u: 2...; 9 P. . :er «I; v: uu P. V ; IVE} n3. :- r. 4?? 23.»; vi v w) 1771.13 v1 7 . 1v .11 :5 , 2.2. . . EMF: 2. :2rfii n ~ ,2 t. 2.713;; t. 32.... V r; r. 71. 45.7.2.7]. :71; i .755; , M: .. 3 .15).. 11;; ‘3 zun ,. iii... "are 1 e .2, 113.: L5 .. .. .13.?! . «EXLS’ "é“‘fl-wa-v-x? av as 1... . 5.: 3’13):va p316... :33r4 . a by 3. . 73.: 13:. :5. . . f... .vté! i; 1.3... , ‘ -r|~ :IIWQD‘UUI-uu-olnu m w-v-x FE?!§§b$VI-1‘!. ‘1‘? Q? 1.99:9”? ».u.ava ..fl.....,?m.,..‘:_.a . A , J _ . ‘ .. .fimwwéifim fiflpgr .. ~ 2 mg,” . .. . s, . a. _, ._ mflfiafigég my: . a” ., . f. J ; :s a . , Linnm‘vfi I}. , 4 This is to certify that the 3‘ thesis entitled v TRANSITION METAL ORGANOMETALLIC CHEMISTRY OF TUNGSTEN AND NICKEL presented by Patrick Lee Burk has been accepted towards fulfillment of the requirements for Ph.D. Chemistry degree in Major professor Date July 31, 1975 0-7 639 1 ‘ IIIDIII Iv “ -l|AE-&.SIIIIS' ‘ III“ I“ III}. -' ”IRA" INDEIS ‘ SPIIIEPIII'JIGIIIII g‘ 2 ABSTRACT TRANSITION METAL ORGANOMETALLIC CHEMISTRY OF TUNGSTEN AND NICKEL BY Patrick Lee Burk The mechanism of olefin metathesis was investigated by using the two catalyst systems butyl lithium—tungsten hexachloride and phenyltrichlorotungsten—aluminum chloride. To distinguish between the "pair—wise" and the "carbene" mechanisms the metathesis of mixtures of l,7-octadiene-l,- 1,8,8-g4 and l,7—octadiene were carried out. The ratios of Q4 : d2 d product ethylene produced were 1 : 3.4 : 3.1 with the tungsten hexachloride catalyst and l : 2.2 : 1.4 with the phenyltrichlorotungsten catalyst. These ratios are most consistent with the proposed "carbene" mechanism. The tetramethylene metallocycles bis——_ ——> J‘- (l) I Early investigations demonstrated the reaction proceeds via scission of the olefinic bond as depicted in eqn. 1, rather than through a transalkylation as shown . 2 3 in eqn. 2. ' i l R dizzzca R RlCH =cn—f——rz2 l ' ——> ____ __.+ _ _.___ T (2) R _1_. _ 1 : CH-- CHRZ Rl CH ::::CHR I 2 l l 3 In one study a 1:1 mixture of 2-butene-d_8 and 2-butene was added to a tungsten metathesis catalyst (eqn. 3). The only product found was CH3CH=CDCD as 3! eqn. 1 predicts. Products derived from cleavage of the carbon-carbon single bonds alpha to the double bond, e. ., CD3CH=CHCH3, were not observed. CD3CD CH CH CD3CD + WC16 - CZHSOH ——————————————__—p. .‘_._ ___________ (3) c2H5A1C12, Benzene CD3CD CH3CH CH3CH Once the bonds involved in the transalkylidenation were established, a concerted mechanism with a four—centered 2’4_7 The transition transition state was proposed (eqn. 4). state for this one step transformation is characterized by having all four carbons equally related to the metal. I... ~ 1 »~ 0‘ The formation of this "quasicyclobutane" transition state and its transformation into a bis olefin complex were viewed as being cycloaddition reactions. From the principles of conservation of orbital symmetry8 of Woodward and Hoffman, these two cycloadditions are formally (2S + 25) transfor- mations, which are thermally forbidden. l Mango 0 postulated it should be possible to switch from a symmetry-forbidden to a symmetry—allowed reaction, if a transition metal complex were used which had atomic d orbitals of the proper symmetries, and an available electron pair. l m Figure l is a correlation diagram for the cyclo- butanation of the ligand-bound olefins of complex 1, which "\1 describes an orbital pathway for the placement of electron pairs into the cyclobutane sigma bonds and for the removal of the appropriate olefin pi electrons. METAL AA‘—————-—————————__________________________ AA COMPLEX AS CYCLOBUTANE SA O*—BONDS ANTIBONDING SA — COMBINATIONS ss —— AS 55 METAL +l— AA ORBITALS METAL ILA—1’1." COMPLEX AS +£— BONDING SIX—1+ COMBINATIONS ss 4+ %\/ —-1-+- AS CYCLOBUTANE +L— ss O—BONDS Figure 1. A correlation diagram for the cyclobutanation of the ligand-bound olefins of complex 1. ’\J Figure 2 shows the orbital interactions which allow the metal to remove the symmetry restrictions on the reaction through a relocalization of ligand-metal AS and SA electron density.11 1.: Eli + n* + n*]s + 0*] dZY Clzx AS +1r— n] dzx AS Figure 2. The removal of symmetry restrictions through a relocalization of ligand-metal AS and SA electron density. 12 In discussing the cyclobutane theory, Petitt noted that the free energy of formation of one molecule of cyclobutane and two molecules of ethylene are very close. Therefore he concluded, if the transition state resembles cyclobutane, then cyclobutane should be one of the products, especially if there is no strong bond between cyclobutane and the metal. Conversely, if cyclobutane was exposed to metathesis conditions, then ethylene should be formed. Experiments were conducted using ethylene and cyclobutane as substrates, and the known metathesis catalyst molybdenum on A1203. Only very small amounts of cyclobutane (<0.1%) and ethylene (3%), respectively, were found. This contrasts with monodeuteroethylene, which under similar conditions, underwent metathesis with itself such that at least 35% of the molecules present entered into a reaction which lead to metathesis products. Petitt suggested there were other more likely orbital symmetry "allowed" pathways. His alternative was a transition state which is a multicentered organometallic system "in which the bonding is most conveniently described as resulting from the interaction of a basic set of metal atomic orbitals and four methylenic units" (eqn. 5). In 1969 and 1970, studies of the isomerization of exg-tricyclo(3.2.1.02’4)octene 2 by Katz,13 and the valence isomerization of cubane by Halpern and Eatonl4 were made. Until these studies, valence isomerizations were mechanis— tically grouped with olefin metathesis as being concerted reactions. A mixture of products consisting of 62% 3a, 32% 4, ’b "b and 6% 5 was formed when 2a was warmed with tris(triphenyl- "u '\. phosphine)rhodium(I) chloride (eqn. 6).13 Although 4 is 'h formally related to 2 by an electrocyclic reaction, 3 is not. X (Ph3 P) 3RhCl X 2 m xendo3 m m a,x=H b,x=D When deuterated starting material 2b was treated with catalyst, the deuterium was found exclusively at C-2 and C-6 in 3b. The metallocycles 6 and 7 were proposed to ’b m ’\1 account for the location and stereochemistry of the deute— rium atoms in 3b. Figure 3 is a scheme which shows how ’\2 metallocycles 6 and 7 could lead not only to 3 but also to 4 and 5. ’\J ’b Shortly after the work by Katz, Halpern and Eaton treated cubane with a rhodium catalyst to give a syn— tricyclooctadiene (eqn. 7).14 EI—afil ET R= COOCl-I3 29% Figure 3. The mechanism of the catalyzed valence isomer- ization of exo-tricyclo(3.2.1.02I4)octene. When a stoichiometric amount of the rhodium catalyst was added to cubane, the expected tricyclooctadiene was not formed, rather the organorhodium complex 3 was isolated and characterized in better than 90% yield (eqn. 8). The complex 3 and arguments based on rate data found in the paper led the authors to conclude that valence isomerization of cubane was multi-step process which included an organorhodium metallocycle. [Rh (c0) 2C1] 2 - (8) is 0 O CC18 '\; 10 The isomerizations of 2 and cubane both involved metal—carbon sigma bonded intermediates, not a direct metal- catalyzed electrocyclic rearrangement. This led Grubbs15 to advance a mechanism for olefin metathesis which involved a carbon—metal sigma bonded intermediate. It was formulated as "(a) a rearrangement of the complexed olefins to a metallocyclic intermediate followed by (b) a rearrangement of the metallocycle and (c) reversal of step a" (eqn. 9). He felt the rearrangement, step b, might involve the formation of a symmetrical intermediate. R R M 5: H 4.35:1: MR The mechanism which Grubbs proposed was based on the following data. When l,4-dilithiobutane was combined with tungsten hexachloride, ethylene was produced quantitatively (eqn. 10). HZC CH Li wc16 f E C ———-> —> /\ /\i (10) Li W HZC W CH2 | I C1 C1 4 4 11 When the labeled l,4—di1ithiobutanes 2 and 10 were ’b added to tungsten hexachloride, a mixture of ethylene-d 6%, 6%, and —d 88% was obtained (eqn. 11). —0 —1 Li CHD=CH 88% + 2 D _—> CH =CH 6% + 2 2 “——‘——‘ (ll) CDH=CDH 6% D L1 10 ’b The trans stereochemistry of the ethylene—g when 2! the d,l-isomer 9 was used, and the cis stereochemistry when ’M the meso isomer 10 was used, is consistent with a metallo- '\/ cycle intermediate as shown in Figure 4. D D D D “’°—*[ )- ——’//\/\\ M M D M D D f D D —>°—>/S -—>//\,\ M M "”D M D Figure 4. Rearrangement of metallocycles to give product ethylenes. 12 If the ethylene—d was arising from metathesis of ethylene-d1, which could be formed by a Grob—type fragmen- tation of an acyclic intermediate, then the stereochemistry of the ethylene-d would be independent of 9 and 10 —2 m m (eqn. 12). D D CHD CHD Grob fra entation metathesis gm ——> ( l 2 ) WCl WCl 5 5 CH2 CHD Another possible nonconcerted pathway for olefin metathesis, initially proposed by Lappertl6 and more recently advanced by Casey17 involves a transition metal— carbene complex. Casey prepared (diphenylcarbene)penta— carbonyltungsten(0) L2, a carbene complex which is not stabilized by electron donating heteroatoms18 attached directly to the carbene carbon atom. When L2 was heated in the presence of isobutylene, a 76% yield of 1,1—diphenyl- ethylene was obtained (eqn. 13). Ph Ph Ph :>== Ph >:WC05 ————>° % + (13) Ph 100 Ph 12 76% m 13 Based on this metathesis type transalkylidenation product, Caseyl7 proposed a scheme, shown in Figure 5, for olefin metathesis which includes the metallocyclobutane intermediate L3. The carbene mechanism is unique in that all of the previously proposed mechanisms, whether concerted or nonconcerted, involved a "pair-wise" interchange. The carbene mechanism instead offers a chain transfer of methylene groups. R R R CR2=CR M R —-——-> :==1 M '——‘-’ CH 2 M I _CH 2 4-— H R R R H R R R 1,3 M CR + M ——— ll ‘\][/ ___CR2 —————.- 2 CH 2 H C 2 CR2 Figure 5. Proposed carbene mechanism for olefin metathesis. The goal of the work in our laboratory was to develop a set of experiments to distinguish between the "pair-wise" interchange mechanism and the methylene chain transfer mechanism. In order to distinguish between these mechanisms, the metathesis of mixtures of 1,7-octadiene— l,l,8,8—d4 l4 and l,7-octadiene 15 was carried out. _ ’\1 ’b 14 l,7-Octadiene under metathesis conditions gives cyclohexene and ethylene. One can predict a priori the ethylene—d z—d2:-d0 ratios of the metathesis of mixtures of 14 and 15 by exam— ’b ’b ining the possible "pair-wise" interchange mechanisms shown in Figure 6. A g4:§2:g0 ratio of 1:0:1 18 predicted 1f kex>>km under nonequilibrating conditions. Ethylene—d could be produced if k >>k . The d :d :d ratio with k >>k is m ex —4 —2 —0 m ex calculated to give a maximum ratio of l:l.6:l. See Appendix A for an explanation as to how this maximum ratio was derived. The carbene mechanism, Figure 7, gives a statistical distribution for d4:d2:d0 of 1:2:1. 15 D D D H H k M— —>m M ‘— 11 D D l D D D D ex ex D .__l_ C2D4+ M CHD+ M\ l 2 2 2 | D Figure 6. Pair—wise interchange mechanism for olefin metathesis. Figure 7. 16 -——.> -+C=C4-M = D2C=M‘_ D 2 CD CH C N o u z + n U n o + z u N _, +C=C+M CD H C=M“ <::> D 2 CH m N o u z + 0 U n o + z n N Scheme for the prediction of product ethylene- g4:g2:do using the carbene mechanism of olefin metathesis. I 17 RESULTS AND DISCUSSION The first goal of this thesis research was the synthesis of l,7-octadiene—1,1,8,8-d4. An effective and efficient means of introducing two deuterium atoms with high isotopic purity at each terminal carbon atom, and the generation of two terminal double bonds was sought. Lithium aluminum hydride is a very effective reducing agent for converting an ester into a primary alcohol through the transfer of two hydride ions to the ester. Using this information, lithium aluminum deuteride should and does convert an ester into an alcohol in which the carbon atom containing the hydroxyl group also contains two deuterium atoms (eqn. 14). O H THF R-C—OR' + LiAlD4 __—_.. R—CDz-OH (14) 2) Workup It follows therefore, that to put two deuterium atoms at each end of a C8—Chain that the C8—diester, dimethylsuberate, be used. The readily available dimethyl— suberate upon treatment with 1.4 equivalents of lithium aluminum deuteride in tetrahydrofuran afforded after workup 1,8-OctanediOl-l,l,8,8-g4 m.p. 62-63° (lit.19 of do analog 62—63°) (eqn. 14). 