Reactions of heterocyclic organotellurium compounds with triironddecacarbonyl: reactions of thiophenes revisitedKaranbir Singh," William R. McWhinnie," Hong Li Chen,b Min Sunb and Thomas A. Hamora Department of Chemical Engineering and Applied Chemistry, Aston University, Aston Triangle,Birmingham B4 7ET, UKSchool ($'Chemistry, University of Birmingham, PO Box 363, Birmingham B1.5 2TT, UKThe reaction of thiophene and of benzothiophene with [Fe,(CO), 2] was greatly accelerated by microwaveheating when carried out in the presence of Fe,O,. Thus the same yields of products are obtained in 50 min asare achieved in 15-1 8 h of conventional heating; however no desulfurisation of the benzothiophene ring wasobserved. By contrast, both tellurophene and, more importantly, dibenzotellurophene undergo detellurationreactions with [Fe,(CO),,], the latter reaction affording a dibenzoferrole Cl,H,Fe206, the structure of whichhas been determined.The reaction of 2-telluraindane with [Fe,(CO),,] gave a complex CI,Hl6FeO3 in whicha novel dimer of the detellurated C,H, fragment is co-ordinated to an Fe(CO), unit, as determinedcrystallographically. The released tellurium was isolated as either FeTe or as [Fe,Te,(CO),], depending on thereaction conditions.In 1960 Stone and co-workers,' in a classic contribution,demonstrated that the reaction of thiophene and [Fe,(CO), ,]afforded the ferrole C4H4Fe,(C0)6 together with FeS andother minor products. It was known that [Fe(CO),] wouldcleave the C-S bonds of alkyl sulfides2 and Stone and co-workers went on to show that the C-S bonds of vinyl sulfideswould cleave in the presence of [Fe,(CO),,].The initialyields of' the ferrole were poor (5% after 15 h),' but othersdemonstrated that more prolonged reaction of thiophene with[Fe,(CO),,] (2 d) increased the yield to 17%.4 The use ofsubstituted thiophenes gave yields of between 0.7 and 10.7%of the appropriate ferroles.This chemistry was believed potentially to provide usefulmechanistic insight to the initial stages of heterogeneouslycatalysed hydrodesulfurisation reactions (HDS) of oils andof coal-derived liquids. In particular the laboratories ofRauchfuss and Angelici have been active in the area. Thus,observations by Rauchfuss and co-workers ' that iron initiallyenters the thiophene ring to give a thiaferrole may model animportant initial step in the heterogeneously catalysed HDSprocess.Angelici and co-workers ' v 9 explored the mode ofbinding of thiophene derivatives with low-oxidation-statetransition-metal centres, arguing that such complexes maymodel the initial interaction of thiophenic moieties with metalsites in the surface of heterogeneous HDS catalysts.Thiophene is not an ideal molecular model for the bulk oforganic sulfur in coal-derived liquids, rather benzothiopheneand, more particularly, dibenzothiophene, are superiormodels. Benzothiophene gives a thiaferrole on reaction with[Fe,(CO) 2] which undergoes oxidative demetallation, but noremoval of ring sulfur is noted.' Dibenzothiophene (dbt) isunreactive under similar conditions but it has been reported toco-ordinate to iron in a monodentate fashion in the complex[Fe(dbt)(CO),(~p)l[BF,],'~ where cp = q-C,H,-.Such co-ordination (through S) has also been observed by Angelici andco-workers l 2 in [M(dbt)Cl,(q-C,Me,)] (M = Rh or Ir), butwhen the metal is in oxidation state (I) the co-ordinationchanges to q6, i.e. [M(q6-dbt)(q-C,Me,)]. An electron-richmetal centre is believed to be required for metal insertion intothe thiophene ring,' and Rauchfuss and his group l4 argue thatmultimetallic reagents are advantageous for sulfur removal.This paper initially explores