Reprinted from Organometallics, 1985,4' 1819. Copyright O 1935by the American Chemical Society and reprinted by permission of the copyright owner. Suppressionof UnwantedHeterogeneous Platinum(0)-Catalyzed Reactionsby Poisoningwith Mercury(O)in SystemsInvolving CompetingHomogeneous Reactions of SolubleOrganoplatinum Compounds:Thermal Decomposition of 1 Bis(triethylphosphine )-3, 3, 4, 4-tetramethylplatinacyclopentane

George M. Whitesides,*2Marifaith Hackett, Robert L. Brainard, Jean-Paul P. M. Lavalleye, Alan F. Sowinski, Alan N. Izumi, Stephen S. Moore,3 Duncan W. Brown,a and Erin M. Staudts

Departmentsof Chemistry,Haruard University,Cambridge, Massachusetts 02138, and MassachusettsInstitute of Technology,Cambridge, Massachusetts 02139

ReceivedDecember 27. 1984

Thermal decompositionof bis(trialkylphosphine)-3,3,4,4-tetramethylplatinacyclopentanesin hydrocarbon solventsyields two major products: 2,2,5p-tntramethylbutaneand l-methyl-l-tert:butylcyclopropane. The former appearsto be generatedby heterogeneousprocess(es) catalyzed by platinum(0) (colloidal and/or solid) formed during the decomposition;the latter is the product of a homogeneousreaction sequence. The existenceof competingheterogeneous and homogeneousreaction pathwayscomplicates the study of this thermal decomposition. The addition of mercury(O)to the systemselectively suppresses the heter- ogeneousreaction by poisoningthe bulk platinum(O). This techniquehas beentested in one well-defined model system-the homogeneoushydrogenation of dineopentylbis(triethylphosphine)platinum(Il)to dihydridobis(triethylphosphine)platinum(Il) and neopentanein the presenceof the heterogeneousplat- inum(0)-catalyzedhydrogenation of l-methylcyclopenteneto methylcyclopentane-and found to eliminate the heterogeneousreaction while leavingthe homogeneousone essentiallyunaffected. Application of this techniqueto the thermal decompositionof 3,3,4,4-tetramethylplatinacyclopentanesimplifies the reaction by eliminating 2,2,3,3-tetramethylbutaneas a product. Similarly, addition of mercury(0) to solutions containing H2 and (1,5-cyclooctadiene)dimethylplatinum(Il)in cyclohexanesuppresses the autocatalytic, heterogeneousplatinum-catalyzed conversion of the organoplatinumcompound to (inter alia) cyclooctane, methane,and platinum(0). In other reactionsexamined-especially the high-temperaturethermal de- compositionsof (1,5-cyclooctadiene)dineopentylplatinum(II)and of cis-bis(cyclopentyldimethyl- phosphine)dimethylplatinum(Il) in hydrocarbonsolvents-mercuq/ poisoningdoes not successfullyseparate homogeneousand heterogeneousreactions. In the latter instance,an additional complicating process- apparent reaction of mercury(O)with the organoplatinum compound-introduces reaction paths which seemto generatemethyl radicals. The paper includes a simple preparation of 1,4-dihalo-2,2,3,3-tetra- methylbutane and illustrates the use of this material as a reagent for the preparation of 3,3,4,4-tetra- methylmetallacyclopentanesvia the correspondingdi-Grignard reagents.