18 l) LiAlD4 2) Workup O O HO OH *fi\b O//—— D2 D2 H3C CH3 The nmr spectrum (CDC13) of the diol showed a broad multiplet at 6 1.33 for the methylene protons. The location of the hydroxyl protons, a sharp singlet at 6 1.60, was determined by the loss of this singlet when the nmr solution was washed with deuterium oxide. There was no detectable absorption at 6 3.61 which corresponds to protons located on the d,w—carbon atoms in the nondeuterated 1,8-octanediol.20 A variety of conditions are available for the con- version of alcohols to olefins. Conditions for converting the diol to the diene had to be such that only terminal olefins be formed, and that loss or scrambling of deuterium be kept at a minimum. A convenient route, taking the above considerations into account, involves conversion of the diol to the diacetate followed by pyrolysis of the diacetate.21 The pyrolysis of esters to give carboxylic acids and alkenes has been known and studied for over a century. The experimental procedure is extremely simple, and the 19 yields are nearly always excellent. The cis character of the elimination is well documented with the eliminations requiring the presence of a B-hydrogen atom. The proposed cyclic nature of the elimination is shown in eqn. 15. (15) /—\3 A O>/_°\ \_____/ /_——__\ Since the diol prepared above has only one source of B—hydrogen atoms, the pyrolysis of the corresponding diacetate should lead exclusively to the desired 1,7-octa— diene. The diol was readily converted to the diacetate in 91% yield, based on dimethylsuberate, by treatment with an excess of acetic anhydride at 140°. A catalytic amount of pyridine was used to facilitate the reaction (eqn. 16). ———> (16) 20 The nmr spectrum (CDC13) of the diacetate showed two broad multiplets at d 1.30 and 1.60 which corresponded to AcO-CDz-(Cflz)6-CD2—OAC. The methyl protons of the acetate functions showed up at 5 2.00. There was no absorption at 6 3.97 indicating, that as with the diol, there was no detectable protons at C-1 and C—8.20 Pyrolysis of the diacetate was accomplished by adding the pure diester to the top of a pyrex tube packed with 48 mm. of glass helices which is heated to 580°. The products were swept from the reaction chamber by a slow stream of nitrogen, 3 l./hr., and collected in a dry ice- ethanol trap. After washing with aqueous sodium bicarbonate, the solution was extracted with pentane. The pentane was dried over molecular sieves type 4A, and it was then care— fully distilled under nitrogen to give in 49% yield l,7-octadiene-l,1,8,8—d (eqn. 17). _. (17) The nmr spectrum (CDC13) of the diene showed two multiplets which corresponded to the methylene protons of 6 1.10 and 1.70. A broad multiplet at 6 5.30 was from the 21 CE=CD2 protons. Integration of the vinyl region in the nmr showed <10% deuterium scrambling had occurred during the pyrolysis. As stated in the introduction, the metathesis of a mixture of l,7—octadiene-l,l,8,8-d_4 L4 and l,7—octadiene 15 should give data which are consistent with either the "carbene" mechanism or the "pair-wise" mechanism for olefin metathesis. A measured amount of L4 under nitrogen was added to a known quantity of freshly distilled 1?. The ratio of lé to 15 was determined by integration of the vinyl region of the nmr, and it was found to be 1:1.1. This solution of the dienes was used in all of the runs below except where noted. The first metathesis catalyst studied is formed from a 2:1 mole ratio of n-butyllithium to tungsten hexa- chloride in benzene, and it is heterogeneous in nature. The active catalyst can be used directly by adding the substrate octadiene-(d /d ) to the system with shaking at room temperature. Also, the catalyst can be isolated by centrifugation, washed repeatedly with dry oxygen-free benzene, and then used as a suspension in benzene. The ratio of ethylene-d4:d2:d was unaffected by whether the catalyst was washed with benzene prior to a run. 22 To the freshly prepared catalyst of tungsten hexachloride and butyllithium in benzene was added the 1:1.1 octadiene mixture. After shaking the mixture at room temperature over 69 min., the reaction was quenched by the addition of water. The gas above the solution was removed by syringe and the ethylene in the gas was isolated by g.l.p.c. Analysis of the ethylene by mass spectroscopy (15 eV. ionization voltage) gave a d4:d2:d0 ratio of 1:3.352t0.20 : 3.08i 0.17. An equilibrium ratio for the ethylene produced in the reaction can be calculated from the equation (1 + A)2 which yields the ratio, d4 : 2Ad2 : Azdo. With no mono- hydride equilibration, see below, A should equal lf/lf. If some equilibration was taking place, A should equal one-half the ratio of ethylene-d : ethylene-d2. Therefore for the d4:d2:d0 ratio obtained using the tungsten hexachloride-butyllithium catalyst the calculated equilibrium ratio is 1 : 3.4 : 2.9. Using the tungsten hexachloride—butyl lithium catalyst a significant amount of ethylene-d3 and —d1 was formed during the reaction. Apparently there is a mono- hydride equilibration reaction taking place in addition to olefin metathesis. Therefore, when calculating the equilib- rium ratio one must take this into account, as stated above. Figure 8 shows how a metal hydride can lead to ethylene-c13 23 and —d —l' and C D but 2H4 2 2' ethylene—d and —d1 can by simple extrapolation be The scheme uses only C generated from C2H2D2. The extent to which this reaction disturbs the ratio of primary reaction products, if at all, is unknown. _ _ -——h> _ -——D- _ M H + CD2—CD2 . M H ‘ Iii —>‘ T D = D — = CD2 CD2 2c T02 CD CDH H T -CD2=CDH + : CH2 CH2 M—H M T‘D -"> = _ —’ = etc. CH CHD H C CH ‘ CH2 CH2 Figure 8. Metal hydride scheme for deuterium scrambling in ethylene. To determine if the ethylene-d was a primary or secondary metathesis product, i;e., the metathesis product of ethylene-d and ethylene—d which are generated during the course of the reaction, another run was made. The conditions were identical to the previous run, except to the system was added a measured amount of ethylene-d at the start of the reaction. Following quenching by water, the gas above the solution was removed, and the ethylene in the gas was isolated by g.1.p.c. Analysis by mass spectroscopy gave an ethylene—d4 : -d2 : —d0 ratio of 1 z 4.14 :0.37 : 14.81 il.44. 24 The d4 : d2 ratios obtained in the presence and the absence of added ethylene are the same within experimental error. This indicates the ethylene—d was not produced by the metathesis of ethylene-d4 and —do, but rather it was a primary reaction product. Despite the monohydride equilibration reaction, the ratios presented above support the "carbene" mechanism. A metal-carbene complex could be initially formed in this system and other systems which use metal alkyls as catalyst activators by an alpha hydride elimination scheme shown in Figure 9. 2W iflfi la—hydride elimination l Hj Figure 9. Alpha hydride elimination scheme for the formation of a tungsten—carbene. 25 There is precedent in the literature for alpha hydride eliminations occurring with alkyl transition metal complexes. When Co(CH )(PPh and Rh(CH3)(PPh were 3 3)3 3)3 exposed to deuterium gas, polydeuterated methanes were formed as shown by mass spectrocopic analysis.23 The authors used the formation of an intermediate hydride carbene complex by an alpha hydride elimination step to explain the incorporation of more than one deuterium atom in the product methane. This is shown in Figure 10. D D H H D 2 I “'H I I 'HD D2 I M—H_> M—CH ——'> _= -—> _ ——> _ —>_ C3‘_| 3 <_D1~lr1CHZ_4_1\|4CH2D <_MCH2D<_I\|’ICH2D D \ D D /D M— + D D CH22 —D+—CH D M 3 Figure 10. Generation of polydeuterated methanes via an alpha hydride elimination mechanism. A new metathesis catalyst from phenyltrichloro— tungsten and aluminum chloride was next used with the octadiene—(d4/d ) solution. Phenyltrichlorotungsten(IV) F? was recently synthesized by Grahlert24 from tetra- phenyltin and tungsten hexachloride. Their procedure was repeated by us to give 16 in good yield. ’\1 26 Phenyltrichlorotungsten was investigated for its potential use as a metathesis catalyst because it is a stable tungsten(IV) complex. Several tungsten metathesis catalysts which are generated in situ including the catalyst from butyllithium and tungsten hexachloride are felt to be tungsten(IV) species. Initially, it was found that in benzene a? showed very low catalytic activity toward l,7—octadiene when compared to the butyllithium—tungsten hexachloride system. The activity of the catalyst was greatly enhanced when the Lewis acid, aluminum chloride, was added to the reaction medium. A control experiment demonstrated that aluminum chloride does not produce ethylene from l,7—octadiene. The function of the aluminum chloride is unclear. The metathesis catalyst was prepared by combining under nitrogen in a l : 1 mole ratio 1? and aluminum chloride. To this mixture was added dry oxygen-free chlorobenzene. After shaking the mixture for 30 min., it was centrifuged. The supernatant liquid contained the active catalyst. The homogeneity of the catalyst solution was verified by laser light scattering experiments. It was stated earlier, but should be noted again that the butyl- lithium—tungsten hexachloride catalyst was heterogeneous in nature.22 27 To do a run with this homogeneous catalyst, the supernatant liquid is transferred by syringe to another vessel to which is added the octadiene—(d /d ), and the resultant solution shaken at room temperature. In one run the gas above the solution was removed periodically by syringe, and the ethylene in the gas was isolated by g.l.p.c. Table 1 shows the ratios of the ethylene-d4 : -d2 : —d and how they are invariant with —2 respect to time. Table l. Ethylene Ratios from the PhWCl3-A1Cl3 and l,7—Octadiene—(d4/d0) System 1.42 :0. 1.43 i0. 1.39 i0. 1.44 i0. 39 176 373 1280 Ethylene 10 10 13 11 28 Following the last gas analysis at 1280 min., the reaction was quenched with water. The unreached diene was isolated by g.l.p.c. and analyzed by mass spectroscopy. Since the parent peak for the diene is very small, even at low ionization voltages, it is impossible to accurately determine the actual ratio of E4 : d2 : d diene present. However, it is possible to compare the relative peak intensities from m/e 114 - 110 for the starting diene and product unreacted diene. If they are the same, it may be assumed deuterium scrambling was minimal during the course of the reaction. Table 2 shows that the starting and product diene were identical. Thus, the ethylene—d2 produced does not come from l,7—octadiene—d4 or —d being first converted to l,7—octadiene—d which is then meta- thesized to ethylene—d . Table 2. Mass Spectral Analysis of l,7—Octadiene—(d4/d0) from the PhWCl3-A1Cl3 System Starting diene Product diene 29 As in the heterogeneous system, it had to be determined if the ethylene-d formed while using the homogeneous catalyst, was a primary reaction product or whether it was from the metathesis of ethylene—d4 and ~d0. The reaction conditions were the same as those described above with the homogeneous system. To the system was added at the start of the run a measured amount of ethylene—d0. The gas above the solution was removed by syringe, and the ethylene was isolated by g.l.p.c. Mass spectroscopic analysis gave a d : 92 ratio of l : 2.29: 0.19 which was identical to the ratio found in the absence of added ethylene. The run with added ethylene was quenched with water, the unreacted diene was isolated by g.l.p.c. and analyzed by mass spectroscopy. Examination of the m/e ratio of 114 : 112 : 110 of the starting diene and the unreacted product diene showed them to be the same, thus, deuterium scrambling was not occurring to a significant extent in this system (Table 3). The ratio of the Q4 : d2 : do ethylenes produced by the phenyltrichlorotungsten—aluminum chloride system is most consistent with the “carbene" olefin metathesis mech— anism. Unlike the tungsten hexachloride-butyllithium system the monohydride equilibration reaction was minimal, as evidenced by the low concentrations of ethylene—d3 and -d _1 in the system. 30 Table 3. Mass Spectral Analysis of l,7—Octadiene—(d4/go) from the PhWCl3—AlCl3 Plus Ethylene System Starting diene Unreacted product diene Since a carbene has been strongly implicated in the phenyltrichlorotungsten-aluminum chloride system, there is the problem as to how a metallocarbene can be generated in a system which does not use an alkyl lithium, tin, or aluminum to activate the catalyst. A likely possibility is that the pentametallocycle proposed initially by Grubbs15 could be the precursor to the carbene. Figure 11 shows how a pentametallocycle could generate a metallocarbene. It is interesting to note that during the metathesis of ethylene, propene is a side product, as this scheme predicts. 31 W W / § I ’CHZ Ln Ln Ln Ln /\ + w 4— II W—L / II n CH2 H CH Figure 11. Generation of a tungsten—carbene from a pentametallocycle. 