the question of the reactivityof materials such as thiophene and benzothiophene with[Fe3(CO>,,] under the influence of microwave heating.Sincethis methodology has been shown to be capable of greatlyaccelerating the passage to equilibrium of a large number ofchemical reactions it was interesting to examine whether,under these conditions, the benzothiophene reaction wouldproceed to the ferrole stage. Our major purpose, however,was to explore the reactivity of some tellurium heterocycliccompounds with [Fe,(CO), ,I. Electronically tellurophenes aresimilar to thiophenesl6 and may therefore be regarded asthiophene 'models'. The carbonxhalcogen bond strength willbe less in the tellurophenes and this may facilitate detellurationreactions, even of dibenzotellurophene, and thereby demon-strate the mechanistic feasibility of the dechalcogenationreaction for the more condensed aromatic molecules.Angeliciand his group 7 3 1 ' have extended their investigations toselenophenes since 77Se NMR spectroscopy allows directobservation of the ligand; a similar advantage based on 125TeNMR spectroscopy is available for tellurophenes: also a newarea of organometallic chemistry is accessed.Experimental and ResultsTriiron dodecacarbonyl was obtained from Aldrich and used asreceived. Tellurophene, ' dibenzotellurophene 2o and 2-tellu-raindane 21 were prepared by the indicated literature methods.Thiophene was distilled from CaH, prior to use. Pentanewas HPLC grade and used as received from Aldrich. Allmanipulations involving reactions of the tellurium compoundswere carried out under an atmosphere of pure argon with theuse of Schlenk techniques.Reactions of triiron dodecacarbonyl with telluropheneTellurophene (3.00 g, 17 mmol) and [Fe,(CO),,] (3.02 g, 6mmol) were refluxed in heptane (100 cm3) with stirring for2.5 h.After 45 min the green colour of the iron carbonyl gaveway to an intense violet which in turn changed to orange after1.5 h, following which no further colour change occurred. Thecooled solution was filtered to give an orange filtrate and ablack solid which adhered to the sides of the flask. The solventwas removed from the filtrate in uucuo to give orange flake-likecrystals of m.p. 51 "C (0.89 g, 45% based on [Fe,(CO),,]). Theinfrared data were consistent with those previously reportedJ . Chem. SOC., Dalton Truns., 1996, Pages 1545-1549 154for the ferrole 2, v(C0) 2078m, 2050m, 2000s, 1967s and1940s cm-' (Found: C, 37.0; H, 1.40.C,,H,Fe,O, requires C,36.2; H, 1.20%). Electron impact (EI) mass spectrum; m/z =332 ( M + ) , 56Fe. The black solid residue was shown to beFeTe.The experiment was then repeated exactly as above, butterminated at the point of development of the violet colour.Removal of heptane under vacuum left a deep violet solid whichwas chromatographed on a 2.0 x 20 cm column of TLC-gradesilica thereby giving a small green band preceded by orange andpurple bands. Elution with pentane and removal of solventafforded a dark red powder from the first eluate, orange crystalsfrom the second (C,,H,O,Fe,), and unreacted [Fe,(CO),,]from the third.The dark red material was recrystallised fromboiling heptane yielding shiny, plate-like crystals, m.p. 41 "C,which were identified as the telluraferrole I , yield 1.82 g {66%based on [Fe,(CO),,]) (Found: C, 26.0; H, 0.90. C,,H,Fe,-0,Te requires C, 26.1; H, 0.90%). Chemical ionisation (CI)mass spectrum: miz = 462 ( M + ) , l3'Te and 56Fe. NMR(CDCI3):'H,69.54(d,1H),7.32(q,1H),5.72(d,1H)and5.22(9, 1 H); 13C, 6 181.8, 147.8, 88.8, 86.8 and 86.4; 125Te, 6 43.0[(J(Te-H') = 93, J(Te-H2) = 19, J(Te-H3) = 5 Hz]. Infra-red (KBr, cm '): v(C0) 2069m, 2030m, 1981s and 1961m.With