDENUM DIMER The react ive species prepared from [(h. 5 -Cp)Mo(C0) 3 ] 2 and ytterbium in the absence of (CH 2 Br) 2 displays an infrared spectrum i n THF reminiscent of that of Al [ (h 5 -Cp)W(C0) 3 ] 3 -3THF 5 0 (section 2 .1) . The large number of bands i n the range 2100 - 1500 cm" 1 strongly suggests isocarbonyl bonding. Furthermore, the ready reaction with water or iodine i s ; charac te r i s t i c of the anion [(h_ 5-Cp)MoC0 3]", a property also found for the above mentioned aluminium complex. I t should not be overlooked, however, that iodine reacts i n s t a n t l y with [(h_ 5-Cp)Mo(C0) 3] 2 i n HCC13 to produce the corresponding iodide 144 in 20 - 70% y i e l d . I t i s conceivable, therefore , that any unreacted dimer would also combine with iodine to y i e l d (h^-Cp)Mo(C0) 3I. A comparison with the infrared spectra of the molybdenum dimerr, however, shows the presence of t h i s compound to be neg l ig ib le i n the react ive mixture. The infrared spectrum of the red-brown solut ion prepared from ytterbium and [(h_ 5-Cp)Mo(C0) 3] 2 i n the presence of (CH 2 Br) 2 shows an absorption at 1585 cm" 1 and a new band at 1850 c m " 1 , w h i l s t the rest of the carbonyl region bears a strong resemblance to the parent dimer. I t i s possible that the band at 1585 c m " 1 , i s caused by e i ther an isocarbonyl l i n k or by a m u l t i p l y -bridging carbonyl r e s u l t i n g from metal-c luster formation. A review of the known molybdenum carbonyl c lus ter compounds, however, reveals no s i m i l a r infrared band. The conversion of the red-brown solut ion to (h_5-Cp)Mo(C0)2N0 with D i a z a l d , together with the lower frequency carbonyl bands suggest some anionic or radica l character of the (Jx5-Cp)Mo(C0)3 group. Furthermore, the lack of reaction of the red-brown solut ion with Ph 3SnCl i s not s u r p r i s i n g , because the corresponding magnesium compound pre-114 pared by Burl i t c h did not react e i t h e r . The exact reason for the lack of r e a c t i v i t y by [(h_5-Cp)Mo(C0)3]~ towards Ph 3SnCl i s not known. The n u c l e o p h i l i c i t y of [Mn(C0) 5]" i s s i m i l a r to that of [(h_ 5-Cp)Mo(C0) 3]~, as 91 measured by the rate of displacement of iodide from methyl i o d i d e , and so i t might be expected that the two anions would react s i m i l a r l y . A more sa l i en t point could be the change i n coordination number i n forming the neutral species: from f i v e to s ix for [Mn(C0)5]~ and from s ix^ to sevenffor [(Jl 5-Cp)Mo(C0)3]7 Presumably, octahedral geometry i s preferred and hence [Mn(C0) 5]" reacts whi l s t [(h5-Cp)Mo(C0)33~ does not (see below for further comment). Assuming (h^-Cp) occupies three coordination s i t e s . 115 3.4d REACTIONS INVOLVING MERCURY COMPOUNDS The two preliminary resul ts involving ytterbium with RHgCl species (R = Ph and (h_ 5-Cp)Cr(C0) 3) indicate that i t i s possible to generate a react ive anionic organometallie l i g a n d , whereas the purely organic '"R" group prefers to be attached to mercury. These resul ts probably r e f l e c t the s t a b i l i t y of the two anions "R" i n THF. In teres t ing ly , chromium has a much lower n u c l e o p h i l i c i t y than molyb-denum in the anions [(h_ 5-Cp)M(C0) 3]~ (M = Cr or Mo), according to Dessy et 91 a l . By using the mercury s a l t method, however, i t i s possible to obtain Ph 3 Sn-Cr(h_ 5 -Cp)(C0) 3 . An explanation of t h i s r e s u l t might be that a more anionic species i s generated by the mercury route. Evidence for t h i s con-c lus ion i s the mull infrared spectrum of the react ive chromium solut ion taken to dryness: the many strong carbonyl s tretching frequencies in the region 1600 - 2050 cm" 1 suggest an isocarbonyl anion (Chapter I I ) . The attempts to prepare a Sn-Mo bond were made using the solut ion prepared using (CH 2 Br) 2 and the infrared spectrum of t h i s so lut ion did not show as many strong bands i n the range 1600 - 1850 c m " 1 , suggesting less anionic character than the chromium counterpart. * * * * * * * * * * * * * * * * * * * * * * * * * As the work progressed, our aim