32 EXPERIMENTAL Melting points were determined using a Thomas-Hoover capillary melting point apparatus and are uncorrected. Infra red spectra were recorded using a Perkin—Elmer 237B spectrometer. Nmr spectra were obtained using a Varian T—60 spectrometer with tetramethylsilane as an internal standard unless otherwise indicated. Gas chromatography was carried out using a Varian Aerograph Model 90-P gas chromatograph with 10 ft. x 0.25 in. 7% paraffin wax on alumina, 5 ft. x 0.25 in. 10% carbowax, 5% SE-30, or 10% FFAP on 60 — 80 Chromosorb W column except where indicated. Preparation of l,8—Octanediol-l,l,8,8-d4 In a flamed three-necked 500 ml. round bottom flask fitted with a dropping funnel, condenser with an oil bubbler, and a mechanical stirrer were added under nitrogen 100 ml. of freshly distilled tetrahydrofuran and 7.5 g. (0.18 mole) lithium aluminum deuteride. To the stirring solution was added dropwise a solution of 25 ml. (0.13 mole) dimethyl- suberate dissolved in 50 ml. of tetrahydrofuran at a rate sufficient to maintain a mild reflux. Following the addition of the diester, the reaction mixture was refluxed over 3 hr. Using a base workup (7.5 ml. of water, 7.5 ml. of a 15% sodium hydroxide solution, and 23 ml. of water), the aluminum salts were removed by vacuum filtration. The 33 solution was dried over molecular sieves type 4A, and a g.1.p.c. (SE—3O at 170°) of this solution indicated < 3% of unreduced diester was present. The solvent was removed in vacuo to give 25.9 g. of crude l,8—octanediol-l,l,8,8-d . An analytical sample was prepared by recrystallization of the diol from benzene—ligroin to give white crystals: m.p. 19 62 — 63 (lit. m.p. of do analog 62 - 63); nmr (CDC13) 6 1.33 (m, 12, -CE ), 1.60 (s, 2, CD OE). Preparation of l,8-Octanediol—l,l,8,8-d1 Diacetate In a 100 ml. three-necked round bottom flask fitted with a stirring magnet, dropping funnel, reflux condenser, and a glass stopper were added 25.9 g. of crude 1,8-octane— diol-l,l,8,8-d4 and 10 drops of pyridine. The mixture was heated to 140° with an oil bath, and 40 m1. of acetic anhydride were added dropwise. After refluxing over 2.5 hr., the acetic acid and the excess acetic anhydride were removed in vacuo to give a pale yellow liquid. The liquid was distilled (121°/1.5 mm.) to give 27.0 g. (0.12 mole; 91% from dimethylsuberate) of the title compound: nmr (CDC1 6 1.30, 1.60 (m, 12, —CH -), 2.00 (s, 6, cng3). 3) _2 Preparation of 1,7—Octadiene-l,1,8,8—d4 Under a 3 l./hr. flow of nitrogen, 27.0 g. of l,8-octanediol—1,1,8,8-d4 diacetate were added dropwise to a hot tube (25 mm. x 580 mm.) filled with 48 mm. of 34 glass helices at 580°. The product was washed with a saturated sodium bicarbonate solution, extracted with pentane (3 x 20 ml.), dried (molecular sieves type 4A), and distilled to give 6.5 g. (0.06 mole; 49%) of 1,7—octa— diene—1,1,8,8—d4: nmr (CDCl3) 6 1.10, 1.70 (m, 8, -C§2-), 5.30 (m, 2, CgéCDz). A multiplet at 6 4.93 which corre— sponded to the terminal vinyl protons of l,7-octadiene was also found. When this peak was integrated versus the absorption at 6 5.30, it was found about 10% of the diene had terminal vinyl hydrogens. Pure l,7—octadiene—l,l,8,8—d _4 should have no absorption at 6 4.93. Preparation of the Tungsten Hexachloride- Butyl Lithium Metathesis Catalyst Into a 25 mm. x 100 mm. test tube with an 8 mm. o.d. mouth fitted with a rubber septum were added under nitrogen 2.5 m1. of 0.1 M. tungsten hexachloride in benzene, 2.5 ml. dry oxygen-free benzene and 0.39 ml. of 2.6 M. butyl lithium. The reaction mixture was shaken for 10 min., centrifuged, and the supernatant liquid was removed by syringe and dis- carded. To the resultant solid was added 5 m1. benzene. The mixture was shaken 10 min., centrifuged, and the super- natant liquid was again removed by syringe and discarded. The washing procedure was repeated again. The washed solid was the catalyst used for olefin metathesis. It was used as a suspension in benzene. 35 Metathesis of a 1:1.1 Mixture of l,7-Octadiene-l,1,8,8—g4 14 and l,7-Octad1ene 15 U51ng the Tungsten Hexachloride-Butyl Lithium Catalyst in Benzene Into two 25 mm. x 250 mm. test tubes fitted with rubber septums were placed, under nitrogen, 5 ml. of 0.1 M. tungsten hexachloride in benzene, 5 ml. of dry oxygen-free benzene, and 0.37 ml of 2.6 M. butyl lithium. To this activated metathesis catalyst was added 1.0 ml. of a 1:1.1 mixture (by nmr) of If to 1?. To one of the tubes 1.0 ml. of ethylene was added. The mixture was shaken at room temperature for 69 min. The reaction was quenched with 2 ml. of water, and the gas above the solution was removed by syringe. The ethylene in the gas was isolated by g.1.p.c. (7% paraffin wax/alumina at 80°), and analyzed by mass spectroscopy (15 eV ionization voltage). The results are shown in Table 4. Preparation of Phenyltrichlorotungsten(IV) In a 250 ml. side—arm round bottom flask fitted with a reflux condenser with an oil bubbler, rubber septum, and a stirring magnet were placed under nitrogen 11.7 g. (0.06 mole) tungsten hexachloride and 25.7 g. (0.06 mole) tetraphenyltin. To this mixture was added 100 ml. dry oxygen—free heptane, and the mixture was heated in an oil bath. wusuxHE COHuomwu ca sumo wwuwm run: DDSUOMQ ocwawnum posmoum mamaxcum 36 Ewumhm A m\vmvtwcmflpmu00Ih.H can flqsmloaoz Eouw moflumm mamawcum .v magma 37 After refluxing over 12 hr., the solution was cooled to room temperature, filtered, and dried under a flow of argon to give the brown solid phenyltrichlorotungsten(IV). Preparation of the Phenyltrichloro— tungsten-Aluminum Chloride Metathesis Catalyst Into a 25 mm. x 100 mm. test tube with an 8 mm. o.d. mouth fitted with a rubber septum were added under nitrogen 0.368 g. (1 mmole) phenyltrichlorotungsten and 0.134 g. (l mmole) aluminum chloride. To this mixture was added 25 m1. dry oxygen-free chlorobenzene. The reaction mixture was shaken for 30 min., and then centrifuged. Unlike the tungsten hexachloridebutyllithium metathesis catalyst, the catalyst derived from phenyltrichlorotungsten was found to be soluble in the solvent, and aliquots of 5 ml. were trans— ferred to other test tubes in which the metathesis reaction was carried out. The catalyst was found to be soluble and homogeneous by laser light scattering techniques. Metathesis of a 1:1.1 Mixture of :4 to 1; U51ng the Phenyltrichloro— tungsten—Aluminum Catalyst in Chlorobenzene In a 25 mm. x 250 mm. test tube fitted with a rubber septum were placed under argon 5 ml. of the title catalyst system, and 0.4 ml. of a 1:1.1 mixture (by nmr) of If to 1?. The solution was shaken at room temperature, and periodi— cally an aliquot of the gas above the solution was removed. 38 The ethylene in the gas was isolated by g.l.p.c. (7% paraffin wax/alumina at 80°), and analyzed by mass spectroscopy (15 eV ionization voltage). Within experimental error the ratio of ethylene-d4 : -d2 : -d0 did not change with time as shown in the table below (Table 1). Table l. Ethylene Ratios from PhWCl —AlCl 3 3, l,7-Octadiene- (g4/do) System Ethylene Time (min.) d4 22 20 39 1.0 2.23 10.13 1.42 10.10 176 1.0 2.23 10.14 1.43 10.10 373 1.0 2.34 10.10 1.39 10.13 1280 1.0 2.31 10.13 1.44 10.11 After 1280 min., the reaction was quenched with 2 ml of water. Analysis of the reaction mixture by g.l.p.c. (10% carbowax at 100°) showed the reaction had proceeded to approximately 10% completion. The unreacted octadiene was purified by g.1.p.c. and analyzed by mass spectroscopy. The octadiene within experimental error was the same before and after the reaction (Table 5). mcwflp poscoum wcmflp mcfluumum 39 Emummm m H03 may mcflmo mammawqm Hmuuowmm mmmz Aom\vmvlmcmflpmu00|>.fi .m DHQMB Hqsm- 40 Metathesis of a 1:1.1 Mixture of 14 ~_——-r———_——-.-—D'- to 15 With Added Ethylene USing the Phenyltrichlorotungsten—Aluminum Chloride Catalyst in Chlorobenzene In a 25 mm. x 250 mm. test tube fitted with a rubber septmnweregflaced, under argon, 5 m1. of phenyl— trichlorotungsten-a1uminum chloride catalyst, 0.4 m1. of a 1:1.1 mixture (by nmr) of 13 to 13, and a measured amount of ethylene. The solution was shaken at room temperature, and periodically a sample of the gas above the solution was removed by syringe. The ethylene in the gas was isolated by g.1.p.c. and analyzed by mass spectroscopy. The ratio of Q4 : dz ethylene did not vary with the amount of added ethylene, nor with the amount of time added ethylene was present in the reaction medium, as shown in Table 6. In a separate experiment using the conditions above, and 0.25 ml. added ethylene, the reaction was quenched with 2 m1. of water after 1 hr. The unreacted octadiene was purified by g.l.p.c. (10% carbowax at 100°) and analyzed by mass spectroscopy. The octadiene recovered from the reaction mixture was identical to the diene starting material (Table 7). do.Oan.o Ho.owwm.o No.0HmN.o mo.onmv.o Apospouav wcwflpmuoo v0.0Hmn.o No.0HmN.o mo.onvm.o mo.onov.o Aucmuomwuv mcwflpmuoo m m I vi mcwawrum mafia Hofiml Hogan magma mflmsamca Hanuommm mmmz A %\ wvlmcmHSQDOOIk.H .k mange 41 ov.owom.v Ho.0flmm.o ma.onmm.m v0.0HmH.o .HE mm.o om.owmm.h no.0HNm.o mo.onao.m H0.0HHN.O .HE o.H ow qmmo wmw©< A.cflec mane mo DGDOE< mamasnpm mam a VI m m H cpm cweu< cum .1 m\ ecumcmflwwpoouk.a . HoH / (19) As in the catalyzed (2-+2) cycloadditions of olefins, metallocycles have been proposed as reactive intermediates in the metal catalyzed (2-+2) cycloadditions 45 of olefins and strained carbon-carbon sigma bonds. Noyori27 found that the central bond in bicyclo(2.1.0)— pentane 28, which has the highest reported28 strain energy (47.4 kcal./mole), readily underwent a cycloaddition across carbon-carbon double bonds when the catalyst bis(acryloni— trile)nickel(0), Ni(AN)2, was present (eqn. 20). 2=CH— —CN + (20) N1 (AN) CN 75% 16% It should be noted that the thermal, uncatalyzed reaction between 28 and olefins is felt to be a multi-step process involving a diradical, where bicycloheptanes are formed only as the minor components of complicated reaction mixtures.29 The catalyzed reaction, in contrast to the purely thermal cycloaddition, led to stereospecific cycloaddition products shown in Figure 12.27 In 1974 Noyori30 demonstrated another difference in the product stereochemistry found in the thermal un- catalyzed and the nickel(O) catalyzed cycloadditions of a? with olefins. Gassman29 had reported that in the thermal uncatalyzed and the nickel(O) catalyzed cycloadditions of 20 ’b z z 32% Figure 12. Products from the nickel(O) catalyzed addition of olefins to bicyclo(2.1.0)pentane. 47 with olefins. Gassman29 had reported that in the thermal uncatalyzed reaction olefins approach from the endo side of the bicyclo envelope, and consequently, the hydrocarbon undergoes the reaction with double inversion of stereochem— istry at the C-1 and the C-4 positions (eqn. 21). D O D / + O (21) H H O As shown in Figure 13, when a mixture of 20-2,3—d , dimethyl fumarate or dimethyl maleate, and a catalytic amount of Ni(AN)2 were heated in benzene at 60° for 48 hrs., products containing stereochemistry opposite to that encountered in the purely thermal reaction were produced.30 When 20—5,5-d was combined with a nickel(O) complex in the presence of an olefin, it was noted that the forma— tion of the monocyclic adduct must involve a specific deuterium shift (eqn. 22). D D D D c = — H2 GTE + Ni(AN)2 E (22) 88% E H exo _5 D E=COOCH 68%(endo _ 1) 3 46% l N'— Z=COOCH3 61% 5% + A(4%) + B(30%) m m Figure 13. Products from the nickel(O) catalyzed addition of olefins to bicyclo(2.1.0)pentane-2,3-d . 49 Before giving the mechanism proposed by Noyori based on the above data for the bicyclopentane system, it is necessary to define certain terms found in transition metal chemistry. The term "oxidative addition" has come to be used to designate a rather widespread class of reactions in which oxidation, i.e., an increase in the oxidation number of the metal, is accompanied by an increase in coordination number31 (eqn. 23). 35 M + —————————————————a- A + oxidative addition I n Mn+2 (23) B Reductive elimination simply refers to the reverse of oxidative addition (eqn. 24). A In+2 reductive elimination + M ————————————- Mn + (24) B It should be emphasized that these two terms do not imply the processes described above are concerted. B — Hydride elimination is the abstraction by the metal of a hydride from a B-carbon atom, and in so doing generating a metal hydride olefin complex (eqn. 25). 50 Ta 8 — hydride elimination M _ICI (25) c | c H The last term, migratory insertion, refers to a process in which a group coordinated to the metal inserts itself into a metal—atom sigma bond. The reverse process is simply a reverse migratory insertion (eqn. 26). 35 migratory insertion ——-—-———.~ m—A—B (26) 3 w The mechanism presented by Noyori,3o depicted in Figure 14, involves (1) the initial oxidative addition of the strained C—l - C—4 bond of 20 onto the nickel(O) atom ’b (formally d10 + d8 conversion) forming the organonickel intermediates 21 (L AN or olefin), (2) the insertion of ’b the coordinated olefin into the nickel—carbon sigma bond to produce the new organonickels 22, and (3) either the reductive elimination to give the bicyclic products 23, and the regeneration of the nickel(O) catalyst (d8 + le), or a B-hydride elimination followed by reductive elimination to give the monocyclic products 24 with the regeneration of ’\J the nickel(O) catalyst. 