dibenzotellurophene. Dibenzotellurophene ( 1.93 g,6.9 mmol) and [Fe,(CO),,] (1.16 g, 2.3 mmol) dissolved inheptane (100 cm3) were heated and stirred under reflux for2.5 h, during which time the solution slowly changed fromdark green to orange.After cooling, the solution was filteredgiving a brown solid residue and a deep orange filtrate whichwas evaporated to dryness affording a dark brown solid. Thelatter solid was chromatographed on a 2.5 x 20 cm columnof TLC-grade silica giving a purple band preceded by a smallyellow band. Elution with pentane and removal of solventgave yellow crystals from the first eluate (unreacted dibenzo-tellurophene) and a red powder from the second, recrystal-formula C,,H,,FeO,, yield 0.344 g (15.4%) based on C,H,Te(Found: C, 65.2; H, 4.80. C,,H,,FeO, requires C, 65.5; H,4.60%). EI mass spectrum: mi. = 348 (Mf), 56Fe. NMR(CDCI,): 'H, 6 6.98 (m, 4 H), 4.74 (d, 2 H), 4.43 (d, 2 H), 3.67(m, 2 H) and 3.05 (m, 6 H); 13C, 6 212.1, 155.0, 154.6, 135.6,128.8-125.8, 102.2 and 40.5-29.8.Infrared (KBr, cm-'): v(C0)2043, 1981 and 1950. Crystals suitable for X-ray diffractionwere grown by cooling a concentrated hexane solution.The unstable yellow oil, yield 0.334 g, gave an infraredspectrum showing the presence of both organic fragmentsand of carbonyl groups: v(C0) 2051, 1983 and 1962 cm-'.NMR (CDCI,): 'H, 6 7.45 (m, 2 H), 7.36 (m, 2 H), 2.44(d, 2 H) and 0.22 (d, 2 H); 13C, 6 132.0, 128.0, 100.0, 35.5 and29.3.The reactions of the tellurium heterocyclic compounds areshown in Scheme 1.With thiophene and benzothiophene. The reaction wasrepeated using the conditions reported by Stone and co-workers.' and in concurrence with those workers after 15 h ofreaction a 5% yield of the ferrole C,H,Fe,(CO), was obtained.Following chromatography of the reaction mixture, anadditional product, C,H,Fe(CO),*Fe,(CO), [trace, El massspectrum m/z = 444 (M'), 56Fe] was isolated.The reaction was then repeated by sealing [Fe,(CO), ,] (1 .Og), thiophene (15 cm3) and Fe,O, (0.5 g) in a Teflon containerand heating the contents in a Sharp Carousel microwave ovenfor 50 min on a 'high' power setting. Work-up affordedidentical products in similar yields.A similar microwaveexperiment using benzothiophene gave a 49% yield of thebenzothiaferrole after 50 min identical to that obtained byRauchfuss and co-workers after 18 h; however microwaveheating was not successful in promoting a desulfurisationreact ion.Physical measurementslisation of which from boiling heptane gave violet crystals,yield 0.28 g (28% based on [Fe,(CO),,]), m.p.176 "C, identi-Fe,O, requires C, 50.1; H, 1.85%). CI mass spectrum: m/z =432 ( M ' ) , 56Fe. NMR (CDCI,): 'H, 6 7.33 (m, 4 H), 7.25(m, 2 H) and 7.15 (m, 2 H); I3C, 6 144.1, 128.6, 128.2, 127.9,127.7 and 119.6. Infrared (KBr, cm-'): v(C0) 2060m, 2025s,Proton, 13C, and '"Te NMR spectra were obtained with aBruker AC-300 instrument for CDCI, solutions using SiMe,with a Bio-Rad FTS-40A spectrometer and EI and CI massspectra from the EPSRC service at University College,fied as a dibenzoferrole (Found: C , 50'2; H, C18H8- (lH, 13C) and Me2Te (12'Te) as standards, infrared spectra1992s, 1974s and 1884m.Crystals suitable for X-ray diffractionmeasurements were grown by cooling a concentrated hexanesolution.With 2-telluraindane. 