Introduction additional complementary methods are still required. In the courseof studies of mechanismsof reactions of A problem that is encountered frequently in studies of platinum(Il) organometallic compounds,we have often mechanismsof reactions of transition-metal organometallic encountered situations in which we wished to distinguish compounds is that of distinguishing between truly homo- between homogeneousstoichiometric reactions and het- geneousreactions and heterogeneousreactions catalyzed catalyzedreactions. Unwanted heterogeneous by bulk or finely divided metal produced by decomposition erogeneous reactionshave been particularly troublesome in high tem- of the (originally) homogeneousorganometallic species. A perature (100-250 oC) reactions involved in studies of number of methods have been proposed for detecting carbon-hydrogen bond activation.l6 We describe here a heterogeneouscomponents in complex mechanisms,Flsbut procedure which addressesthe problem of distinguishing between homogeneous and heterogeneous reactions by (1) Supported by the National Science Foundation, Grants CHE- selectivepoisoning. Addition of mercury(0) or other cat- 05143 and CHE-08702. alyst poisons to reaction mixtures suppressesthe catalytic (2) Correspondence should be addressed to G.M.W. at Harvard activity of the bulk platinum metal (by either physisorp- University. (3) National Institutes of Health Postoctoral Fellow, 1979-1981. tion or amalgamation) but, under appropriate circum- (4) D. W. Brown was supported by a grant from Hercules, Inc. stances,does not influence the homogeneous,nonhetero- (5) Massachusetts Institute of Technology Undergraduate Research geneouslycatalyzed reactions of the organometallic species Opportunities participant. yet (6) Hamlin, J. E.; Hirai, K.; Millan, A.; Maitlis, P. M. J. Mol. Catal. of interest. We do not have a sufficiently wide range 1980, 7, 543-544. of experiencewith this technique of selectivepoisoning to (7) Crabtree, R. H.; Mellea, M. F.; Mihelcic, J. M.; Quirk, J.M. J. Am. be able to set well-defined limits to its applicability: in Chem.Soc. 1982,104,107-LLl. Crabtree,R. H.;Demou, P. C.;Eden, D.; Mihelcic, J. M.;Parnell, C. A.;Quirk, J. M.;Morris, G. E. J. Arn. Chem. some circumstancesit is clearly a useful technique; in Soc. 1982,104,6994-7001. Crabtree, R. H. CHEMTECH 1982,506-512. (8) Laine, R. M.;Rinker, R.G.; Ford, P. C. J. Am. Chem. Soc.1977, 99,252-253. (13) Baudry, D.; Ephritikhine, M.; Felkin, H. J. Chem. Soc., Chern. (9) Gregov, H. P.; Gregov, C. D. Sci. Am. 1978,239 (I), lL2-728. Commun.1982,606-607. (10) Berne, B. J.;Pecora, R.'D5rnamic Light Scattering";Wiley: New (14) Collman, J. P.; Kosydav, K. M.; Bressan, M.; Lamanna, W.; York. 1976. Garret, T. J. Am. Chem. Soc. 1984, 106,2569-2579. (11) Laine, R. M. J. Mol. Catal.1982,14,137-169. (15) Anton, D. R.;Crabtree, R. H. Organometallics1983,2,855-859. 'Catalysis: (12) Martino, G. Heterogeneous and Homogeneous'; (16) Whitesides,G. M.; Reamey,R.H.;Brainard, R. L.; Izumi, A.N.; Delman, G.;Jannes, G., Eds.; Elsevier: Amsterdam, 1975;p 439. McCarthy, T. J. Ann. New York Acad. Sci.1983,56-66.

0276-7333I 85 I 2304-1819$01.50/0@ 1985American Chemical Societv al. 1820 Organometallics, Vol. 4, No. 10, 1985 Whitesideset

SchemeI. Preparationof Bis(triethylphosphine)'3,3,4,4' Scheme II. Preparation of tetramethylplatinacyclopentane ( 1 ) Perdeuterated Triethylphosph ine

and Related Compounds o zn, riupn o No, ooac t) HBr, H2s04(78%) ccl3coc5Hl'l--CD.CcCo.__:__^-cDac02oH+:c|]cf,2)3P (86 | ) LiAtH4,THF %) DME,DZo EtzO 2) Mg, Er2O(99%) o v'n o' - t2 t"oo 2) TsCl (717.) (4O%l t42"/") 3) (pho)lp (54%) N .-1,tn )-'( -;;- - t IO N 4.n er'o h 3) LiX (X=Cl,93%l o 4.n X= Br, 90Y") geneousreaction and that 2,2,3,3-tetramethylbutane3 is an artifact resulting from a heterogeneous,platinum-cat- alyzed reaction of undetermined mechanism. Tetra- Ms,rHF (coD)Prcl2 methylcyclobutane is not an important product of decom- Yx r 1^"n, | W\-r position of 1. +'x 4o"t t"nx 60% "Q 0o, X.Cl to Results b,X=8r X2MLn PEr3 Syntheses. To study the thermal decomposition of - (557o) bis(triethylphosphine) 3,3,4,4-tetramethylplatinacyclo- pentane, we required efficient preparations of 2,2,3,3- per- -\.a. 7 PEt" tetramethylbutane-1,4-diyldi-Grignard reagentand lPt F \ert, deuterated tti.thylphosphine. An improved synthesis22 of 1,4-dic hlor o-2,2,3, 3-tetramethylbutane23 used 2,2'- azo- bis(2-methylpropionitrile) (AIBN) as starting material others, although it may solve the problem of heterogeneous (SchemeI). Organometallic compounds were prepared platinum-catalyzed reactions, it appears to introduce other using the di-Grignard reagent. Metallacyclopentaneswere competing and complicating processes.The purpose of this obtained from the reaction of the di-Grignard reagentwith paper is to outline our experiencewith this technique. cis-dichloro(1,5-cyclooctadiene)platinum (II), dichlorodi- The use of mercury(O) to poison undesired heterogene- methylsilane, dichlorodiphenylsilane,dichlorodiphenyl- ous transition-metal-catalyzed reactions is not original with germane,dichlorodiphenylstannane, and dichlorophenyl- us. We have not, however, been able to identify its ori- phosphine. Details of the first of these reactions are ginator(s).u-le This technique is, apparently, a part of summarizedin the Experimental Section;the other prep- the oral tradition of organometallic experimental practice arations are describedin supplementary material to this but has not previously been examined independently. paper contained in the microfilm edition. No isolable The immediate focus of this study was €rninvestigation products were obtained from reaction with dichlorobis- of the mechanism of formation of 1-methyl-L-tert-butyl- (triphenylphosphine)palladium(II),bis(benzonitrile)di- cyclopropane (2) on thermal decomposition of bis(tri- chloropaltadium(II), or dichlorobis(triphenylphosphine)- ethylphosphine) -3,3,4,4-tetr amethylplatinacyclopentane nickel(II). Reactionwith dichlorobis(cyclopentadienyl)- (1) (eq 1; L = PEh). Our interest in I stemmed from its zirconium and dichlorobis(cyclopentadienyl)hafnium '.o,^l yielded thermally unstable,air-sensitive complexes which c-c6Hr2 ry * V + Pt(o)+others (r) were not carefully characterized. The reaction of the di- ,/"+ # ,,f... A Grignard reagent with palladium, zirconium, nickel, and t23 hafnium complexeswas not pursued thoroughly and more persistent might result in the successfulsynthesis structural similarity to bis(triethylphosphine)-3,3-di- efforts metallacyclopentanes. methylplatinaryclobutane (5). This compound decomposes of the corresponding synthesisof perdeuteratedtriethyl- to platinum(O) and dimethylcyclopropane (eq 2; L = A straightforward phosphine is outlined in SchemeII. This procedure, which t -c(cHr)4t.rr1ar. source,appears amenable ..r,x _.+ usesonly D2O as a deuterium | L^pt(o) f>