sh i f ted s l i g h t l y from a desire to i so la te an RM^X species to invest igat ing the synthetic u t i l i t y of the reac-t i v e solutions we could generate. I t was for t h i s reason that we were i n t e r -ested i n the general i ty of synthesis of the red manganese carbonyl solut ion and i n turn i t led us to begin using other metals as s t a r t i n g mater ia ls . F i n a l l y , i t i s obviously necessary to perform further experiments to determine the nature of the s y n t h e t i c a l l y v e r s a t i l e , red manganese so lu-116 tion; to ascertain the changes in oxidation state of the metals during its formation, and to discover how these results relate to the data that we have obtained by utilizing other metal carbonyl and organometallic complexes as precursors. 117 CHAPTER IV  CONCLUDING REMARKS I t would not be amiss to make a few general comments i n order to place these studies in the context of what i s known and to indicate where these resul ts might lead. The s p e c i f i c conclusions, which are to be found wi th in each chapter, w i l l not be re i t e ra ted . The object of these invest igat ions was to gain some ins ight into the chemistry of the rare earth elements in organic media. The f i r s t part of the project demonstrated that a wide range of substrates were receptive to some form of interact ion with the Lewis acids R 3Ln (R = MeCp or Cp). The establishment of isocarbonyl linkages between neutral as well as anionic bases and a lanthanide has helped to determine the general i ty of t h i s recently discovered bonding mode for the carbonyl l i g a n d . Furthermore, i t was shown that the Lewis a c i d i t y of R 3Ln could be used to form i s o n i t r o s y l or metal-lanthanide bonds ( i e . with (h^-Cp)WH2) and also to support the involvement of organolanthanides i n the t r i m e r i z a t i o n of PhCsCH. In each of these cases no d e f i n i t i v e evidence existed regarding the role of the rare earth. C lear ly further studies of the i so la ted complexes using other physical techniques, such as x-ray d i f f r a c t i o n , would be u s e f u l . The resul t s obtained from the reactions employing elemental metals indicate that t h i s i s an area worthy of more deta i led study. It i s conceiv-able that s y n t h e t i c a l l y v e r s a t i l e reagents could be developed from th i s preparative route. In a d d i t i o n , the exact nature of these solutions requires considerable further examination, i n p a r t i c u l a r by spectral methods. One aim of these studies was to i s o l a t e a compound containing a lanthanide- t rans i t ion metal bond. In spi te of using various approaches, t h i s 118 end has not been achieved. Often the rare earth appeared to prefer to form isocarbonyl bonds, whereas i n other cases thellanthanide acted as a halogen abstractor . Although the p o s s i b i l i t i e s of forming new organolanthanides may be l i m i t e d , the ro le of rare earths in organometallie chemistry should not be viewed s i m i l a r l y . These invest igat ions have shown that the properties of organolanthanides and the metals themselves can produce a wide range of e f f e c t s . 119 REFERENCES 1. T. Moel ler , D.F. Mar t in , L . C . Thompson, R. Ferrus, G.R. F e i s t e l , and W.J. Randal l , Chem. Rev. , 65, 1 (1965). 2. T. Moel ler , E.R. Birnbaum, J . H . Forsberg, and R.B. Gaybart i n "Progress in the Science and Technology of the Rare Earths" , Ed. L. E y r i n g , Vol 3, Chpt 3, Pergamon Press, New York, 1968. 3. J . M . Birmingham and G. Wilkinson, J . Amer. Chem. S o c , 78, 42 (1956). 4. M. Tsutsui and H . J . G y s l i n g , J . Amer. Chem. S o c , 91_, 3175 (1969). 5. F .A . Hart , A . G . Massey and Mohan Singh Saran, J . Organometal. Chem., 21_, 147 (1970). 6. R.G. Hayes and J . L . Thomas, J . Amer. Chem. S o c , 91_, 6876 (1969). 7. F. Mares, K.f'Hodgson, and A. S t re i twieser , J r . , J . Organometal. Chem., 24, C68 (1970). 8. F. Mares, K.O. Hodgson, and A. S t re i twieser , J r . , J . Organometal. Chem., 28, C24 (1971). 