2-Telluraindane (1.93 g, 8.3 mmol)and [Fe,(CO),,] (2.8 g, 5.3 mmol) were dissolved in heptane(100 cm3) and heated, with stirring, under reflux for 3 h duringwhich time the solution changed from green to deep purple.After cooling the solution was filtered affording a black residueand an intense purple solution. The solution was taken todryness under vacuum leaving a purple solid and a red oil. Themixture was dissolved in pentane (10 cm3) and chromato-graphed on a 2.5 x 20 cm column of TLC-grade silica, therebygiving a small orange band preceded by a small yellow and largepurple band. Elution with pentane and removal of solvent gavepurple crystals from the first eluate and a red-brown solid fromthe second, m.p.76 "C; a third component. an unstable yellowoil as yet unidentified, was also eluted. The purple crystalsshowed infrared absorption only in the carbonyl region, gaveno proton NMR resonances and were identified as thepreviously reported cluster compound [Fe,Te,(C0),],8 yield1.13 g, 40% (based on C,H,Te) (Found: C, 16.6; H, 0.10.C,Fe,O,Te, requires C, 16.0; H, 0.00%). I3C NMR (CDCI,): 621 1.6 and 209.3. EI mass spectrum: m/z = 680 ( M ' ) , 13'Te,56Fe. Infrared (KBr, cm-'): v(C0) 2077m, 2052s, 2020s, 1998s,1976s and 1960s.The red-brown solid (4) was shown to have the molecularCrystallographygroup Pi, u = 7.357(2), b = 8.386(8), c = 13.585(3) A,Crystal data.C,,H,Fe,O, 3, M , = 431.95, triclinic, space91.49(4), p = 92.10(10), y = 102.21(2)", U = 818 A3, Z = 2,D, = 1.753 g cm ', F(OO0) = 432, p(Mo-Ka) = 1.806 mm-'.C,,H,,FeO, 4, M , = 348.17, monoclinic, space group P2,/c,u = 19.748(1), b = 6.904(1), c = 12.342(3) A, f3 = 107.34(1)",U = 1606 A3, Z = 4, D, = 1.1440 g cm ,, F(OO0) = 720,~(MO-KR) = 0.950 mm '.Cell dimensions and intensity data were measured with anEnraf-Nonius CAD4 diffractometer operating in the w 2 8 scanmode using Mo-KK radiation. The angular range for datacollection was 2-25" for both compounds. 5733 Reflections for3 and 5877 for 4 were scanned, of which 2867 (Rint = 0.037)and 2818 (0.034) were unique. Three standard reflections weremeasured every 2 h to check the stability of the system. Small(1%) decay corrections were applied to the data.Bothstructures were determined by direct methods with SHELXSand refined by least squares using anisotropic thermalparameters for Fe, C and 0 atoms. Hydrogen atoms wereplaced in calculated positions, riding on their respectivebonding atoms.Structure 3 was refined using the TEXSAN package,23 thefunction minimised being Ccv(IF,I - IFc1)2, w = l/02(F,); R,R'0.048, 0.049 for 2474 reflections with I > 2.5o(I). Residualelectron density was within the range +0.9 to -0.9 e A '. An=1546 J. Chem. SOC., Dalton Trans., 1996, Pages 1545-154empirical absorption correction was applied using DIFABS,24transmission factors in the range 0.51-1.00. Structure 4 wasrefined using SHELXL 93;25 the function minimised wasCw(FO2 - f-L2)2, LV = l/[02(Fo2) + 0.004P2 + 0.13P], whereP = :Fo2 + iFc2; final R,R' 0.05, 0.127 for all reflections;residual electron density was within the range + 0.23 to - 0.42 eA '.Diagrams were drawn with Atomic coordinatesare listed in Tables 3 and 4.Complete atomic coordinates, thermal parameters and bondlengths and angles have been deposited at the CambridgeCrystallographic Data Centre. See Instructions for Authors,J . c'lzc~m. Soc.. Dulron Trans., 1996, Issue 1.DiscussionCoal. like many materials of natural origin, is a difficultsubstance with which to work. Recently, however, it has beenestablished that microwave heating can be used with benefitin some aspects of coal science.For example derivatisationreactions of coal functional groups have been greatlyaccelerated leading to the development of new analyticalmethods.2- At the outset of this work it was debated thatsimilar acceleration of the desulfurisation reactions ofthiophene rings using [Fe,(CO), 2] might be possible and,further, that in the case of benzothiophene the reaction mightbe driven to the desulfurisation stage. The classic