51 D' D' D. D. . . H H OXidative H addition H —--———-———————————- \\\N L _ | l n l Nan D D 21 D D m olefin insertion B—elimination ~1———————————_______ D Ln 22 D m reductive reductive elimination elimination Figure 14. Mechanism of the nickel(O) catalyzed addition of olefins to bicyclo(2.1.0)pentane. 52 The first evidence that transition metal catalyzed isomerizations of strained carbocyclic rings involved the possible intermediacy of metallocycles was presented by Katz13 and was soon followed by Halpern and Eaton.14 2 4 When exo-tricyclo(3.2.l.0 ' )octene 2? was heated in the presence of tris(triphenylphosphine)rhodium(I) 13 chloride three products were obtained (eqn. 27). Product 26 was not related by an electrocyclic process to 2? and experiments using 25b demonstrated that a specific deuterium ’\1 shift was required in converting 25b to 26b. 'b ’b X [(C6 H 5) 3P] 3Rhc1 x (27) ’\1 X N6 ’\.- ’\.r8 endo The authors advanced a mechanism shown in Figure 15 which involved the intermediacy of the metallocycles 29 and ’1; 30. m Shortly thereafter it was reported that cubane derivatives underwent valence isomerization with a catalytic amount of (Rh(CO)2Cl)2 in chloroform to give syp—tricyclo- octadienesl4 (eqn. 28). This is formally a (2-12) retro— cycloaddition reaction. When a stoichiometric amount of the catalyst was used, complex 31 was isolated in better ’b Figure 15. The mechanism of the catalyzed valence isomer- ization of exo-tricyclo(3.2.1.02r4)—octene. than 90% yield (eqn. 29). The trapping of this rhodacycle led the authors to propose that for substituted cubanes and possibly for other valence isomerizations which use rhodium, the mechanism should include organorhodium sigma-bonded intermediates, i.e., metallocycles. R R R s ——» D Q 71% R=COOCH 3 2 9 96 54 [Rh(CO)2C1]2 -———————————————.» (29) o l/Rh 0c I C1 31 ’b A more recent example of metal catalyzed rearrangements of strained ring systems was the zinc catalyzed rearrangement of 1—phenylbicyclo(2.1.0)pentane 32.32 The sole product was 3-phenylcyclopentene (eqn. 30). m X X X Ph ZnI2 —————————————.> PhH,60° x (30) a,x=H b,x=D Ph 32 ’\1 When 32-5,5-d2 was combined with zinc iodide the products isolated were consistent with the intermediacy of a zinc metallocycle (Figure 16). Metallocycles were also implicated as reactive intermediates in the mechanism of olefin metathesis. Grubbs15 found that the ip situ generation of a tungsten I metallocycle gave ethylene quantitatively (eqn. 31). For a more detailed discussion of this work see the first chapter of this thesis. 55 D Ph Ph M D Ph D -M ———D- ———I> D D M Ph D D Ph Ph D _> M Figure 16. The mechanism of the ZnI2 catalyzed valence isomerization of l-phenylbicyclo(2.1.0)pentane. Li W Li 1L __, /\ \\ (31) C. Q ( C14 C14 Because of the importance of the role which metallo- cycles as reactive intermediates appeared to be playing in organometallic chemistry, Grubbs33 and Whitesides34 inde- pendently sought to synthesize a stable metallocycle. This is the first instance of a stable metallocycle being the target molecule of an organic preparation. 56 The same metallocycle, bis(triphenyphosphine)tetra— methyleneplatinum(II) 33, was prepared by both groups using the procedure Whitesides35 had developed earlier for the preparation of a set of stable dialkyl platinum(II) compounds (eqn. 32). CHZLi CH2 L PtCl + (CH ) ———.-Et20 Pt/ >CH ) 2 2 2 n L2 / 2 n (32) \GiL' \\ 2 1 CH2 L = (C6H5)3P a? The crystal structure of 33 was done by Grubbs.33 ’D The tetramethyleneplatimum ring was found to be significantly puckered. Grubbs felt this geometry provided an indication of the mechanism of how the rearrangement of metallocyclic intermediates in the olefin metathesis reaction occurred. The mechanisms in eqn. 34 and 35 are reasonable and fit the requirements of the observed geometry. R R R 2 I; ——-—> I 7 CH —-> (i... (34) <———- 4 2 +—-—— M M M —-> ——> (35) 2 g <——- :---7 <— M Whitesides34 examined the thermal chemistry of 33 and other platinum metallocycles. Two features of the thermal decomposition of the platinocycles when n = 2,3 were significant. First, they were remarkably more stable thermally than were the acyclic platinum(II) dialkyls. Second, the products of their decomposition suggested the mechanism by which they were produced resembled that estab- lished for their acyclic analogs, iyef, initial hydride elimination followed by reductive elimination of alkene from an intermediate hydrido-platinum(II) alkyl (eqn. 36). —Pt° —> /\:> 36 <¢—-———— ( ) Pt- H 33 'b H The platinocycles did not give any carbon—carbon bond cleavage products on thermal decomposition, 211°! ethylene or cyclopropane. Thus, no conclusions could be drawn on the validity of metallocycles as reactive inter— mediates in the earlier discussed reactions. However, their 58 thermal stability did shed light on the importance of the elimination of a metal hydride in the thermal decomposition of many transition metal alkyls. Since these eliminations probably occur most readily from conformations of the organometallic compounds in which M-C-C-H dihedral angles are 0°, incorporation of B-CH moieties into a tetramethylene metallocycle, in which the M-C—C-H dihedral angles would be greater than 90°, should result in a decrease in the rate of thermal decomposition relative to an acyclic analog by inhibiting metal hydride elimination. This was what was observed. Whitesides36 more recently synthesized a titanium(IV) metallocycle 3f (eqn. 37). Several reactions of this complex have no close analogy to the chemistry of similar acyclic titanium(IV) dialkyls. This presumably resulted from the inability of metallocyclic species to achieve the 0° M-C-C-H dihedral angle most favorable for metal hydride elimination. He felt this slow rate of metal hydride elimination allowed metallocycles to display unusual types of reactions normally masked by the more facile hydride eliminations. Li EtZO Cp\ H—szTiCIZ + Li —— /Ti (37) (‘300 Cp 59 The metallocycle 36 is stable only below -30°, and its structural assignment rests on the reactions shown in Figure 17. Li Et20 U A Cp TiCl —-> me? 2 2 . -78° = Li Ti 2 2 + CH2 CH | C92 H 34 ’\1 O Brz CO —78 Et 0 HCl pentane 2 Br E CH3 @4— 0 Br Ti CH3 Figure 17. Reactions on which the structural assignment of the titanium metallocycle 34 rests. "U Of the reactions in Figure 17 note should be made of the thermal decomposition of 34 to give ethylene, which ’1: probably arises from a carbon—carbon sigma bond fragmenta- tion route encouraged by the near 0° Ti—C—C-C dihedral angle (eqn. 38). O ——?———> //\Ti/\\ (38) <—-——-——— 60 One other result is significant. When ethylene was introduced into a system containing Cp2TiN2TiCp2, which was below -30°, reaction mixtures, 35, were produced whose ’b properties strongly suggested the presence of titanium metallocycles (Figure 18). CH2=CH2 Cp2TiN=NTiCp2 ————> 3&5 Br C1 Et 0 C Br C CH3 Br CH3 0 Figure 18. Reactions supporting a titanium metallocycles formation from szTiNZTiCp2 and ethylene. The goal of this thesis research project was to synthesize a metallocycle which exhibited the most desirable properties of the platinum and titanium metallocycles, 1;§" isolatability and chemistry other than that derived from hydride elimination schemes. For the following reasons nickel was selected as the transition metal to be used. 61 Nickel metallocycles have been proposed as reactive intermediates in a number of catalyzed reactions. As shown earlier in this introduction, nickel(O) complexes catalyzed the addition of strained carbon-carbon sigma bonds to olefins in a (2-+2) fashion27’30 (eqn. 22). A nickel(O) catalyst was also found to cyclodimerize methylenecyclopropane37 (eqn. 39). Ni(COD)2 D:———+ O: pentane -15° The scheme shown in Figure 19 was proposed by the authors to explain how the dimers 36 and 37 could be formed. No actual evidence other than the structure of products was given to support the metallocycles 38 and 39. A similar reaction finds nickel cyclodimerizing and cyclotrimerizing allene.38’39 Again metallocycles were cited as likely intermediates. Until 1973, nickel catalyzed cycloaddition of two olefins had been limited to strained olefins such as norbornadiene and methylenecyclopropane. Hall40 found that if a nickel complex such as nickelocene or cyclo- pentadienylnickel carbonyl were used at 250° and 1300 psi 62 [>: + Ni (COD) 2 -COD (COD) Ni k x‘ N (COD) Ni 38 ’\J (COD)Ni Figure 19. Mechanism for the nickel(O) catalyzed cyclodimerization of methylenecyclopropane. 63 of ethylene in the presence of acrylonitrile, that cyclobutanecarbonitrile could be produced. The reaction was found to be more stoichiometric than catalytic in nickel (eqn. 40). 0 ll -IC\ l @Nl Nl"© \C/ CN CH2=CHCN g + (40) CH2=CH2 As to the feasibility of preparing a nickel metallo- cycle, dialkyl nickel complexes have been prepared, and when there is a stabilizing ligand present, they are thermally quite stable. Yamamoto41 prepared diethyldipyridylnickel 40 from acetylacetonate, a,d'-dipyridy1 and diethylaluminum ethoxide (eqn. 41). N dipyridyl O \ /Et Ni(acac) +Et AlOEl ————-> Hi (41) (DH Et 40 ’b He found that temperatures above 100° were necessary to thermally decompose this complex (eqn. 42). 64 A 4 __—_-. + = + _. A? CH3(CH2)2CH3 CH2 CH2 CH3 CH3 (42) It is interesting to note that the thermal decomposition went through ethyl radicals, and not by a B-hydride elimination scheme. This seems to be the rule rather than the exception for nickel alkyl and aryl chemistry. Taking these facts together, nickel seems to be an ideal choice for a metallocycle which will be thermally stable and should give chemistry other than that of B—hydride elimination. 65 RESULTS AND DISCUSSION This chapter deals with the preparation and characterization of two nickel metallocycles. Most organonickel compounds are prepared by com— bining nickel halide complexes with very electropositive organometallic species. In general, monosubstitution of L2NiX2, reagents at temperatures not greater than 25°, whereas where L = monodentate ligand, occurs with Grignard disubstitution occurs when lithium and sodium reagents are used. Thus, the easily accessible l,4-dilithiobutane was chosen as the organometallic reagent to be used in the preparation of the tetramethylenenickel complex. Prior to the work of Chatt and Shaw42 which described the synthesis and characterization of stable aryl and alkynyl nickel complexes, transient dialkyl and diaryl nickel species had been prepared, but the necessary conditions were not achieved for the isolation of these 43’44 Chatt's and Shaw's success was attributed compounds. to their having additional ligands simultaneously bound to the nickel which imparted stability to the product. There is no general agreement as to why certain ligands stabilize transition metal—carbon sigma bonds. One theory suggests the ligands possess a proper balance between G-donor and H—acceptor properties which favorably influences the level of the 0 and 0* molecular orbitals of the complex.45’46 66 Another theory has the ligand only tightly occupying a coordination site on the metal.47 Thus, any decomposition route which requires the metal to make a coordination site available through loss of a ligand is effectively blocked. When nickel complexes of the type L NiR(X) or 2 LZNiRR' were prepared from organometallic reagents and a nickel halide complex, the ligand which proved most satisfactory was a tertiary phosphine. Taking the above into consideration, the first nickel substrate chosen for study by us was the easily prepared bis(tripheny1- phosphine)nickel dichloride a}. In a 500 ml. side-arm round bottom flask flushed with argon were combined 16.0 g. (24 mmoles) of 43 and 50 ml. of ether. The reaction mixture was brought to dry ice—ethanol bath temperature, and 200 ml of 0.25 M l,4-dilithiobutane in ether was slowly added by syringe. After stirring at the bath temperature for 5 min., the bath was removed. As the reaction mixture warmed, it turned dark and homogeneous, and on warming further a precipitate formed. When precipitation appeared to be complete, the mixture was filtered under a flow of argon and the solid was washed with ether until the wash was only faintly yellow. The bright yellow solid was dried under a flow of argon to give 2.91 g (20%) of bis(triphenyl— phosphine)tetramethylenenickel(II) 42 (eqn. 43). Because 43 67 is extremely sensitive to dioxygen, it was transferred and stored for later use into a dry box. It also follows that it was important to rigorously exclude dioxygen from reaction vessels in the reactions described below. CH / 2\CH2 ((C6H5)3P)2N1C12 + L1-(CH2)4-L1 ——.