9. D.F. Evans, G.V. Fazakerley, and R.F. P h i l l i p s , J . Chem. S o c , (A) , 1931 (1971). 10. E. Murphy and G.E. Toogood, Inorg. N u c l . Chem. L e t t . , 7_, 755 (1971). 11. B . L . K a l s o t r a , R.K. M u l t a n i , and B.D. J a i n , J . Inorg. Nucl . Chem., 34, 2679 (1972) and r e f s . there in . 12. J . L . S l a t e r , T .C. DeVore, and V. Calder, Inorg. Chem., 1_2, 1918 (1973). 13. F. Calderazzo, P. Pappalardo, and S. L o s i , J . Inorg. N u c l . Chem., 28, 987 (1966). 14. S .A. Cotton, F .A. Hart , M.B. Hursthouse, and A . J . Welch, J . Chem. Soc Chem. Commun., 1225 (1972). 120 15. A . J . Shortland and G. Wilkinson, J . Chem. S o c , Dalton Trans . , 872 (1973). 16. M.R. C o l l i e r , M.F. Lappert, and R. Pearce, J . Chem. S o c , Dalton Trans . , 445 (1973). 17. H. Gysling and M. T s u t s u i , Advanc. Organometal. Chem. !3, 361 (1970). 18. L . J . Nugent, P .G. Laubereau, G.K. Werner and K . L . Vander Shus, J . Organometal. Chem., 27, 365 (1971). 19. R.G. Hayes and J . L . Thomas, Organometal. Chem. Rev. , Sect. A , 7_, 1 (1971). 20. E.O. Fischer and H. F ischer , J . Organometal. Chem., 6_, 141 (1966). 21. C . J . Ballhausen and H.B. Gray, "Molecular Orbi ta l Theory", W.A. Benjamin, New York, 1965, p. 53. 22. R.G. Pearson, J . Amer. Chem. S o c , 85£ 3533 (1963). 23. J . L . Hoard, T .B . Owen, Z . B u z z e l l , and O.N. Salmon, Acta . C r y s t a l l o g r . , 3, 121 (1950). 24. D.F. Shr iver , J . Amer. Chem. S o c , 85, 1405 (1963). 25. K .F . Purcel l and R.S. Drago, J . Amer. Chem. S o c , 88, 919 (1966); K . F . P u r c e l l , i b i d , 89, 247 (1967); K .F . P u r c e l l , i b i d , 89, 6139 (1967); K .F . P u r c e l l , i b i d , 91_, 3487 (1969). 26. N . J . Nelson, N.E. Kime, and D.F. Shr iver , J . Amer. Chem. S o c , 91_, 5173 (1969). 27. R .F . Bryan, P.T. Greene, M . J . Newlands, and D.S. F i e l d , J . Chem. Soc (A) , 3068 (1970). 28. R .F . Bryan and P.T. Greene, J . Chem. Soc (A) , 3064 (1970). 29. D.F. Shriver (1971) private communication. 30. J . L . Atwood and G.D. Stucky, J . Amer. Chem. S o c , 89, 5362 (1967). 121 31. Y. Okaya, M.S. Ahmed, R. Pepinsky, and V. Vand, Z. C r y s t a l l o g r . , 109, 367 (1957). 32. S.R. Rasmussen and N.C. Broch, Acta . Chem. Scand., 20, 1351 (1966). 33. F .A. Cotton and G. Yagupsky, Inorg. Chem., 15 (1967). 34. A. A l i c h , N . J . Nelson, D. Strope and D.F; Shr iver , Inorg. Chem., Y\_, 2976 (1972). 35. G. A l l e g r a , G. Perego, and A. Immirzi , Macromol. Chem., 61_, 69 (1963); J . C . Huffman and W.E. S t r e i b , Chem. Commun., 911 (1971) and K. G r o s l i n g , G.M. McLaughlin, G.A. Sim, and J . D . Smith, i b i d , 1617 (1970). 36. (a) F .A. Cotton, J . G . B u l l i t t , and T . J . Marks, Inorg. Chem., 1J_, 679 (1972); (b) F .A . Cotton, J . G . B u l l i t t and T . J . Marks, J . Amer. Chem. S o c , 92_, 2155 (1970). 37. R.D. Adams and F.A. Cotton, J . Amer. Chem. S o c , 94, 6193 (1972). 38. P. McArdle and A.R. Manning, J . Chem. S o c (A) , 2128 (1970). 39. F .C. Wilson and D.P. Shoemaker, J . Chem. Phys. 27, 809 (1957). 40. P. Braterman and D. Thompson, J . Chem. S o c (A) , 1454 (1968). 41. R.D. Fischer and K. Noack, J . Organometal. Chem., 1_6, 125 (1969). 42. E.W. Abel and F.G. A. Stone, Quart. Rev. Chem. S o c , 23_, 325 (1969). 43. J . F . T i lney-Basset t , J . Chem. S o c , 577 (1961 ). 44. J . C . Kotz and C D . Turnipseed, Chem. Commun, 41 (1970). 45. J . C . Kotz and D.G. Pedrotty, Organometal, Chem. Rev. , Sect. A , 4_, 479 (1969). 46. J . M . B u r l i t c h and R.B. Petersen, J . Organometal. Chem. 24, C65,(1970). 47. E . L . Brown and D.B. Brown, Chem. Commun., 67,(1971). 48. D.A. Dows, A. Haim and W.K. Wilmarth, J . Inorg. Nuc l . Chem., 21_, 33 (1961). 122 49. F .A. Cotton, R.R. Monchamp, R . J .M. Henry, and R.C. Young, J . Inorg. N u c l . Chem., 10, 28 (1959). 50. R.B. Petersen, J . J . Stezowski, Ch'eng Wan, J . M . B u r l i t c h , and R.E. Hughes, J . Amer. Chem. S o c , 93, 3532 (1971). 51. S.W. Ulmer, P.M. Skarstad, J . M . B u r l i t c h , and R.E. Hughes, J . Amer. Chem. Soc . , 95, 4469 (1973). 