reaction ofStone and co-workers' was repeated and found to proceedprecisely as reported, although chromatography did reveala trace of a compound not previously reported (identified bymass spectroscopy), namely C,H,Fe(CO),*Fe(CO),. Whenthe reaction was repeated in a sealed Teflon container in amicrowave oven the yield of the same products obtained after50 min was identical to that obtained after 15 h of bench reflux.Thiophene does not heat rapidly in a microwave field of 2.45GHz but the addition of 0.5 g of otherwise inert Fe,O, to thereaction vessel ensured rapid heating." Similar experimentswith benzothiophene also produced the same yield of thethiaferrole after 50 min to that obtained by Rauchfuss andco-workers' after 18 h, however, no desulfurisation of thebenzothiophene was noted.The use of microwave heating evidently leads to accelerationof previously reported reactions of thiophene derivatives with[Fe,( CO), J .Since benzothiophene and dibenzothiophene arebetter models for sulfur in coal and in coal-derived liquids,"the failure to observe the desulfurisation of benzothiopheneunder microwave conditions was disappointing.Attention wasdirected to reactions of tellurium heterocyclic compounds forreasons stated in the Introduction. All such reactions aresummarised in Scheme 1.Oefele and Dotzauer 29 were, to our knowledge, the first toconsider reactions of tellurophene with metal carbonyls. Inparticular they treated [Fe,(CO), ,] with tellurophene inbenzene to obtain black [Fe,Te,(CO),] and the yellow ferroleC,H,Fe,(CO),, together with an 18% yield of a third materialbelieved to be C,H,FeTe-Fe(CO),. On repeating the reactionunder the conditions described in the Experimental section(heptane, 2.5 h) a 45% yield of the ferrole was obtained.Earlier termination of the reaction gave the telluraferrole(found to melt at 41 "C) in 66% yield.The black residue fromthe 2.5 11 reaction was, in our case, FeTe which doubtlessarose from [Fe,Te2(C0)9] via loss of CO and Fe(CO),.Encouraged by the good yields from the tellurophenereaction, attention was turned to dibenzotellurophene withno less success; thus the dibenzoferrole, C, ,H,Fe,O,, wasobtained in 28% yield. In this case no intermediate wasisolated. but extraction of soluble material from the solidresidue remaining in the reaction flask gave a black substancegiving no IR absorptions above 400 cm and shown tocontain only Fe and Te (electronic spectroscopy for chemicalanalysis. ESCA).1 234Scheme 1 Reactions with iron carbonylTable 1 Bond lengths (A) for complex 3Fe( 1 )-C( 1 7)Fe( 1 )-C( 18)Fe( 1 )-C( 16)Fe( 1 )-C( 6)Fe( 1 )-C( 8)Fe( 1 )-C( 1 )Fe( 1 )-C( 7)Fe( 1 )-Fe( 2)Fe(2)-C( I 3 )Fe( 2)-C( 1 5)Fe( 2)-C( 14)Fe( 2)-C( 8)Fe( 2)-C(6)Fe( 2)-C( 1 6)O( 1)-C( 13)o(2m I 4)o ( 3 ~ 15)1.767(5)1.767( 5 )1.796( 5 )2.122(4)2.126(4)2.2 1 6( 4)2.21 6(4)2.468( 1 )1.787(5)1.807(4)2.002(4)2.005(4)2.320(5)1.129(5)1.150(5)1.773(5)1 .I 39( 5 )1.149(5)1.140( 6)I . I36(6)1.41 6(6)1.454( 6)1.362(7)1.398(9)1.361(7)1.431(6)1.41 8(7)1.409(6)I .448( 5 )1.358(7)1.382(7)I .363(6)1.424(7)To maintain the theme of reactions of heterocycliccompounds of tellurium with [Fe,(CO),,], but to model athioether rather than an aromatic environment, 2-telluraindanewas used as a substrate.As detailed in the Experimental section,tellurium was removed from the ring and, in this case, recoveredas the cluster compound [Fe,Te,(CO),] initially reported byHieber and Gruber.,' In addition, a 15% yield of a novelcomplex C,,H,,FeO, 4 was obtained which was shown by X-ray crystallography (see below) to contain an organic moietyderived from the unsymmetrical coupling of two CsH,fragments (see Scheme 1). This is in contrast to the formationof a benzocyclobutane following thermochemical extrusionof tell~rium.