- ((C6H5)3P)2N1 | (43) \ /CH2 CH2 41 42 (\J ’b It should be noted that the ratio of l,4—dilithio— butane to 41 should not exceed 1.8 : 1.0. When larger ratios were used either a brown solid precipitated from solution or else no solid was formed. The properties of the brown solid were not investigated, but when it was washed with ether a dark red solution resulted, as opposed to the light yellow coloration 42 gave the wash ether. Also the brown solid appeared to be more soluble in ether than 42. The chemical characterization of 42 is discussed below. Transition metal-carbon sigma bonds generally can be cleaved by the addition of a strong acid to the metal complex. Thus, if 42 were decomposed by strong acid, then ’b it should generate butane and/or other C -hydrocarbons 4 (eqn. 44). 68 H+ 42 ——-————> - - '0 CH3 (CH2)2 CH 3 + other C4-hydrocarbons (44) In a 250 ml. Erlenmeyer flask fitted with a rubber septum was placed an accurately weighed amount of 43, approximately 0.4 g., and to the complex was added 5 ml. of concentrated hydrochloric acid. After gas evolution had ceased and a homogeneous solution formed, 1.0 ml. of propane, to be used as a g.l.p.c. standard, was added by syringe to the flask. The yield of the C —hydrocarbons 4 produced was determined by removing the gas above the solution by syringe and analyzing it by g.l.p.c. Near quantitative yields of C4—hydrocarbons were found along with trace amounts of ethylene and ethane. The homogeneous solution from the addition of concentrated hydrochloric acid to 43 was next analyzed for its nickel content. By Chemical means nickel is most often determined by precipitation with dimethylglyoxime from a neutral or ammoniacal solution.48 The sparingly soluble precipitate is scarlet in color and has the formula Ni(C4H7N202)2. After being dried, it is weighed, and from its weight the percent nickel of the original complex can be determined. This procedure is virtually specific for the determination of nickel. 69 Before the nickel analysis was carried out on the solution, it was first diluted with 150 ml. of water. At this point triphenylphosphine and/or triphenylphosphine oxide precipitated out of solution, and was subsequently removed by vacuum filtration. The filtrate was ready for nickel analysis. The percent nickel was consistently found to be 9.11:0.2% (theo. 9.1%). If a preparation of 42 gave a nickel analysis or gas analysis which was not satisfactory, it was not used in further reactions. These two reactions, which make use of HCl and dimethylglyoxime, provide an efficient and routine manner for evaluating the purity of 43. In the reaction above, butane was generated from 42 and HC1. If the butane was arising from the acid decomposition of a tetramethylene metallocycle and if DCl was used in place of HCl, then butane-l,4-d should be produced from 42 (eqn. 45). If the butane was instead coming from the acid decomposition of a C4-acyclic nickel complex, then DCl would produce butane-l-d (eqn. 46). CH / 2\ CH DC]. M l 2 --——--—--O' D—(CH2)4-D (45) \ //CH2 CH 2 70 DCl M-(CH2)3—CH —————- D-(CH ) —CH (46) 3 23 3 Deuterochloric acid DCl was generated by slowly combining under nitrogen 10 ml. of deuterium oxide and 5 ml. of acetyl chloride (eqn. 47). II o CH3-C-Cl + D20 —————-———- CH3-g-OD + DCl + D20 (47) The DCl solution was then added to 42. The gas above the resultant solution was removed from the flask by syringe and the butane in the gas was isolated by g.l.p.c. The butane was then analyzed by mass spectroscopy (Table 9). The mass spectrum exhibited a parent peak of m/e 60 which corresponds to butane—d . This spectrum matches well with the spectrum reported in the literature for butane-l,- 4 4—9 . 9 Both have a base peak m/e 44 which corresponds to the loss of CHZD. Loss of methyl is characteristic for butanes and generally this fragmentation gives the base peak for the spectra. The relative intensities of base to parent peaks are the same for both the literature and product spectra. The small discrepancies in relative peak intensities between certain peaks in our spectra versus those reported 71 Table 9. Mass Spectral Analysis of the Butane—dz's Produced from DCl and the Tetramethylene Nickel(II) Metallocycles Relative Intensities Butane—l—d Butane—1,4—d2 From Buli 72 in the literature could have at least two sources. The spectrometer could be the source, or H20 and HCl as a contaminant in the DCl solution could be the source. To check the latter possibility butyllithium was slowly added to the DCl solution. This should generate only butane—l-d . The butane generated was isolated and analyzed as before by mass spectroscopy (Table 9). The correct parent peak m/e 59 was exhibited. Since loss of methyl is equally likely from each end of the chain, then the base peak can be either m/e 43 or 44. If the relative intensi— ties of m/e 43 and 44 are significantly different then it may be concluded that there was a significant amount of water present in the DCl solution. As Table 9 shows, the relative intensities of m/e 43 and 44 are very close, therefore an insignificant amount of water is present in the DCl. Further confirmation of the presence of a tetra- methylene metallocycle through the specific labeling of the d,w—positions of the tetramethylene unit was sought. Bromine is known to undergo oxidative additive reactions with transition metal complexes. If bromine was added to 42 or any other tetramethylene metallocycle, l,4—dibromo- butane should be generated. A possible route for its formation is shown in Figure 20. 73 /CH2\CH oxidative Br\ /CH2\CH reductive BrCH2\CH Br2 + M I 2 addition M I 2 elimination M I 2 \ l/CHZ //’\ //CH2 //\ //CH2 CH Br CH . r) Br CH 2 2 . atlo 2 .mln ere” H H CH— — B B O / \ CH2 CH2 r r2,Et2 M -—————— M— Br Br CH2 + Br I CHz—CHZBr CH =CH—CH -CH Br—M | 2 2 3 \ CH2 Br CH2 /Br M + Br-(CH ) -Br \ 2 4 Br Figure 20. Formation of l,4—dibromobutane from a tetramethylene metal complex and bromine. 74 When bromine was added to 42, l,4—dibromobutane was generated. The temperature at which the reaction was carried out was found to be important. When done at ~78° in ether, near quantitative yields of l,4—dibromobutane (85-100%) were obtained. When the reaction was carried out at 0° or room temperature the yields were substantially reduced. This loss in yield could be due to an intermediate such as 43 at higher temperatures having available to it a hydride elimination pathway (see Figure 20). This is in line with the observation that substantial amounts of l— and 2—butenes were formed when acid was added to 42. As noted in the introduction, migratory insertion is a common reaction of transition metal complexes. Inser— tion of carbon monoxide in many cases is a very facile process (eqn. 48). co To lo M—R ————§ M—R ——> M—C—R (48) Halpern and Eaton14 and more recently Blum59 used carbonyl insertion to trap the unstable rhodacycles formed from cubane and 1,3—bishomocubane as their acylderivatives (eqns. 49 and 50). o ,L co r' Rh ———" 2C1 (49) \cl “1 CO 75 0 C1 C1 Rh/ CO / \ —> Rh CO \\ CO (50) Whitesides36 as part of his structure proof of bis(H-cyclopentadienyl)—tetramethylenetitanium(IV) 34 added ’\.a carbon monoxide (10 psi) to 34 and cyclopentanone was formed. ’b The scheme in Figure 21 shows a route by which cyclopentanone can be generated from a metallocyclic species. 0 CH CH 0\\C CH IC! 2 2 migratory _ / \CH2 CO / \CHZ insertion / \2 'M CH2/ \CH2 M\ I ———> oc-M\ I ——’ M\ /CHz —"I I CH CH CH —— CH / 2 / 2 2 2 CH2 CH2 CHZCHZ Figure 21. Formation of cyclopentanone from a tetramethylene metal complex and carbon monoxide. We sought to determine if cyclopentanone could be prepared from 42. In a pressure bottle were placed 1.12 g. of 42 and 20 ml. of heptane. While stirring the mixture at 0°, carbon monoxide (10—20 psi) was introduced into the bottle. Based on color change of the solid present (yellow + off—white), the reaction was complete within 2 hr. Analysis of the solution by g.1.p.c. gave a 75% yield of cyclopentanone. 76 The off-white solid formed during the reaction is bis(triphenylphosphine)nickel dicarbonyl m.p. 205-206 dec. (lit.51 m.p. 210—215 dec.). The ir spectrum (mull) exhib— ited strong metal carbonyl bands at 2005 and 1954 cm.—1 (lit.51 2007 and 1952 cm.-l). The dicarbonyl complex was recovered from the reaction medium by filtration in 94% yield. When Yamamoto41 added ethanol to the dialkyl nickel complex diethyldipyridylnickel, the complex decomposed giving ethane. Water was also found to decompose the dipyridyl complex, but not as rapidly as ethanol. Again only ethane was found. When ethanol or water was added to 42 no gaseous products were observed, and the complex appeared to be stable under these conditions. When hydrochloric acid was subsequently added to these mixtures of water (ethanol) and 43, the complex 42 decomposed to give C ~hydrocarbons m 4 as shown by g.1.p.c. Thus, the tetramethylene bridge is inert to both alcoholysis and hydrolysis. The nmr spectrum of 42 was determined in benzene. The complex 42 appeared to be stable at room temperature for days in the solid state or when added to a solvent in which it was insoluble. The normal nmr solvents CCl4 and CDCl3 could not be used since the complex was unstable in their presence; it probably did an oxidative addition across ‘Mn 77 the carbon-chloride bond of the solvent. Freon 113 was also found to be unsatisfactory. The complex 42 was partially soluble in acetone—d but a sufficient concen- tration for an nmr spectrum could not be achieved. The complex is soluble in aromatic solvents to the extent that an nmr could be taken. The problem is that when dissolved in an aromatic solvent, the half life of 42 decreases from 'b days to approximately 1 hr. The nmr of 42 (benzene) showed two singlets at 6 1.63 and 1.70. The positions of the peaks were based on ether as an internal standard. As stated above, the half life of 42 in benzene is approximately 1 hr. This was based on the loss of peak intensity in the nmr with respect to time as the complex decomposed. The products from this decomposition will be discussed later. The second metallocycle prepared was ethylene(bis- (diphenylphosphine))tetramethylenenickel(II) 43. Using the same procedure which gave 42, 50 m1. of 0.20 M.1,4- dilithiobutane in ether was added to 3 g. of dichloro— ethylene(bis(diphenylphosphine)nickel(II) 45 under argon at -78°. Upon allowing the solution to warm a precipitate 44 formed. The solid was isolated by filtration, washed with ether until the wash was only faintly yellow, dried under a flow of argon to give 0.67 g. (23%) of the bright yellow 44. 78 The reactions used to characterize 44 are summarized in Figure 22. Li EtZO HCl diphos NiCl2 + —.' diphos Ni —- C4-hydrocarbons + 45 m 45 Li m DCl PhCH Br2 ,2Et 0 35° = + CH2 CH2 [:3 7;] Br Figure 22. Reactions on which the structural assignment of 44 rests. ’b When hydrochloric acid was added to 44, quantitative yields of C4-hycrocarbons were obtained as determined by g.l.p.c. with propane serving as the internal standard. Unlike 42, when HCl was added to 44, a homogeneous solution was not obtained but rather 45 was recovered in greater than 80% yield as shown by its ir spectrum and mixed melting point. Using the same DCl solution which was used to decompose 42, 44 was decomposed and the butane so obtained ’b ’1; was found to be primarily butane-l,4-d . The mass spectra 79 of the butanes produced from the decomposition of 42 and 44 were nearly identical so the same arguments which showed the butane— —d2 produced by 42 was butane— l, 4— —d2 can now be made for the butane- d produced by 44 (Table 9). —2 As with 42 the formation of l,4-dibromobutane from the addition of bromine to 44 was very temperature dependent. ’\.v The maximum yield 67% was achieved when the reaction was carried out in ether at -78°. When the reaction was con- ducted at 0° or room temperature yields of 20% were common. The thermal decomposition of 42 and 44 in toluene 'b ’b was examined. Several gaseous products were obtained (Table 10). Table 10. Thermal Gaseous Decomposition Products of the Tetramethylene Nickel(II) Metallocycles Gaseous _roducts Solvent Temp . Complex Ethylene Cyclobutane PhMe 30 42 34 61 30 44 28 —— 80 It is important to note that carbon—carbon sigma bond cleavage products were observed in the decomposition products. This is the first case of metallocycle which is stable at room temperature giving products other than those derived from a hydride elimination scheme. It will be recalled that carbon—carbon sigma bond cleavage products, e;g., ethylene, have been previously reported for the tungsten and the titanium metallocycles, which are stable only at relatively low temperatures. This is the first reported case of a prepared metallocycle giving cyclobutane as one of its thermal decomposition products. The cyclobutane was isolated by g.l.p.c. and analyzed by mass spectroscopy and it gave the correct parent peak, i/e 56. Mass Spectrums were run on the spectrometer of l—butene and 2—butenes and their mass spec- trums were significantly different from the cyclobutane mass spectrum. The mechanism of the formation of cyclobutane is currently under study in our laboratory. 