52. G.B. McVicker and R.S. Matyas, J . Chem. Soc. Chem. Commun., 972 (1972). 53. T. Blackmore and J . M . B u r l i t c h , J . Chem. Soc. Chem. Commun., 405 (1973). 54. R.R. Schrieke and J . D . Smith, J . Organometal. Chem., 31_, C46 (1971 ). 55. G . J . Gainsford, R.R. Schrieke and J . D . Smith, J . Chem. SoCnChem. Commun., 650 (1972). 56. I . L i n d q u i s t , "Inorganic Adduct Molecules of Oxo-Compounds", Academic Press, New York, 1963. 57. T . J . Marks, J . S . K r i s t o f f , A. A l i c h , and D.F. Shr iver , J . Organometal. Chem., 33, 035 (1971). 58. J . Chatt , R.H. Crabtree, and R .L . Richards, J . Chem. Soc, Chem. Commun., 534 (1972). 59. J . Chatt , R.H. Crabtree, E .A. Je f f rey and R .L . Richards, J . Chem. S o c , Dalton Trans . , 1167 (1973). . 60. R. Chim" and E. E r e o l i , Gazz. Chim. I t a l . , 88, 1170 (1958). 61. G.G. Summer, H.P. Klug , and L . E . Alexander, Acta . C r y s t a l l o g r . , V7, 732 (1964) 62. K. Noack, Spectrochim. Acta , 19, 1925 (1963). 63. G. Bor, Spectrochim. Acta , 1_9, 2065 (1963). 64. G. Bor and L. Marko, Chem. Ind. (London), 912 (1963). 65. F .A. Cotton and R.R. Monchamp, J . Chem. S o c , 1882 (1960). 66. R. Whyman, Nature (Physical Science) (London), 230, 139 (1971). 123 67. G. Schmid and V. B a t z e l , J . Organometal. Chem., 46, 149 (1972). 68. J . S . K r i s t o f f , N . J . Nelson, and D.F. Shr iver , J . Organometal. Chem., 49, C82 (1973). 69. D.F. Shriver and A. A l i c h , Inorg. Chem., 1J_, 2984 (1972). 70. M. T s u t s u i , T.Takino, and D. Lorenz, Z. Naturforsch, 21B, 1 (1966). 71. R.S. M a r i a n e l l i and M.T. Durney, J . Organometal. Chem., 32, CA1 (1971). 72. T .A. Manueland, F .G.A. Stone, J . Amer. Chem. S o c , 82, 366 (1960). 73. R.B. King, "Organometal1ic Syntheses", V o l . 1, Academic Press , New York, p. 79. 74. i b i d , p. 161. 75. i b i d , p. 167. 76. F .A. Cotton, T.S . P i p e r , and G.Wilkinson, J . Inorg. and Nucl . Chem., 1_, 165 (1955). 77. R.E. Maginn, S. Manastyrskyj, and M. Dubeck, J . Amer. Chem. S o c , 85_, 672 (1963). 78. T.S. Piper and G. Wilkinson, J . Inorg. and Nucl . Chem., 3_, 104 (1956). 79. S.D. Robinson and M.F. U t t l e y , J . Chem. S o c , Dalton Trans . , 1 (1972). 80. R.B. King and M.B. B isnet te , Inorg. Chem. _3» 791 (1964). 81. D.D. P e r r i n , W.L.F. Armarego, and D.R. P e r r i n , " P u r i f i c a t i o n of Labora-tory Chemicals", Pergamon Press, Oxford, 1966. 82. A.R. Pray, Inorg. S y n . , V, 153 (1957). 83. L . F . Fieser and M. F ieser , "Reagents for Organic Synthesis" , John Wiley and Sons, New York (1967), V o l . 1, p. 1158. 84. K.W. Barnett , J . Organometal. Chem., 44, 1 (1972). 85. B .F .G . Johnson and J . A . McCleverty, Progr. Inorg. Chem., 7_, 277 (1966). 86. T . J . Marks and J . S . K r i s t o f f , J . Organometal. Chem., 42, C91 (1972). 124 87. A. Storr and B.S. Thomas, Can. J . Chem., 49, 2504 (1971). 88. D.F . Shr iver , Accounts Chem. Res. , 3_, 231 (1970). 89. (a) M. Pankowski, B. Demerseman, G. Bouquet, and M. Bigorgne, J . Organo-metal. Chem., 35_, 155 (1972). (b) W:R. C u l l e n , D . J . Patmore, and J .R . Sams, Inorg. Chem., 1_2, 867 (1973). 90. A . J . C le land, S .A. Fieldhouse, B .H. Freeland, and R . J . O 'Br ien , J . Chem. L Soc. Chem. Commun., 155 (1971). 91. R .E. Dessy, R.B. King, and R.L. P o h l , J . Amer. Chem. S o c , 88, 5121 (1966). 92. R.B. King and M.B. Bisnet te , Inorg. Chem., 6, 469 (1967). 93. T . H . C o f f i e l d , U.S. Patent 2,967,087. 94. R.B. King, Inorg. Chem., 9_, 1936 (1970). 95. F. Bonati and G. Wilkinson, J . Chem. S o c , 3156 (1964). 96. F .A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry", 1st Edn. Interscience, p. 853. 97. J . Rajaram and J . A . Ibers, Inorg. Chem., 1_2, 1313 (1973). 98. L . Friedman, A . P . Irsa and G. Wilkinson, J . Amer. Chem. S o c , 77_, 3689 (1955). 99. C. Wong, T. Lee, and Y. Lee, Acta . C r y s t a l l o g r . , B25, 2580 (1969). 100. P . L . Timms, Advan. Inorg. Chem. Radiochem., 1£, 121 (1972). 101. See, for example, P .S . Skel l and J . Howe!, J . Amer. Chem. S o c , 93_, 6687 (1971 ); and K . J . Klabunde, M.S. Key, and J .Y. .F. Low, J . Amer. Chem. Soc . , 94, 999 (1972). 