~' It is to be expected that 2-telluraindane willbe a better Lewis base than tellurophene, hence it is possiblethat the initial stage of the reaction is the monodentate co-ordination of two molecules of the base to the two equivalentiron atoms of [Fe,(CO),,] thus providing a starting pointfrom which the observed products might plausibly emerge.Theretention of an exocyclic double bond is of interest; doubtlessthe co-ordination of that fragment of the molecule to the irontricarbonyl unit is responsible for the stabilisation of this form.Indeed, it is possible that in an intermediate stage the C,H,unit is co-ordinated as 1,6-dirnethylenecyclohexa-2,4-diene thusJ. Chem. SOC., Dalton Truns., 1996, Pages 1545-1549 154facilitating the addition of a similar fragment across oneexocyclic double bond.Structural studiesThe structure of the C,,H&,O, 3 molecule is shown in Fig. 1;bond lengths are in Table 1. The biphenyl residue is planar towithin k 0.018 8,. One of the iron atoms, Fe( 1) in our numberingscheme, is displaced by 1.67 8, from the plane and is n bonded tothe aromatic bonds C(6)-C(1) and C(7)-C(8). The Fe-Cdistances involving inner carbon atoms, C(1), and C(7), at2.216(4) 8, are significantly longer than those involving the outeratoms, C(6) and C(8), at 2.122(4) and 2.126(4) A, respectively.Atom Fe(2) lies close to the biphenyl plane, displacement 0.22 A,and is CJ bonded to C(6) and C(8), forming a metallocyclic ring.The Fe-C bonds at 2.005(4) and 2.002(4) 8, are shorter than then interactions.The central C( 1)-C(7) bond is 1.454(6) 8,, some 0.03-0.05 8,shorter than is commonly found in biphenyls, and may indicatesome inter-ring electron delocalisation, presumably triggeredby the n bonded atom.The Fe(1)-Fe(2) distance is 2.468(1) A,corresponding to a bonding interaction. The carbonyl groupC(16)-0(4) is involved in bonding to both Fe(1) and Fe(2),Fe(1)-C(16) 1.796(5) and Fe(2)-C(16) 2.320(5) A, forming anunsymmetrical bridge. This Fe,CO iron carbonyl system isbent, angles Fe( 1)-C( 16)-0(4) and Fe(2)-C( 16)-0(4) 158.4(5)and 129.2(4)', respectively. The other iron carbonyl groupingshave normal dimensions, Fe-C 1.767-1.807, C-0 1.129-1.150A, Fe-C-0 177.5-179.0'. The longest Fe-CO bonds, apartfrom those of the bridging carbonyl group, are those trans tothe C(6)-Fe(2) and C(8)-Fe(2) G bonds, indicating a possibleweak trans influence of these bonds. The C-0 bond of thebridging carbonyl group at 1. I49(5) 8, falls within the range oflengths of the other, non-bridging carbonyl groups.Both ironatoms obey the 18-electron rule.The central Fe,C,(CO), residue bears a striking resemblanceto the corresponding residue in (hexacarbonylcyclododeca-177-diyne)diiron the crystal structure of which has beendetermined.,* Here the Fe-Fe separation is 2.462(3) 8, and theFe-CO (bridging) distances are 1.75(2) and 2.32(2) 8, withFe-C-0 angles 162(3) and 125(3)'.The structure of the C,,H,,Fe03 molecule 4 is shown in Fig.2. Bond lengths and angles are in Table 2. The iron atom is co-ordinated to three carbonyl groups and to the C(5)-C(6) andC(7)-C(8) double bonds of the organic system. The Fe-COdistances average 1.787(3) 8,. The Fe-C(5), -C(6), -C(7) and-C(8) distances are 2.135(2), 2.056(3), 2.042(3) and 2.112(2) 8,respectively.These distances may be compared with thoseFig. I Structure of complex 3 showing the atom labellingfound in a selection of 21 structures containing thecyclohexadiene-Fe(CO), system extracted from the CambridgeStructural Database.33 In 18 of these structures theFe-C(diene) bond lengths