81 EXPERIMENTAL Preparation of Bis(tri hen 1 hos hine)- nickelZII) Dichloride The procedure of Venanzi52 was modified to prepare bis(triphenylphosphine)nickel(II) dichloride. To a hot solution of 52.6 g. (0.2 mmole) triphenylphosphine in 500 ml. of glacial acetic acid was added with stirring a solution of 23.8 g. (0.1 mole) nickel(II) chloride hexahydrate, 20 ml. of water, and 250 ml. of glacial acetic acid. The olive green microcrystalline precipitate, when kept in contact with its mother liquor overnight, gave dark blue crystals of bis(triphenylphosphine)nickel(II) dichloride (80%) which were vacuum filtered, washed with glacial acetic acid and dried in vacuo. Preparation and Standardization of l,4—Dilithiobutane In a 500 m1. three—necked round bottom flask fitted with an argon line, 250 ml. dropping funnel, oil bubbler, and stirring magnet were placed 100 cm. of shaved lithium ribbon and 250 m1. of dry oxygen-free diethyl ether. To the reaction was added dropwise over 2 hr. a solution of 25 ml. l,4—dibromobutane and 150 ml. of ether. The reaction was initially allowed to proceed at room temperature for a few minutes, then it was cooled to 0°. Following the addi- tion of the dibromobutane, the mixture was stirred for 30 82 min., and then the mixture was filtered and stored under nitrogen at 0°. The l,4—dilithiobutane was stable for approximately one week at this temperature. The concentration of the l,4-dilithiobutane solution was determined by removing a 10 ml. aliquot of the filtered solution and adding it to 2 ml. of chloro— trimethylsilane under nitrogen. This generated quanti- tatively 1,4-bis(trimethylsilyl)butane. Durene (1,2,4,5— tetramethylbenzene) was added as a standard and the solution was analyzed by g.1.p.c. (5% QF—l at 70°). Pre aration of Bis(triphen l hos hine)— tetramethylenenickel(II) The following reaction and the isolation of the product were done under argon. In a 500 ml. side-arm round bottom flask which was fitted with a stirring magnet, oil bubbler, and rubber septum were placed 16.0 g. (24 mmoles) of bis(triphenylphosphine)nickel(II) dichloride and 50 m1. of dry oxygen—free ether. After cooling the reaction mixture to dry ice—ethanol bath temperature, 200 ml. of 0.25 M. l,4-dilithiobutane in ether was slowly added by syringe. Following the addition of the l,4-dilithiobutane, the mixture was stirred for 5 min., then the bath was removed and the stirring maintained. On warming, the solution became very dark. On warming still further, 83 the bis(triphenylphosphine)tetramethylenenickel(II) 42 precipitated out as a bright yellow solid. When precipita— tion appeared complete, the nickel complex was isolated by filtration, and it was washed with ether (21:30 ml.) under a flow of argon. At this point the complex 42 was contaminated with the starting nickel dichloride. It was observed while washing 42 with ether that 42 remained suspended in the ether longer than the starting nickel complex. Using this information the product mixture was purified by introducing ether into/the funnel, stirring, and letting the starting nickel complex settle out of solution. The product, which was still suspended in the ether, was transferred by syringe to another filter funnel, dried under a flow of argon to give 2.91 g. (20%) of the bright yellow 42: nmr (benzene) 6 1.63 (s), 1.70 (5). Analysis: Ni 9.1:t0.2% (theo. 9.1%); gaseous products from the complex with HCl = >95%. Gaseous Products from HCl and Bis(tri hen 1 hos hine)tetra- methylenenicEeIZII) and Ethyl— ene(bis(diphenylphosphine)- tetramethylenenickel(II) The purity of the tetramethylene nickel complexes was determined in part by measuring the amount of gas evolved when 5 ml. of concentrated hydrochloric acid and 1.0 m1. of propane (g.1.p.c. standard) were added to an accurately weighed amount of the complex in a 250 ml. 84 Erlenmeyer flask which was fitted with a rubber septum. The yield of the gaseous products was determined by g.1.p.c. (7% paraffin wax/alumina at 80°). Nickel Analysis of Bis(triphenylphosphine)- tetramethylenenickel(II) A standard procedure for determining percent nickel by using dimethylglyoxime was employed.48 To the solution which resulted from the determination of the percent yield of gaseous products of bis(triphenylphosphine)tetramethyl— enenickel(II) with hydrochloric acid was added 150 ml. of water. A white precipitate believed to be triphenylphosphine and/or triphenylphosphine oxide formed immediately. To facilitate the removal of this white solid, the mixture was first heated to 70°, then cooled to room temperature, and finally vacuum filtered to give a clear solution. To the clear solution at 60° was added 20 m1. of a freshly prepared 1% solution of dimethylglyoxime in absolute ethanol. Concentrated ammonium hydroxide was added dropwise very slowly until no more of the bright red flocculent precipitate formed. The mixture was digested for one hour, cooled to room temperature, and vacuum filtered onto a tared fritted glass curcible. The solid was dried for several hours at 130°, its weight determined, and the percent nickel calculated using the following equation. wt. of ppt.(g.) X 20.313 %N1 = wt. of sample(g.) 85 Formation of 1,4-Dideuteriobutane from Bis(triphenylphosphine)tetra— methylenenickel(II) and DCl In a 25 mm. x 250 mm. test tube fitted with a rubber septum were carefully combined 5 ml. of acetyl chloride and 10 ml. of deuterium oxide. This generates a solution which contains DCl. The DCl solution was introduced into an Erlenmeyer flask which was fitted with a rubber septum, and which contained bis(triphenylphosphine)tetramethylenenickel(II). When gas evolution had ceased, the gas was removed by syringe, and the butane in the gas was isolated by g.1.p.c. (7% paraffin wax/alumina at 80°). The butane was analyzed by mass spectroscopy and as shown in the table below (Table 11) a significant amount of butane—l,4—d was produced. Table 11. Butane—d2 from DCl and Bis(triphenylphosphine)— tetramethylenenickel(II) Relative Intensities Relative Intensities Butane— —l, 4 -d2 42 with Lit .49 100 ll 22 7 86 Preparation of Butane-l-dl Into a 250 ml. Erlenmeyer flask fitted with a rubber septum was placed 5 m1. of a DCl solution prepared from 10 ml. of D20 and 5 ml. of acetyl chloride. To this solution was carefully added 0.5 ml of 2.4 M. butyllithium in hexane. The gas above the solution was removed by syringe and the butane in the gas was isolated by g.1.p.c. (7% paraffin wax/alumina at 80°), and it was analyzed by mass spectroscopy (Table 12). Butane—l—d was the only butane found. The correct parent peak m/e 59 was exhibited and the relative intensities of m/e 43 and 44 were very close. Table 12. Butane—dl from DCl and Butyllithium Relative Intensities Relative Intensities m/e 59 58 45 44 49 From BuLi 49 From BuLi Lit. and DCl m/e Lit. and DCl 93 100 24 21 43 5 7 42 33 34 3 4 41 28 31 100 93 40 9 8 87 Formation of l,4—Dibromobutane from Eis(triphenylphosphine)tetramethylene- nickel and Bromine Into a 125 ml. Erlenmeyer flask fitted with a rubber septum were placed 0.664 g. (1.04 mmole) of bis(triphenyl— phosphine)tetramethylenenickel and 10 m1. of ether. The reaction mixture was cooled to dry ice—ethanol bath tem- perature and 3 m1. of bromine were added by syringe. The bath was removed and the mixture was shaken until it had reached room temperature at which point 0.10 ml. (0.98 mmole) of dodecane was added as a g.1.p.c. standard. An aliquot of the solution was removed and added to an aqueous sodium bisulfite solution which destroyed the excess bromine. The solution was then analyzed by g.1.p.c. (10% carbowax at 100°) and l,4—dibromobutane (100%) was found. Cyclopentanone Formation from Carbon Monox1de and Bis(triphenylphosphine)- tetramethylenenickel(II) Into a pressure bottle fitted with a stirring magnet were placed 1.12 g. (1.75 mmoles) bis(triphenylphosphine)- tetramethylenenickel and 25 m1. of dry oxygen free heptane. The reaction mixture was cooled to 0° and with vigorous stirring carbon monoxide (20 psi) was added. As carbon monoxide was taken up by the system, more was added so as to maintain the pressure above the solution at 20 psi. After stirring for 2 hr., the reaction was judged complete 88 since none of the bright yellow starting complex was present. Instead, a grayish white solid was present at the end of the reaction. Analysis of the resultant solution by g.1.p.c. (10% FFAP at 80°) showed using undecane as a standard, that cyclopentanone had been formed in 75% yield. The light gray solid was isolated by vacuum filtration and has been characterized as bis(triphenyl— phosphine)nickel(O) dicarbonyl in 94% yield: m.p. 202—203 dec. (lit. 210—215 dec.); ir (mull) 2005, 1954 cm‘l. Thermal Decomposition of Bis(tri henyl- phosphine)tetramethyelenenickel(II) in Toluene Into a 250 ml. Erlenmeyer flask fitted with a rubber septum was placed 0.34 g. of the title nickel complex. The flask was placed in a constant temperature oil bath at 30°, and after several minutes 20 m1. of toluene was added. After shaking for a few minutes the solution was allowed to stand in the bath for several hours. The gas above the solution was analyzed by g.1.p.c. (7% paraffin wax/alumina at 80°) and the results are presented in Table 13. The cyclobutane was isolated by g.1.p.c. and analyzed by mass spectroscopy. The correct parent peak m/e 56 was exhibited by the compound. Mass Spectrums of 1—butene and the 2-butenes were determined and they were significantly different from that of the cyclobutane (Table 14). 89 Table 13. Thermal Gaseous Decomposition Products from Bis(triphenylphosphine) tetramethylenenickel(II) Gaseous Products Ethylene Cyclobutane Table 14. Mass Spectrums of l—Butene, trans-Z—Butene, and i Cyclobutane Relative Intensities Trans—2—butene Cyclobutane 42 53 22 25 3 2 100 100 7 8 32 34 90 Preparation of Ethylene(bis(diphenyl- phosphine))nicke1(II) Dichloride A solution of 80 ml. 2-propanol, 20 m1. methanol, and 1.29 g. of nickel(II) chloride hexahydrate was added to 2.00 g. of 1,2—bis(diphenylphosphine)ethane dissolved in 200 ml. of hot 2—propanol. The ethylenebis(diphenyl— phosphine)nicke1(II) dichloride crystallized as dull orange feathery needles. The product was isolated by vacuum filtration, washed with ether, and dried in vacuo to give 2.35 g. (87%). Preparation of Ethylene(bis(diphenyl— phosphine))tetramethylenenickel(II) The following reaction and isolation of the product were carried out under argon. In a 100 m1. side-arm round bottom flask fitted with an oil bubbler, rubber septum, and stirring magnet were placed 2.99 g. (5.68 mmoles) of ethylene(bis(diphenyl- phosphine)nickel dichloride and 10 ml. of dry oxygen-free ether. The mixture was cooled to dry ice—ethanol bath temperature, and 50 ml. of 0.20 M. 1,4—dilithiobutane in ether was slowly added by syringe. After stirring for 5 min., the bath was removed and the stirring resumed. Upon warming a bright yellow solid precipitated from the solution. The solid was isolated by filtration, washed with ether (2 x 15 ml.), and dried under a flow of argon to give 0.67 g. (23%) of 91 the title compound. Analysis: >95% C4—hydrocarbons produced from the complex and HCl; Ni: >80% ethylene(bis— (diphenylphosphine))nickel dichloride was recovered from the reaction of HCl with the complex. Formation of Butane—l,4-d from _TT—Ethylene bis diphenyl—tho-s—ph—Yine - tetramethylenenickel(II) and DC1 In a 250 ml. Erlenmeyer flask fitted with a rubber septum was placed approximately 0.4 g. of ethylene(bis(di- phenylphosphine)tetramethylenenickel(II). To the complex was added 5 ml. of a DC1 solution prepared from 10 m1. of D20 and 5 m1. of acetyl chloride. When gas evolution had ceased, the gas in the flask was removed by syringe and the butane in the gas was isolated by g.1.p.c. (7% paraffin wax/alumina at 80°). The butane was analyzed by mass spectroscopy and the parent and base peak m/e 60 and 44, respectively, agree with those reported for butane-1,4-d (Table 15). Formation of l,4—Dibromobutane from Bromine and Ethylene(bis(di hen 1— phosphine))tetramethylenenickeliII) Into a 125 m1. Erlenmeyer flask fitted with a rubber septum were placed 0.94 g. (1.83 mmoles) of ethylene(bis(di- phenylphosphine))tetramethylenenickel(II) and 10 ml. of ether. The mixture was cooled to ethanol—dry ice bath temperature, and 3 ml. of bromine was added by syringe. 