102. H. Gilman and R.G. Jones, J . Org. Chem., lp_, 505 (1945). 103. E.O. Fischer and H. Fischer , J . Organometal. Chem., 3_, 181 (1965). 125 104. R.6. Hayes and J . L . Thomas, Inorg. Chem., 8, 2521 (1969). 105. J . M u l l e r , Chem. Ber, 102, 159 (1969). 106. D.F. Evans, G.V. Fazerkerley, and R.F. P h i l l i p s , J . Chem. Soc. Chem. Commun., 244 (1970). 107. A. Carrick and F. Glocking, J . Chem. Soc. (A) , 913 (1968). 108. J . M . B u r l i t c h and S.W. Ulmer, J . Organometal. Chem., J_9, P21 (1969). 109. J . M . B u r l i t c h , personal communication. 110. P. Legzdins and E.R. Sigurdson, unpublished r e s u l t s . 111. S .E. Pedersen, W.R. Robinson, and D.P. Schussler, J . Organometal. Chem., 43, C44 (1972). 112. J . M . B u r l i t c h , J . Chem. S o c , Chem. Commun., 887 (1968). 113. G. Pass and H. S u t c l i f f e , " P r a c t i c a l Inorganic Chemistry", Chapman and H a l l , London (1968). 114. F. Bonat i , S. C e n i n i , and R. Ugo, J . Chem. S o c (A) , 932 (1967). 115. W. Hieber, W. Beck, and G. Z e i t l e r , Angew. Chem., 73, 364 (1961). 116. R.D. Gors ich, J . Amer. Chem. S o c , 84, 2486 (1962). 117. Reference 73, p. 147. 118. W.R. McCle l lan , H.H. Hoehn, H.N. Cr ipps , E . L . Muetter t ies , and B.W. Howk, J . Amer. Chem. S o c , 83, 1601 (1961 ). 119. F .A. Cotton, J.W. F a l l e r , and A. Musco, Inorg. Chem., 6_, 179 (1967). 120. A . G . Osborne and M.H.B. S t iddard, J . Chem. S o c , 634 (1964). 121. Reference 73, p. 164. 122. M.A. El-Sayed and H.D. Kaesz, Inorg. Chem., 2, 158 (1963). 123. H.D. Murdoch, Z. Naturforsch, 20B, 179 (1965). 124. Reference 73, p. 176. 126 125. T .S . P i p e r , F .A. Cotton, and G. Wilkinson, J . Inorg. Nucl . Chem., 1_, 165 (1955). 126. F. F i e g e l , "Spot Tests" , V o l . ]_, E l sev ier 4th E d i t i o n , English Edi t ion (1954). 127. F .A. Cotton and G. Wilkinson, Progr. Inorg. Chem., J_, 40 (1959). 128. A. Miyake, H. Kondo and M. Aoyama, Ang. Chem. I n t l . E d n . , 8, 520 (1969). 129. G.E. Coates, "Organometallic Compounds", Methuen, London, 1960, Second Edn. , p. 80. 130. D. Murdoch and E . A . C . Lucken, Heilv. Chim. Acta . 47, 1517 (1964). 131. R.E. Dessy and P.M. Weis.sman, J . Amer. Chem. S o c , 88, 5124 (1966). 132. M.S. Kharasch and 0. Reinmuth, "Grignard Reactions of Nonmetallic Sub-stances", Prent ice-Hal l I n c . , New York, 1954, Chapter V and references there in . 133. J . March, "Advanced Organic Chemistry; Reactions, Mechanisms and Structure" , McGraw-Hil l , New York, 1968, p. 345. 134. P.M. T r e i c h e l , E. P i t c h e r , R.B. King , and F .G.A. Stone, J . Amer. Chem. S o c , 83, 2593 (1961 ). 135. S .A. Hal lock and A. Wojc i ck i , 0. Organometal. Chem., 54, C27,(1973). 136. G.O. Evans, J . S l a t e r , D. Giusto and R.K. Shel ine , Inorg. Nucl . Chem. L e t t . , 7, 771 (1971 ) . 137. A . T . T . Hsieh and M . J . Mays, J . Chem. S o c , Dalton Trans . , 516 (1972). 138. F .A. Cotton and G. Wilkinson, J . Chem. S o c , 833 (1959). 139. R.B. King, Accounts Chem. Res . , 3_, 417 (1970). 140. P. Legzdins and J .T . M a l i t o , unpublished r e s u l t s . 141. S .P. Gubin and L . I . Denisovich, J . Organometal. Chem., 15_, 471 (1968). 142. H.R.H. P a t i l and W.A.G. Graham, Inorg. Chem., 5_, 1401 (1966). 127 143. R .F . Heck and C R . Boss, J . Amer. Chem. S o c , 86, 2580 (1964). 144. E.W. A b e l , A. Singh, and G.V'Wilkinson, J . Chem. S o c , 1321 (1960). 145. T .G. Se l in and R. West, Tetrahedron, 5_, 97 (1959). 146. H. Gilman and A. Zuech, J . Org. Chem., 26, 3035 (1961). 147. F. Glocking and K.A. Hooton, J . Chem. S o c , 3509 (1962). 148. W. Steudel and H. Gilman, J . Amer. Chem. S o c , 82_, 6129 (1960). 149. N.N. Greenwood and T . C Gibb, "Mossbauer Spectroscopy", Chapman and H a l l , London, 1971, Chapter 3 and references there in . 150. K. Burger, II. Korecz, and G.LBor, J . Inorg. Nuc l . Chem., 31_, 1527 (1969). 151. M.P. Johnson and D.F. s h r i v e r , J . Amer. Chem. S o c , 88, 301 (1966). 