follow the pattern observed in ourt::Fig. 2 Structure of complex 4 showing the atom labellingTable 2 Bond lengths (A) and selected angles (") for complex 41.783(3)1.788(3)1.789(3)2.042( 3)2.056(3)2.135(2)2.1 12(2)1.133(3)1.134(4)1. I35(3)1.327(4)1.459(4)1.5 17(3)1.424(4)C(btC(7)C(7FCG3)C@tC(9)C(9bC( 18)C(9FC( 10)C(l0)-C(l1)C(ll)-C(12)C( 12)-C( 17)C( 12)-C( 13)C(13)-C(14)C( 14)-C( 15)C( 15)-C( 16)C( 16)-C( 17)C( 17)-C( 18)1.397(4)1.404(4)1 .529( 3)1.5 18(3)1.554(3)1.523(4)1.496(4)1.389(4)1.414(4)1.365(5)1.383(5)1 .3 74(4)1.389(4)1.520(4)178.4(2) C(7)-C(8)-C(9) 118.5(2)179.3(3) C(4)-C(9)-C( 18) 1 13.1(2)178.4(3) C(4)-C(9)-C(8) 106.9(2)120.8(3) C(18)-C(9)-C(8) 11 1.7(2)125.1(3) C(4)-C(9)-C( 10) 109.2(2)122.4(2) C(8)-C(9)-C( 10) 108.7(2)114.7(2)1 14.1(2) C(18)-C(9)-C(lO) 107.3(2)Table 3 Atomic coordinates ( x lo4) for complex 3Atom X Y1770(1)1991(1)5028(2)388(2)- 660(3)- 1881(2)273(4)1466(2)4368(3)5243(3)470 1(4)3288(4)2396(3)29 1 O( 3)4783(3)3666(2)3990(3)5327(3)6406(3)6 1 54( 3)38 52( 3)1020( 3)357(3)- 340(3)850(4)1 564( 3)7250( 1 )4394( 1 )2682( 2)3061(2)19 1 7(3)5439(2)8 7 39(4)9883(2)7063(3)7981(3)75 14(4)61 35(4)5247(3)5663(3)7430(3)6337(3)6646(3)7928(3)8968(3)8764(3)3606(3)2913(3)5871 (3)8 179(4)8854(3)3354(3)22559( 1)22 18( 1)2334(1)274(1)3 225( 2)1886(2)42 14(2)1259( 1)3 3 8 3( 2)4224(2)5 139(2)5252(2)4456(2)3474(2)2364(2)1670(2)640(2)38 l(2)1083(2)2064(2)2282(2)1018(2)28 36( 2)2 1 30(2)3 5 52(2)1773(2)1548 J.Chem. SOC., Dalton Trans., 1996, Pages 1545-154Table 4 Atomic coordinates ( x lo4) for complex 4x1 156( 1)1789( 1)628(2)1538(2)295(2)831(2)2335( 1)1585( 1)1294(2)1578(1)2118(1)2701(1)3427(2)3829( 1)4324(2)4709(2)4618(2)41 34( 1)3737( 1 )3227(2)2643(2)-248(1)3120(2)Y1687(1)4653(3)1421(4)4070( 5 )3523(4)1523(4)3 120( 5 )- 127(3)-601(4)- 1267(4)-417(4)942(4)430(3)- 1345(4)- 946(4)92 l(4)1263(5)2935(5)4324(5)4040(4)2349(4)2076(4)- 178(5)Z1534(1)3221(2)1854(2)2557( 2)1723(2)285(2)3183(2)270 1 (2)15632)769(2)12 19( 2)2308(2)2070(3)1098( 3)1260( 2)656(3)- 495(2)794( 3)1 547( 3)2 1 34( 2)1996(2)2695(2)4297(2)structure, with bonds to the inner carbon atoms (mean over all21 structures 2.05 A) shorter than those to the outer carbonatoms (mean 2.1 1 A).The C-0 lengths average 1.134( 1) 8, andthe Fe-C-0 angles are all within 1.6" of being linear. As in 3,here also the iron atom obeys the 18-electron rule.The bonds C(5)-C(6), C(6)-C(7) and C(7)-C(8) involvingthe complexed diene moiety are 1.424(4), 1.397(4) and 1.404(4)8, respectively, essentially.equal, intermediate between singleand double bond distances. It is, nevertheless, of interest thatin 17 of the 21 structures cited above the central bond of thediene system is slightly shorter (mean over 21 structures 1.40 A)than the outer two bonds (mean 1.42 A), in good agreement withour results, indicating a tendency to bond alternation of theopposite sense to that of the parent free ring system. Thiseffect had been noted34 previously, but not considered to besignificant. Reverse bond alternation is more pronounced instructures where there is a cyclopentadienyl ligand co-ordinatedto the iron atom trans to the cy~lohexadiene.