92 Table 15. Butane-d2 from DC1 and Ethylene(bis(diphenyl- phosphine))tetramethylenenickel(II) Relative Intensities Relative Intensities The bath was removed, and the flask was shaken until it reached room temperature. Dodecane, 0.10 ml, was added as a g.1.p.c. standard and an aliquot of the mixture was removed and quenched with aqueous sodium bisulfite. G.l.p.c. analysis (SE—30 at 130°) gave l,4—dibromobutane in 67% yield. Thermal Decomposition of Ethylene(bis(di— phenylphosphine))tetramethylenenickel(II) in Toluene Into a 250 ml. Erlenmeyer flask fitted with a rubber septum was placed 0.40 g. of the title compound. The flask was placed in a constant temperature oil bath at 30°, and to the flask was added 20 m1. of toluene. The solution was allowed to stand in the bath for several 93 hours. The gas above the solution was analyzed by g.1.p.c. (7% paraffin wax/alumina at 80°) and the are presented in Table 16. Table 16. Thermal Gaseous Decomposition Products Ethylene(bis(diphenylphosphine))tetrame nickel(II) Gaseous Produc ’b Cyclobutane 72 CHAPTER III THE OLEFIN CATALYZED DECOMPOSITION OF TETRAMETHYLENE NICKEL METALLOCYCLES INTRODUCTION The coordination of olefins to alkyl transition metal compounds and the resulting activation of the alkyl- metal bonds constitute the crucial steps in various cata— lytic reactions, particularly in olefin polymerization with Ziegler—type catalysts.54’55 The mechanism of Ziegler-Natta polymerization involves the formation of a transition metal-alkyl to which the olefin coordinates. Next, migratory insertion of the olefin into the metal—carbon bond occurs to generate the new metal alkyl, etc. (Figure 23). R— ' = = M CH CH CH2 CH2 M—X —‘—.' M—R —2‘—25- M—R —'—.' M-CHz-CHz-R _1 = H CH2 C 2 etc. Figure 23. Mechanism of olefin polymerization using Ziegler-Natta catalysts. 94 95 It is important to note that these Ziegler-Natta polymerization catalysts, transition metal alkyls, are stable in the absence of olefins, i;g., the coordination of an olefin is necessary to labilize the metal-carbon bond. The most important systematic study of olefin activation of alkyl transition metal bonds was carried out by Yamamoto56 using the dialkyldipyridyl system. Diethyldipyridylnickel is thermally quite stable at room temperature.41 However, the metal—alkyl bonds (R—M) are readily cleaved at room temperature by acryloni— trile with the formation of zero—valent nickel complexes and butane with small amounts of ethane and ethylene being generated (eqn. 51). ow. .. ‘ // 2 5 CH =CHCN . 2 .N1 ——> = ' I x \ C4H10+ (CH2 CHCN)nN1(d1py) (51) It was found that a correlation could be drawn between the electron deficiency of the olefin and the ease with which the metal alkyl bonds were cleaved. As the electron deficiency of the olefin increased, higher rates for the decomposition were found. 1: I155 96 Their explanation for this observation is based on the theory initially proposed by Chatt and Shaw,42 and later advanced by Cossee46 to explain the stability of alkyl metal complexes. They felt the stability of the alkyl metal complexes was based on the energy gap between the bonding oR-M orbital and a vacant d orbital. Electron promotion from the bonding o to a vacant d orbital would result in the splitting of R—M the R-M bond. Thus, if the energy gap were to decrease, promotion of the electron would be easier, and the R—M bond would become more susceptible to cleavage. Also an indirect electron promotion may be another path to cleavage of the R—M bond, but this still leads to the same conclusions as far as the stability of the R—M bond is concerned. An electron may be promoted to the lowest vacant energy level from a highest occupied energy level to which in turn an electron may be promoted. It was found that coordination of olefins, which have electron withdrawing substituents, to the metal caused a marked lowering of the energy levels of the d orbitals, as demonstrated by blue shifts of the charge transfer band for the nickel complex. They assumed that the magnitude of the ligand field splitting is chiefly determined by a strong field ligand, dipyridyl, and therefore the energy gap between the highest occupied energy levels and the lowest 97 vacant energy levels, Al and Aé, in Figure 24, may not be so different. If the energy level of the Ni—C bond is not seriously affected by coordination of the olefin to the empty site of the nickel complex, then the lowering of the vacant orbital would make excitation from the Ni-C orbital much easier (the excitation energy would decrease from a to B in Figure 24), and thus splitting of the Ni-C bonds would result. (a) in hexane ————1 0.5°). It was found with 42 that temperature did influence somewhat the composition of the gaseous products, so the constant temperature bath was employed for all the runs described below for both metallocycles. Table 17 is a listing of the olefins used and the gaseous products which were generated by the olefin and 42. '\J The olefins are arranged in decreasing order of the Alfrey- Price 3 values,57 which are considered to reflect the polarity, ipg., the electron deficiency, of the Vinyl group. The more positive the value for e the more electron deficient is the olefin. The following observations were made on Table 17. The relative ratio of ethylene increased as 3 increased. In general linear alkenes, which could arise from hydride elimination schemes were absent when an electron deficient olefin was present in the solution. Olefins with negative e's influence only slightly the composition of the gaseous products. ngpm H%:H> kusm mcmanoHo>u mcoz ®COpmx H>CH> Hmnumz wHHsDHCOsto< 100 mummama Hwnuwaeo mcocflso openpxccm oewamz mcwawnumocwmomuuwe E mcmudnoaomo mamamnpm demon. meQEoo cammao muospoum msommmo mo oflumm m>HumHmm a me meQEOU Eouw mauspoum cofluemomfioomo pr>HMDm0 aflwwao .PH wanme 101 As with the dialkyldipyridyl complexes, it was noted from color changes in the system that, the decomposi— tion rate of 42 increased as the electron deficiency of the olefin increased. A run was made using the olefin acrylonitrile to determine if the products were stable to the reaction con- ditions with respect to time. The product ratios remained constant with time. Again using acrylonitrile, it was established that when lithium bromide or bis(triphenylphosphine)nickel(II) dichloride, which are possible contaminants arising in the preparation of 42, were added to 42 in toluene, no change in the composition of the gaseous products was observed. As noted in the previous chapter, 42 is very sensitive to dioxygen. Dioxygen could be acting as a radical initiator or as an electron deficient species with a fl-bond. When 10 ml. of dioxygen was added to a flask containing 42 and toluene, cyclobutane and ethylene in a ratio of 10 : 1 were the major gaseous products. From this it appears the n—bond is complexing to the metal, promoting the decomposition of 42. The Lewis acid diethylaluminum ethoxide was used to prepare the diethyldipyridylnickel species.41 When the Lewis acid aluminum chloride was added to a solution of 42 and toluene, the solution immediately turned black, and ’b 102 the only gas found above the solution was ethylene in low yield. From this experiment it might be concluded that it is not feasible to prepare a tetramethylene nickel phosphine complex from an aluminum metallocycle. The presence of the tertiary amine triethylamine had little influence on the gaseous product composition which arose from the decomposition of 42 in toluene. Table 18 gives the gaseous products of the olefin catalyzed decomposition of the nickel complex 44. The complex 44 appears to give more ethylene than 42. One trend is that the lower the e for the olefin, the more ethylene is generated. This is just the opposite trend which was observed in the olefin catalyzed decomposition of 43. The complex 44 also produces more butenes than 42. ’b 103 mamawnum meQEOU muospoum msowwmo mo oflumm m>fipwamm 8 vv xMHmEou mo muOSCOHm cofipflwomfioomo pmumamumu cemwao wHHHuHCOH>MU< openpwrcw UHmHmz mamawnumocmwomuuwe porno H>CH> H>p5w# ®COZ cflwmao .mH magma 104 EXPERIMENTAL A General Procedure Used in the Olefin Catalyzed Decomp051tion of the Tetrameth lene Nickel Metallocycles 42 and 44 Into a 200 m1. Erlenmeyer flask fitted with a rubber septum was placed approximately 0.4 g. of freshly prepared metallocycle 42 or 44. The flask was placed in a constant temperature oil bath at 30°, and 20 m1. of dry oxygen—free toluene was added by syringe. The flask was shaken until the solid appeared to be dissolved, approximately 1 min., and to the solution was added the olefin by syringe. The solution was again shaken, and then allowed to stand in the bath. Large excesses of the olefin were generally employed. In most cases the ratio of olefin to nickel was >20 : 1. If a solid olefin was used, it was either combined with the nickel complex and then toluene added or it was first dissolved in the toluene and this solution was added to the nickel complex. When the reaction was judged complete, generally noted by a color change, the gas above the solution was removed by syringe and analyzed by g.1.p.c. (7% paraffin wax/alumina at 80°) with propane serving as an internal standard. APPENDIX A CALCULATION OF POSSIBLE ETHYLENE-D4 : D2 RATIOS FOR THE "PAIR-WISE" MECHANISM OF OLEFIN METATHESIS APPENDIX A CALCULATION OF POSSIBLE ETHYLENE-D4 : D2 RATIOS FOR THE "PAIR-WISE" MECHANISM OF OLEFIN METATHESIS For the "pair—wise" exchange mechanism there are three possible cases which must be considered. Product Product Consider l: metathesis, no rotation or exchange 2: metathesis, with rotation no exchange 3: metathesis, with rotation and exchange For each case number the carbons and write all permutations allowed for that case. Then write a matrix for the following reaction. 105 106 l,7—octadiene + metal—ethylene + ratio of deuterated D4 D4 D4 D0 D0 D0 ethylenes and metal D4 ethylene complexes produced D2 D0 D4 D2 D0 Next take the values in the matrix and calculate the various ethylenes and metal ethylenes produced based on the following equation: E-DX (or ME-DX) = pl(l,7-D4)(ME-D4) + p2(l,7-D4)(ME-D2) ME—DX l,7—DX = p: + p3(l,7-D4)(ME-D0) + p4(l,7—D0)(ME—D4) + p5(l,7—D0)(ME—D2) + p6(1,7-D0)(ME—D0) product ethylene—d.X metal ethylene—dx complex l,7—octadiene-d__X probability factor molar concentrations CASE 1: Metathesis with no rotation 107 1. Possible ethylenes which could be generated are 1-3 and 2—3. 2. Possible metal ethylenes which could be generated are 2-4 and 1—4. Permutations Case A: 4 = l = 2 = 3 = CD (or CH2) 2 Ethylene produced will be D4 (or D0) Metal ethylene produced will be D4 (or D0) Case B: 4 = 1 2 # 3 then 1 = CD (or CH2) 2 Ethylenes produced are 1 D2 from 1-3 and 1 D2 from 2-3 Metal ethylenes produced are 1 D4 (or D0) from 2-4 and 1 D4 from 1—4; or 4 = 1 = 3 # 2 then 1 CD2 (or CH2) Ethylenes produced are 1 D4 (or D0) from 1—3 and 1 D2 from 2-3 Metal ethylenes produced are 1 D2 from 2-4 and 1 D4 (or D0) from 1—4 Thus, there are four modes of formation of ethylene, three produced D2 and one produced D4 (or D0) and four modes of formation of metal ethylene, three produced D4 (or D0) and one produced D2. Case C: l = 4 # 2 = 3 1 = CD2 (or CH2) Ethylenes produced are 1 D2 from 1—3 and 1 D4 (or D0) from 2—3 Metal ethylenes produced are 1 D2 from 2—4 and 1 D4 (or DO) from l-4 108 Thus, there are two modes to formation of ethylene, one produces D2 and the other D4 (or D0) and there are two modes of formation of metal ethylene complexes and one produces D2 and the other D4 (or D0). This information can be put into the matrices below. METAL ETHYLENE ETHYLENE PRODUCT PRODUCT CASE l,7-DX ME-DX p4 pp pp p4 pp pp A D4 D4 1 o o 1 o o B D4 D2 1/4 3/4 0 3/4 1/4 0 C D4 D0 1/2 1/2 0 1/2 1/2 0 C D0 D4 0 1/2 1/2 0 1/2 1/2 B D0 D2 0 3/4 1/4 0 1/4 3/4 A D0 D0 0 o 1 o o 1 To get the starting ratios of metal ethylene, sum the columns for the metal ethylene matrix: ME—D4 = 9/4; ME—D2 = 6/4; ME-DO = 9/4 or a ratio of ME—D4 : D2 : D0 = 3 : 2 : 3. The probability factors are read from the matrix, e.g., for ethylene—d4 P1 = 1, p2 = 1/4, P3 = 1/2: P4 = PS = p6 = 0. Ethylene-D4 (l)(l)(ME-D4) + (l/4)(l)(ME-D0) + (1/2)(1)(ME-D0) (l)(3) + (1/4)(2) + (1/2)(3) 5 Ethylene-DO Ethylene-D2 Ethylene-D4 II 109 (3/4)(l)(ME-D2) + (1/2)(1)(ME—D0) + (1/2)(l)(ME-D4) + (3/4)(l)(ME-D2) (3/4)(2) + (l/2)(3) 6 1.2 Check for convergence: 3 r? U as II + (1/2)(3) (l)(3) + (3/4)(2) + (1/2)(3) = 6 + (3/4)(2) ME-D2 = (1/4)(2) + (1/2)(3) + (1/2)(3) + (1/4)(2) CASE 2: Metathesis with rotation no exchange 1. Possible ethylenes: 2. Possible metal ethylenes: therefore it has converged Permutations Case A: Case B: 4 All lead to D4 4 1—2 I 1-3, 2—3 3-4, 2—4, 1 = 2 = 3 = CD2 (or CH2) 1 ¢ 2 = 3 = CD 2 (or D0) (or CH2) 1—4. 4 Possible Ethylenes 1-2 1-3 2-3 Ethylenes Formed D2 D2 D4 (or D0) 110 Possible be 3-4 2—4 104 ME Formed D2 D2 (or D0) D4 (or D0) There are 2 D2 and 1 D4 (or D0) paths to ethylene. There are 2 D2 and 1 D4 metal ethylenes. Case C: 4 Possible Ethylenes 1—2 1-3 2-3 =2,e3 Ethylenes Formed D4 D2 D2 There are 2 D2 and ethylene. There are 1 D2 and 2 D4 metal ethylenes. Possible 111 3-4 2-4 1—4 (or D0) paths to ME Formed D2 D4 D4 1 D4 (or D0) paths to (or D0) paths to This information is put into the matrices below for CASE 2. 