128 APPENDIX The Determination of Lanthanides in Organometal!ic Complexes by the Closed Oxygen Flask Method INTRODUCTION The work reported in th is appendix i s almost e n t i r e l y that of Mr. P. Borda. I t has been included, however, for two reasons. The develop-ment of th is a n a l y t i c a l technique was essent ial to characterise several so l ids that were i s o l a t e d , hence the nature of the problems involved i n the project are more f u l l y displayed. Secondly, th i s information-Ms v i t a l for other workers entering th i s f i e l d . The quant i tat ive determination of the lanthanide content i n organometallie compounds becomes a necessity for those species whose charac-t e r i z a t i o n by common physical techniques i s hampered by various f a c t o r s . 4 The to ta l or very poor s o l u b i l i t y of organolanthanides i s the major reason that prompted th is work. Since a l l organolanthanides are a i r - and moisture-sens i t ive to varying degrees, the customary procedures involved i n metal est imation, the dry combustion of a sample in oxygen to the corresponding metal oxide or the wet degradation of the sample by an appropriate acid medium in a Kjeldahl f l a s k , become extremely cumbersome. Moreover, the dry combustion technique i s t o t a l l y inappl icable i f metals other than the rare earths are also present in the compound to be analyzed. I t was, therefore , necessary to develop a method of lanthanide determination which can be carr ied out s imply, accurately and r a p i d l y i f the a n a l y t i c a l sample is i n i t i a l l y decomposed by means of the oxygen f l a s k . 129 Since 1955 considerable attention has been focused on the 7 14 15 analysis of non-metals by Schdniger's oxygen f lask method.' ' However, despite i t s apparent u t i l i t y for the analogous determination of metals, t h i s 1 o 12 "]3 15 combustion technique has only been infrequently employed ' ' ' ' during the estimation of both main group (eg. Mg, Ca, Ba and Sn) and t r a n s i t i o n metals (eg. Zn, Cd, Hg, Mn and Co). Furthermore, the metal-containing species which were studied were pr imar i ly a i r - s t a b l e coordination compounds. True organometallic complexes ( i e . those containing d i r e c t metal-carbon bonds) have large ly been neglected. The present studies represent the f i r s t exten-sive appl ica t ion of the oxygen-flask technique during the analysis of a class of these l a t t e r compounds. Accounts of the basic p r i n c i p l e s , methods and problems of metal 1 8 analysis by the Schdniger method have been published. * 130 EXPERIMENTAL MATERIALS AND METHODS Reagents and Apparatus A l l reagents used were of ana ly t i ca l grade or comparable p u r i t y . The t i t r a t i o n s were carr ied out with e i ther a Gilmont micrometer burette (Cole-Parmer catalog number 7876) having a 2.5000 ml capacity and capable of measuring to 0.0001 m l , or a Mettler E457 micrometer burette having a 5.000 ml capacity and capable of measuring to 0.001 m l , as required. The hydrogen-ion concentration in the various solutions was monitored by an Orion model 801 d i g i t a l pH meter. Procedure Since a l l of the compounds examined were sens i t ive in d i f f e r i n g degrees to both a i r and moisture, a l l manipulations of the s o l i d species g p r i o r to combustion were performed i n a glove bag f i l l e d with prepuri f ied nitrogen. Analy t i ca l samples were prepared by placing 5-10 mg of the organo-lanthanide der iva t ive into pre-weighed adhesive c e l l u l o s e containers f i t t e d with a f i l t e r paper l i n i n g . The containers were then sealed and t h e i r weight g was determined e i ther by the procedure out l ined by Pickhardt and co-workers , or by d i r e c t weighing i n the nitrogen atmosphere on a Cahn e lec tronic balance. The samples were then igni ted in a 500-ml oxygen f lask charged with 10 ml of e i ther IN.HCl or 1N"HN03 as the absorbent s o l u t i o n . Af ter combustion was com-p l e t e , the f l ask was shaken thoroughly for ten minutes, and then the stopper and platinum gauze were rinsed with d i s t i l l e d