~~ BondsC(8)-C(9) and C(4)-C(9) are purely single, while C(4)-C(5)(1.459 A) is slightly shorter than a single bond and C( 19)-C(4)(1.327 A) is very slightly longer than a double bond.Thus theC( 19)-C(4) double bond appears to be partly conjugated withthe irondiene system.The complexed cyclohexadiene-like ring has atoms C(5)--C(S)coplanar to within k 0.005 8, with C(4) and C(9) displaced on thesame side of this plane by 0.725 and 0.974 A, respectively. Theiron atom lies 1.655 A on the opposite side of this plane. Thecyclohexene ring C(9)-C( 18) adopts the half-chair conform-ation. Atoms C(l l ) , C(12), C(17) and C(18) are coplanar towithin k 0.004 A and C(9), C( 10) are located on opposite sides ofthis plane at distances of 0.35 and 0.44 A.ConclusionWhilst microwave heating is successful in accelerating reactionsof thiophene and benzothiophene with [Fe,(CO), J, it isunable to induce sulfur removal from the thiaferrole formedfrom benzothiophene.Tellurium is readily removed fromtellurophene and, more importantly, from dibenzotellurophene,thus demonstrating the mechanistic feasibility of removal ofchalcogen from more condensed systems. Tellurium is alsoremoved from a cyclic telluride, but the major point of interestis an unsymmetrical dimerisation of two organic fragments,which is in contrast to the product of thermal decomposition ofthe telluride.References1 H. D. Kaesz, R. B. King, T.A. Manuel, L. D. Nichols andF. G. A. Stone, J. Am. Chem. Soc., 1960,82,4749.2 W. Hieber and C. Scharfenberg, Ber., 1940,73, 1012.3 R. B. King, P. M. Triechel and F. G. A. Stone, J. Am. Chem. Soc.,4 G. Dettlof and E. Weiss, J. Organomet. Chem., 1976, 108, 213.5 P. Huebener and E. Weiss, J. Organomet. Chem., 1977,129, 105.6 T. B. Rauchfuss, Prog. Inorg. Chem., 1991,39,259.7 R. J. Angelici, Coord. Chem. Rev., 1990, 105, 61.8 A. G. Ogilvy, M. Draganjac, T. B. Rauchfuss and S. R. Wilson,9 J. Chen, L. M. Daniels and R. J. Angelici, J. Am. Chem. Soc., 1991,10 X. D. Hu, L. Zhang, S. M. Cui, Z . H. Wang and J. S. Gao, Fuel Sci.11 J. D. Goodrich, P. N. Nickias and J. P. Selegue, Inorg. Chem., 1987,12 K. M. Rao, C. L. Day, R. A. Jacobson and R. J. Angelici, Inorg.13 S.Harris, Organometallics, 1994, 13,2628.14 S. Luo, A. E. Ogilvy, T. B. Rauchfuss, A. L. Rheingold and15 D. M. P. Mingos and D. R. Baghurst, Chem. SOC. Rev., 1991,20, 1.16 F. Fringuelli and A. Taticchi, J. Chem. Soc., Perkin Trans. 1 , 1972,17 M.-G. Choi and R. J. Angelici, J. Am. Chem. Soc., 1991, 113,18 C. J. White, R. J. Angelici and M.-G. Choi, Orgunometallics, 1995,19 W. Lohner and K. Praefke, Chem. Ber., 1978,111,3746.20 J. D. McCullough, Inorg. Chem., 1975,9, 2285.21 A. Z . Al-Rubai, W. R. McWhinnie, P. Granger and S. Chapelle,22 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990,46,467.23 TEXSAN, Single Crystal Analysis Software, version 1.6, Molecular24 N. Walker and D. Stuart, Acta Crystallogr., Sect. A , 1983,39, 158.25 G. M. Sheldrick, University of Gottingen, 1993.26 W. D. S. Motherwell and W. Clegg, University of Cambridge,27 P. Monsef-Mirzai, W. R. McWhinnie, M. C. Perry and P. Burchill,28 P. Monsef-Mirzai, M. Ravindran, W. R. McWhinnie and P.29 K. Oefele and E. Dotzauer, J. Organomet. Chem., 1972,42, C87.30 W. Hieber and J. J. Gruber, 2. Anorg. Allg. Chem., 1958,2%,91.31 E. Cuthbertson and D. D. MacNichol, Tetrahedron Lett., 1975,32 H. B. Chin and R. Bau, J. Am. Chem. Soc., 1973,95,5068.33 F. H. Allen, J. E. Davies, J. J. Galloy, 0. Johnson, 0. Kennard,C. F. Macrae, E. M. Mitchell, G. F. Mitchell, J. M. Smith andD. G. Watson, J. Chem. In$ Comput. Sci., 1991,31, 187.34 F. A. Cotton, V. W. Day, B. A. Frenz, K. I . Hardcastle andJ. M. Troup, J. Am. Chem. Soc., 1973,95,4522.35 M. R. Churchill and R. Mason, Proc. R. Soc., London, Ser. A, 1967,301,433.1961,83,3600.Organometallics, 1988,7, 1 17 1.113,2544.Technol. Int., 1995, 13, 16 1.26, 3424.Chem., 199 1,30,5046.S. R. Wilson, Organometallics, 1991, 10, 1002.199.565.14, 332.J. Organomet. Chem., 1982,234,287.Structure Corporation, The Woodlands, TX, 1993.1988.Fuel, 1995, 74, 675.Burchill, Fuel, 1995, 74, 20.1893.Received 26th June 1995; Paper 5/04116CJ. Chem. SOC., Dalton Trans., 1996, Pages 1545--1549 154