111 1 METAL ETHYLENE ETHYLENE PRODUCT PRODUCT CASE 1 , 7-DX ME—DX p4 pp p_(_)_ p4 2 Q A D4 D4 1 0 0 1 0 0 C D4 D2 1/3 2/3 0 2/3 1/3 0 B D4 D0 0 2/3 1/3 1/3 2/3 0 B D0 D4 1/3 2/3 0 0 2/3 1/3 C D0 D2 0 2/3 1/3 0 1/3 2/3 A D0 D0 0 0 1 _Q_ _g_ _g_ Total for ME's 1 1 1 therefore, ME—D4 ME—D2 = ME—DO = l H Ethylene—D4 = (l)(l) + (1/3)(1)(1) + (1/3)(l)(l) = 5/3 Ethylene—D2 = 2/3 + 2/3 + 2/3 + 2/3 = 8/3 Ethylene-D2 _ Ethylene-D4 CASE 3: Metathesis with exchange and rotation 1. Possible ethylenes: 1—2, 1-3, 1—4, 2-4, 2-3, 3-4. 2. Possible metal ethylenes: 3-4, 2-4, 2-3, 1-3, 1—4, 1-2. Permutations Case A: 4 = l = 2 = 3 = CD (or CH 2 2) All ethylenes and metal ethylenes produced are D4 (or D0) Case B: 4 = l # 2 = 3 112 Possible Ethylenes Ethylenes Produced 1-2 D2 1—3 D2 1-4 D4 2-3 D0 2-4 D2 3-4 D2 There are 4 D2 paths, to ethylenes. There are 4 D2 paths, to metal ethylenes. Case C: 4 = l = 2 # 3 Possible Ethylenes Ethylenes Produced 1—2 D4 1—3 D2 1—4 D4 2-3 D2 2-4 D4 3-4 D2 Possible 12 3—4 2-4 2-3 Possible be 3-4 2-4 2—3 1-4 1-3 1-2 ME Produced D2 #2 D0 D4 D2 D2 1 D4 path and 1 D0 path 1 D4 path and 1 D0 path ME Produced D2 D4 D2 D4 D2 D4 There are 3 D2 and 3 D4 (or D0) paths to ethylene. There are 3 D2 and metal ethylene. 3 D4 (or D0) paths to This information is put into the matrices below for CASE 3. CASE l,7-DX A D4 C D4 B D4 B D0 C DO A D0 Ethylene—D4 Ethylene—D2 Ethylene—D2 Ethylene—D4 The system converges on 1.33 = ME-DX D4 D2 D0 D4 D2 D0 113 ETHYLENE PRODUCT 242m 1 0 0 1/2 1/2 0 1/6 4/6 1/6 1/6 4/6 1/6 0 1/2 1/2 0 0 1 Total for ME's METAL ETHYLENE PRODUCT 131—22 91 1 0 0 1/2 1/2 0 1/6 4/6 1/6 1/6 4/6 1/6 0 1/2 1/2 1 _0_ 1 11/6 14/6 11/6 1(11/6) + 1/2(l4/6) 1/6(11/6) + l/6(11/6) + 1/6(11/6) = 130/36 1/2(14/6) + 4/6(1l/6) + 4/6(11/6) + (l/2)(14/6) = 1.32 172/36 Ethylene-D4 Ethylene—D2 10. 11. 12. 13. 14. 15. REFERENCES N. Calderon, Accounts Chem. Res., 5, 127(1972). .__________________ m N. Calderon, H. Y. Chen, and K. W. Scott, Tetrahedron Lett., 3327(1967). N. Calderon, E.A. Ofstead, J. Ward, W. Judy, and K. Scott, J. Amer. Chem. Soc., 40, 4133(1968). C. Bradshaw, E. Howman, and L. Turner, J. Catal., 7, 269(1967). “ E. Zeuch, Chem. Comm., 1182(1968). C. Adams and S. Brandenberger, J. Catal., 13, 360(1969). W R. Pettit, H. Sugahara, J. Wristers and W. Merk, Disc. of Far. Soc., 47, 71(1969). For a complete presentation see R. B. Woodward and R. Hoffmann, "The Conservation of Orbital Symmetry," Verlag Chemie, Weinheim, Germany, 1970. F. Mango and J. Schachtschneider, J. Amer. Chem. Soc., 49, 2484(1967). F. Mango, Advan. Catal., 40, 291(1969). F. Mango et al., J. Amer. Chem. Soc., 93, 1123(1971). _____________1_____ m R. Pettit and G. Lewandos, Tetrahedron Lett., 789(1971). T. Katz and S. Cerefice, J. Amer. Chem. Soc., 91, 2405(1969). ” J. Halpern, P. Eaton, and L. Cassar, ibid., 92, 3515(1970). m R. Grubbs and T. Brunck, ibid., 94, 2538(1972). 114 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 115 M. Lappert, D. Cardin, and M. Doyle, J.C.S. Chem. Comm., 927(1 2). C. Casey and T. Burkhardt, J. Amer. Chem. Soc., 96, 7808(1974). N For recent reviews of transition metal carbene complexes see D. Cardin, B. Cetinkaya, and M. Lappert, Chem. Rev., 12, 545(1972); F. Cotton and C. Lukehart, Progr. Inorg. Chem., 46, 243(1972). R. Lespieau, Compt. Rend., 158, 1188(1914); C.A. 3:2678. W The reactions used to convert dimethylsuberate to 1,7-octadiene—l,l,8,8-g4 were first carried out using nondeuterated material so that yields could first be maximized. Thus, nmr spectra for all nondeuterated intermediates were available. For a review on pyrolytic cis eliminations see C. Depuy and R. King, Chem. Rev., 40, 431(1960). E. Mutterties and M. A. Busch, J.C.S. Chem. Comm. 754(1974). A. Yamamoto, L. Pu, J.C.S. Chem. Comm., 9(1974). W. Grahlert, K. Milowski, and U. Langbein, Z. Chem., 44, 287(1974). T. Katz et al., J. Amer. Chem. Soc., 94, 5446(1972). ___________________ m J. Osborn et al., ibid., 9?, 597(1973). R. Noyori, T. Suzuki and H. Takaya, ibid., 93, 589(1971). ” R. Turner, Theo. Org. Chem., Pap. Kekule Symp., 1458, 67(1959); R. Bohn and Y.-H. Tai, J. Amer. Chem. Soc., a2, 6447(1970). P. Gassman, K. Mansfield, and T. Murphy, ibid., 91, 1684(1969) and references cited therein. “ R. Noyori, Y. Kumagai, and H. Takaya, ibid., 96, 634(1974). W J. Halpern, Accounts Chem. Res., 2, 386(1970). 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 116 M. McKinney and S. Chou, Tetrahedron Lett., 1145(1974). R. Grubbs, H. Eick, and C. Biefield, Inorg. Chem., 12, 2166(1973) . ’” G. Whitesides, J. White, and J. McDermott, J. Amer. Chem. Soc., 95, 4451(1973). G. Whitesides, J. Gaasch, E. Stedronsky, ibid., 94, 5258(1972). m G. Whitesides and J. McDermott, ibid., 26, 947(1974). P. Binger, Angew. Chem., Int. Ed. Engl., 11, 309(1972). S. Otsuka et al., Chem. Comm., 863(1971). R. Pasquale, J. Organomet. Chem., 32, 381(1971). H. Hall, C. Smith, and D. Plorde, J. Org. Chem., 38, 2084(1973). N A. Yamamoto et al., J. Amer. Chem. Soc., 88, 5198(1966). J. Chatt and B. Shaw, J. Chem. Soc., 1718(1960). H. Gilman, R. Jones, and I. Woods, J. Amer. Chem. Soc., 16, 3615(1954). M. Tsutsui and H. Zeiss, ibid., 81, 6090(1959). J. Chatt and B. Shaw, J. Chem. Soc., 705(1959). P. Cossee, J. of Catal., 3, 89(1964). ‘—-—-—-—- m P. Braterman and R. Cross, J.C.S. Dalton, 657(1972). R. Belcher, and A. Nutten, "Quantitative Inorganic Analysis," AMG MacDonald, Butterworth and Co., Ltd., London, 1970, p. 101. A. Cornu and R. Massot, "Compilation of Mass Spectral Data," Heyden and Son Ltd., London, 1966. J. Blum, C. Zlotogorski, and A. Zoran, Tetrahedron Lett., 1117(1975). R. King, "Organometallic Syntheses," Vol. I, Academic Press, New York, 1965, p. 181. CASE l,7-DX ME-DX A D4 D4 C D4 D2 B D4 D0 B D0 D4 C D0 D2 A D0 D0 therefore, ME-D4 = Ethylene—D4 Ethylene—D2 Ethylene-D2 Ethylene-D4 CASE 3: 1. Possible ethylenes: 1-4, Permutations Case A: (l)(l) + (1/3)(l)(l) + (1/3)(l)(l) = 2/3 + 2/3 + 2/3 + 2/3 = 1.6 ME-D2 111 ETHYLENE PRODUCT gym 1 0 O 1/3 2/3 0 0 2/3 1/3 1/3 2/3 0 0 2/3 1/3 0 0 1 Total for ME's = ME-DO = 1 8/3 2—4, METAL ETHYLENE Metathesis with exchange and rotation PRODUCT 22 “_“2 m 1 0 0 2/3 l/3 0 1/3 2/3 0 0 2/3 1/3 0 1/3 2/3 L L L l 1 1 5/3 2-4, 2-3, 3-4. 2-3, 1—3, l-2, 1-3, 1—4, 2. Possible metal ethylenes: 3-4, 1—2. 4 = l = 2 = 3 = CD2 (or CH2) All ethylenes and metal ethylenes produced are D4 (or D0) 4 = 1 ¢ 2 3 Case B: 112 Possible Ethylenes Ethylenes Produced 1—2 D2 l-3 D2 1-4 D4 2—3 D0 2-4 D2 3-4 D2 There are 4 D2 paths, to ethylenes. There are 4 D2 paths, to metal ethylenes. Case C: 4 = l = 2 ¢ 3 Possible Ethylenes Ethylenes Produced 1-2 D4 1-3 D2 1-4 D4 2—3 D2 2-4 D4 3—4 D2 This information is put into the matrices below for CASE 3. Possible PE 3-4 2-4 2—3 Possible PE 3—4 2—4 2-3 1-4 1-3 1-2 ME Produced D2 #2 D0 D4 D2 D2 1 D4 path and 1 D0 path 1 D4 path and 1 D0 path ME Produced D2 D4 D2 D4 D2 D4 There are 3 D2 and 3 D4 (or D0) paths to ethylene. There are 3 D2 and metal ethylene. 3 D4 (or D0) paths to CASE A D4 D4 D4 D0 D0 3’ 0 CO to 0 D0 Ethylene—D4 Ethylene-D2 Ethylene-D2 Ethylene—D4 The system converges on 1.33 = 1,7-DX ME-DX D4 D2 D0 D4 D2 D0 113 ETHYLENE PRODUCT mam l 0 0 1/2 1/2 0 1/6 4/6 1/6 1/6 4/6 1/6 0 1/2 1/2 0 0 1 Total for ME's METAL ETHYLENE PRODUCT a—a m l 0 0 1/2 1/2 0 1/6 4/6 1/6 1/6 4/6 1/6 0 1/2 1/2 L L L 11/6 14/6 11/6 1(11/6) + l/2(l4/6) 1/6(11/6) + l/6(1l/6) + 1/6(ll/6) = 130/36 l/2(l4/6) + 4/6(ll/6) + 4/6(11/6) + (l/2) (14/6) = 1.32 172/36 Ethylene-D4 Ethylene-D2 10. ll. 12. 13. 14. 15. REFERENCES N. Calderon, Accounts Chem. Res., 5, 127(1972). ____________________ m N. Calderon, H. Y. Chen, and K. W. Scott, Tetrahedron Lett., 3327(1967). N. Calderon, E.A. Ofstead, J. Ward, W. Judy, and K. Scott, J. Amer. Chem. Soc., 9?, 4133(1968). C. Bradshaw, E. Howman, and L. Turner, J. Catal., 7, 269(1967). “ E. Zeuch, Chem. Comm., 1182(1968). C. Adams and S. Brandenberger, J. Catal., 13, 360(1969). “ R. Pettit, H. Sugahara, J. Wristers and W. Merk, Disc. of Far. Soc., 47, 71(1969). For a complete presentation see R. B. Woodward and R. Hoffmann, "The Conservation of Orbital Symmetry," Verlag Chemie, Weinheim, Germany, 1970. F. Mango and J. Schachtschneider, J. Amer. Chem. Soc., 89, 2484(1967). N F. Mango, Advan. Catal., 29, 291(1969). F. Mango et al., J. Amer. Chem. Soc., 93, 1123(1971). ____________________ m R. Pettit and G. Lewandos, Tetrahedron Lett., 789(1971). T. Katz and S. Cerefice, J. Amer. Chem. Soc., 91, 2405(1969). ” J. Halpern, P. Eaton, and L. Cassar, ibid., 92, 3515(1970). ” R. Grubbs and T. Brunck, ibid., 24, 2538(1972). 114 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 115 M. Lappert, D. Cardin, and M. Doyle, J.C.S. Chem. Comm., 927(1 2). C. Casey and T. Burkhardt, J. Amer. Chem. Soc., 96, 7808(1974). —_ N For recent reviews of transition metal carbene complexes see D. Cardin, B. Cetinkaya, and M. Lappert, Chem. Rev., 12, 545(1972); F. Cotton and C. Lukehart, Progr. Inorg. Chem., L6, 243(1972). R. Lespieau, Compt. Rend., 158, 1188(1914); C.A. 8:2678. ” The reactions used to convert dimethylsuberate to l,7—octadiene-l,l,8,8-d4 were first carried out using nondeuterated material so that yields could first be maximized. Thus, nmr spectra for all nondeuterated intermediates were available. For a review on pyrolytic cis eliminations see C. Depuy and R. King, Chem. Rev., 29, 431(1960). E. Mutterties and M. A. Busch, J.C.S. Chem. Comm. 754(1974). A. Yamamoto, L. Pu, J.C.S. Chem. Comm., 9(1974). W. Grahlert, K. Milowski, and U. Langbein, Z. Chem., L4, 287(1974). T. Katz et al., J. Amer. Chem. Soc., 94, 5446(1972). ___________________ m J. Osborn et al., ibid., 96, 597(1973). R. Noyori, T. Suzuki and H. Takaya, ibid., 93, 589(1971). W R. Turner, Theo. Org. Chem., Pap. Kekule Symp., 1958, 67(1959); R. Bohn and Y.—H. Tai, J. Amer. Chem. Soc., 92, 6447(1970). P. Gassman, K. Mansfield, and T. Murphy, ibid., 91, 1684(1969) and references cited therein. “ R. Noyori, Y. Kumagai, and H. Takaya, ibid., 96, 634(1974). W J. Halpern, Accounts Chem. Res., 3, 386(1970). 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 116 M. McKinney and S. Chou, Tetrahedron Lett., 1145(1974). R. Grubbs, H. Eick, and C. Biefield, Inorg. Chem., 12, 2166(1973). ” G. Whitesides, J. White, and J. McDermott, J. Amer. Chem. Soc., 96, 4451(1973). G. Whitesides, J. Gaasch, E. Stedronsky, ibid., 94, 5258(1972). “ G. Whitesides and J. McDermott, ibid., 96, 947(1974). P. Binger, Angew. Chem., Int. Ed. Engl., L1, 309(1972). S. Otsuka et al., Chem. Comm., 863(1971). R. Pasquale, J. Organomet. Chem., 32, 381(1971). H. Hall, C. Smith, and D. Plorde, J. Org. Chem., 38, 2084(1973). “ A. Yamamoto et al., J. Amer. Chem. Soc., 88, 5198(1966). J. Chatt and B. Shaw, J. Chem. Soc., 1718(1960). H. Gilman, R. Jones, and I. Woods, J. Amer. Chem. Soc., 16, 3615(1954). M. Tsutsui and H. Zeiss, ibid., 81, 6090(1959). J. Chatt and B. Shaw, J. Chem. Soc., 705(1959). P. Cossee, J. of Catal., 3, 89(1964). _— ’\J P. Braterman and R. Cross, J.C.S. Dalton, 657(1972). R. Belcher, and A. Nutten, "Quantitative Inorganic Analysis,“ AMG MacDonald, Butterworth and Co., Ltd., London, 1970, p. 101. A. Cornu and R. Massot, "Compilation of Mass Spectral Data," Heyden and Son Ltd., London, 1966. J. Blum, C. Zlotogorski, and A. Zoran, Tetrahedron Lett., 1117(1975). R. King, "Organometallic Syntheses," Vol. I, Academic Press, New York, 1965, p. 181. 52. 53. 54. 55. 56. 57. L. Venanzi, 117 J. Chem. Soc., 719(1958). G. van Hecke and W. Horrocks, Inorg. Chem., 5 1968(1966). ’\2 K. Ziegler et al., Angew. Chem., 67, 541(1955). G. A. J. Natta and I. Pasquon, Yamamoto, T. Yamamoto and S. Ikeda, J. Amer. Soc., 93, 3350(1971). Bandrup and E. Immergut, Vol. II, Interscience, New "Polymer Handbook," York, N.Y., 1966, p. Adv. in Catal., £1, 1(1959). Chem. 341. x} Lx..\.. x9: \ ERSITY LIBRARIES A.\\.....3su. \ (3W 4. x 1.1.. \.vl. AN STATE Uva H‘ . n . n: .3. {1.71. . 0.1.52 1 :1 nu. _ xi... writ... .1. law... . .fi .....t.kh24.\¢.}.xu. 3.5.31. 4 E54! . «In. 53% 1.1:...{ky .3: .51! 3...}... 3...“: it}? 5.3,...»- Flkufiw. 23‘. 5.3% . - . . an. 5.5 if“: 3.52.5.3 5 \u' I? if... .\\u 6.1} v1... .1. .QIuA.au\“ 3.11“... <\ 4 law . is... a ..:...........2......-. 1. : . . . . . . , 3.5.15.6}. 1v. F. 213:... 1.2.3:; , . . , . . :4 575.13.. . 4.1%. «T»: 3““? I umkfistf run hut-(vi. $4! aEAIXMIVE ‘IL Yr.” «a... snug. .233..." MW. m... i. .u .............§...a .14 . . mean. V . .. x . . . . . .2 . .L..thfia§vd immussvnf :21. V .. . z. s . 5“. 1.? ii“! 1? um Vii 9- .‘I 4 \ 1. 3. u)..».».f u . Hawtifiuflga... .5 5 f t X. : ... . «8.x . u. .6: hrs}... .u tdvfivfifiiEuan.u\uww-VVR6.§J“2§ . . . .. u?- Ea .. Eviacifikfiu an .3... .s._,.1vt§:m\.?un .. ghwfluVEn .11 5., L’swofienhwvt rota”; J » Ill»). x is . : 5. {3:3 it” ,2! . a». 5:... It 1...... .1 3:23.». .. 3 €3.73}... .. “Hugh: 5:: JADE" , which; . any“. . .. l V? ,7. . 4. .2: A w .1. u... 2.3:... w...“ id 1.. qiflwfiwrflh Kc .. L. “W... .. Lew _ . .161 E . .5. L. M; 5:. 2”,... w , g 2.. .fi 3,. 5.2%“. migmihséz