water. During th i s operation i t was noted that the s o l u b i l i t i e s of the lanthanide residues in the absorbent solutions diminished as the atomic weight of the metal increased. Thus, for 131 example, the L a , Sm and Gd oxides r e a d i l y dissolved i n the acid solut ion whereas the corresponding Dy, Er and Yb species required gentle heating to e f fec t complete d i s s o l u t i o n . Once the metal oxides were d i s s o l v e d , the contents of the oxygen f lask were then washed into a 150 ml beaker, the total volume of the so lut ion at th i s stage being approximately 50 ml . The most rapid and simple means of f i n a l determination involved the t i t r i m e t r i c evaluation of the lanthanides with 0.01 N or 0.005 N EDTA. Two types of t i t r a t i o n environments were u t i l i z e d and the end points of the t i t r a t i o n s were ascertained by the threenmethods outl ined below. A. A basic medium with Eriochrome Black T as indicator The hydrochloric acid absorbent solut ion was neutral ized at room temperature with IN NH^OH, with sodium tar t ra te being added at pH 6 to prevent p r e c i p i t a t i o n of hydroxide der ivat ives of the metals. An appropriate N H 4 C I - N H 4 O H buffer solut ion was next introduced to maintain the pH of the solut ion in the optimum range of 8.3 - 8.6. The solut ion was then heated to ^80°C, and was t i t r a t e d d i r e c t l y at th i s temperature i n the presence of Eriochrome Black T. Throughout th i s t i t r a t i o n , str ingent adherence to a spec i f ied pH range (narrower than that previously reported for analogous q determinations ) , was found to be mandatory because the indicator was very sens i t ive to changes i n the hydrogen-ion concentration, espec ia l ly when small amounts of metals such as ytterbium were being t i t r a t e d . For example, at pH 8.8 the end points were attained very slowly and the recoveries of the heavier lanthanides were never quant i ta t ive . Below pH 8.3 , on the other hand, the color change of the indicator near the end point was very poor. 132 B. An a c i d i c medium with Xylenol Orange as indicator The pH of the hydrochloric acid absorbent solut ion was adjusted to approximately 4 with lNNH^OH. While the solut ion was gently warmed, an appropriate CH3C00Na-CH3C00H buffer was added to maintain the pH of the solut ion in the range 4 . 8 - 5 . 5 . This environment was found to be the most sa t i s fac tory of a l l for obtaining good end points in the d i rec t t i t r a t i o n of the warm solut ion with Xylenol Orange as i n d i c a t o r . I f the pH was allowed to increase above 5.8 , not only were repeated fa l se end points prematurely observed, but also the reddish t i n t of the indicator i t s e l f obscured the desired color change at the true end point . Our f indings thus substantiated recent reports [c f . 5] that the true working pH range of Xylenol Orange i n such complexometric t i t r a t i o n s i s a c t u a l l y lower than had been claimed i 6 previously . C. An a c i d i c medium with the end point being detected potent iometr ical ly One normal n i t r i c acid was used as the absorption medium during the i n i t i a l combustion. The pH of the f i n a l absorbent solut ion was adjusted to 4.3 - 5.0 with IN iNaOH. Then 5 ml of an acetate buffer (pH 4.8) and four drops of a 10" 3 M solut ion of the mercury-EDTA complex were added. The resultant so lut ion was heated to 80°C and was t i t r a t e d while hot with EDTA, the end point being detected potent iometr ical ly with a mercury e l e c t r o d e ] ^ ' 1 1 A calomel electrode f i l l e d with a saturated solut ion of KN03 was employed as a reference. No d i f f i c u l t i e s were experienced with t h i s method and a l l t i t r a t i o n s were e a s i l y performed. Moreover, the presence of chlor ine in the organo-133 lanthanide sample did not in ter fere with the mercury indicator electrode. Hence, because of i t s general a p p l i c a b i l i t y , espec ia l ly when very small {<1 mg) amounts of metal were to be determined, and less rigorous experimental condi t ions , th is procedure eventually became the method of choice. The closed oxygen f lask technique was also successful ly u t i l i z e d during the analysis of more complex organometallic compounds. For instance, the compounds which contained both tungsten and a lanthanide were i n i t i a l l y burned in the usual manner. The absorbent solut ion was then simply b o i l e d , the tungstic acid was removed from the hot so lut ion by f i l t r a t i o n , and the lanthanide content of the f i l t r a t e was determined by one of the methods described previously . 134 RESULTS AND DISCUSSION Good precis ion end points were obtained with a l l the methods of f i n a l determination as shown by the representative data displayed i n the Table. The resul ts indicate that the closed oxygen f lask method provides a general means of decomposing very reactive compounds which contain d i r e c t metal-carbon bonds. Moreover, the f lask method expedites the desired metal analyses i n those cases for which a rapid f i n a l method of determination i s known or can be developed. In th i s connection i t should be noted that when small amounts ( i e . 1-3 mg) of metals are being analyzed, conditions d i f f e r e n t from those employed for larger -sca le determinations may be required. For example, these studies reveal that for the use of Eriochrome Black T as an indicator in lanthanide-EDTA t i t r a t i o n s , the hydrogen-ion concentration range in which the indicator-metal complex i s not too i n e r t and the protonated form of the indicator does not in ter fere i s , unexpectedly, quite small when micro-determinations of metal are performed. Cer ta in ly the closed oxygen f lask method i s invaluable as the f i r s t step during analysis of the lanthanide content of a l l types of organolanthanide complexes. 135 TABLE Analysis of Various Organolanthanide Complexes % Lanthanide Compound Theoretical Found Tris(cyclopentadienyl)lanthanum, ( C 5 H 5 ) 3 L a 41.56 41.47 Tris(cyclopentadienyl)samarium, (C5H5)3Sm 43.51 43.21 Tris(methylcyclopentadienyl)samarium b , (C 6 H 7 ) 3 Sm 38.79 38.64 Tr i s(methylcyclopentadi enyl)gadoli ni um, (C 6 H 7 ) 3 Gd 39.84 39.81 Tris(methylcyclopentadienyl)dysprosium, (C 6 H 7 ) 3 Dy 40.64 40.69 Tris(cyclopentadienyl)erbium, ( C 5 H 5 ) 3 E r 46.12 46.34 Bis(cyclopentadienyl)erbium c h l o r i d e , ( C 5 H 5 ) 2 E r C l 50.23 50.47 Bis(cyclopentadienyl)ytterbium c h l o r i d e 3 , (C 5H5) 2YbCl 51.08 51.31 T r i carbonyl tr i s(cyclopentadi enyl)erbi umtungstenb (C 5H5) 2ErW(C 5H5)(C0) 3 26.54 26.42 T r i carbonyl tr i s(cyclopentadi enyl)tungstenytterbi um3 (C 5 H 5 ) 2 YbW(C 5H5)(C0) 3 27.20 27.04 3 Determined by Method A; b Determined by Method B; A l l others determined by Method C. 136 APPENDIX REFERENCES 1. R. Belcher, A.M.G. Macdonald, and T.S. West, Talanta, 1_, 408 (1=958). 2. M. Corner, Analyst (London), 84, 41 ( 1 9 5 9 ) . 3. H.A. Flaschka, Mikrochim. A c t a . , 55 ( 1 9 5 5 ) . 4. F .A. Hart , A .G. Massey and M.S. Saran, J . Organometal. Chem., 21_, 147 ( 1 9 7 0 ) . 5. O.H. Kriege and M.L. Theodore, Talanta , 13, 265 ( 1 9 6 6 ) . 6. S . J . Lyle and M.M. Ratiman, Talanta, l f j , 1177 ( 1 9 6 3 ) . 7. A.M.G. Macdonald, Analyst (London), 8 6 , 3 ( 1 9 6 1 ) . 8. A.M.G. Macdonald and P. S i r i chanya , Microchem. J . , 14, 199 ( 1 9 6 9 ) . 9. W.P. Pickhardt , L.W. Safranski and J . M i t c h e l l , Ana l . Chem., 30, 1298 ( 1 9 5 8 ) . 10. C . N . R e i l l e y and R.W. Schmid, Anal . Chem., 30, 947 ( 1 9 5 8 ) . 11. C . N . R e i l l e y , R.W. Schmid and D.W. Lamson, Anal . Chem., 30, 953 ( 1 9 5 8 ) . 12. R. Reverchon and Y. Legrand, Chim. Anal . ( P a r i s ) , 47_, 70 ( 1 9 6 5 ) . 13. A . B . Sakla , S'.W. Bishara and S.A. Abo-Taleb, Microchem. J . , J_7, 436 ( 1 9 7 2 ) . 14. W. Schoniger, Mikrochim. Acta. . , 123 ( 1 9 5 5 ) . 15. W. Schoniger, Mikrochim. A c t a . , 869 ( 1 9 5 6 ) . 16. B .C. Southworth, J . H . Hodecker, and K.D. F le i scher , Anal . Chem., 30_, 1152 ( 1 9 5 8 ) . 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099929/manifest

Comment

Related Items