MCP c. o ttr14 2Hz L... 6 /'H c, + 2 c(cH3)4 (4) o50 L/t .pt + H^ (K c ,rtt-a o () OL,'CC 4 De o The choice of MCP as the olefinic substrate to be used in this study was based on its rate of hydrogenation: under o the conditions used, the two reactions summarized in eq loo 3 and 4 proceeded at comparable rates. Reaction of 4 with I dihydrogen at 40 oC in the presenceof MCP proceedsby a fiist-oider reaction having rate constant k = 10.7 x 10-4 c .9 good o s-1(Figure 1A). This rate constant is in agreement I previous of the hydrogenolysis of 4 in the with studies 3so ( ll + H C \r/ 2 absenceof olefin.% Addition of a heterogeneousplatinum ()o (platinum on scintered glass) increasesthis rate catalyst ;e constant by approximately 30%. This small increase ap- pears to represent a small heterogeneouscontribution of unknown character. MCP present in the reaction mixture containing the heterogeneous platinum catalyst hydro- o2040 genates smoothly (Figure 1B). In the presence of 4 but Time (min) in the absence of the heterogeneous platinum catalyst, no Figure 1. Reaction in the absence of catalyst poisons. (A) reduction of MCP occurs under these conditions (Figure Reaction of cis-dineopentylbis(triethylphosphine)platinum(4) 1B). Thus, neither 4 itself, nor its reaction products, are (0.022rrmol in 6 mL of n-octane) with 2 atm H2 at 40 "C: O, in hydrogenation catalysts for MCP. The observation that the presenceof 0.030mmol of MCP; O, in the presenceof 10 mg glass and MCP. (B) Reaction MCP hydrogenates smoothly in the presence of 4 is im- of a-29VoPt on scintered catalyst (0.030 mmol in 6 mL of n-octane) with 2 atmH2 at 40 portant that neither this compound nor its of MCP in indicating oC: E, in the presenceof 10 mg of a29Vo Pt on scintered glnss reaction product dihydridobis(triethylphosphine)plati- catalyst and 4 (0.022mmol);1, in the presenceof 4 (0-022mmol). num(Il) is a catalyst poison for bulk platinum(0), although free triethylphosphine is a strong catalyst poison for roo d q platinum(0).25 The data given in Figure 2 summarize the course of the I reactionsgiven in eq 3 and 4 proceeding in the same re- c o action uesselin the presenceand the absenceof mercury .; -50 8Lr (added of metallic mercury to o in the form of a small drop A * * *, c ,'Pt',1- (-! magnetically stirred reaction mixtures). The rate of the o o H" ""t<* stoichiometric hydrogenolysis of 4 is essentially unaffected s by the presenceof mercury(O);that is, it has a value in- I distinguishable from that in homogeneoussolution (Figure o o2040 2) (the Experimental Section contains ratp constants). The (min) heterogeneousplatinum-catalyzed rcaction is completely Time presence mercury (Figure 2). Figure 2. Reaction in the presence of catalyst poisons. (A) suppressedin the of oC: We examined a number of other potential heterogeneous Reaction of 4 with 2 atm of H2 in n-octane at 40 O, in the presenceof Hg; a, in the presenceof CO; D, in the presenceof catalyst poisons for species which might be compatible oC: Pfnr. (B) Reaction of MCP with 2 atm H2 in n-octane at 40 reactions. Carbon with homogeneous organometallic o, in the presenceof Hg; A, in the presenceof CO; r, inthe monoxide26(added at the start of the reaction) and tri- presence of PPh3. Both 4 and MCP were present in the same phenylphosphine gave results indistinguishable from those solution for these studies; quantities are as given in Figure 1. observed with mercury(0). Other catalyst poisons- including thiophene and dibenzothiophene-were less addition to other minor products (summarized in the = successful. Experimental Section) (eq 5; L PEh). The reaction

Thermolysis of Bis(triethylphosphine't-3,3,4,4- ceHr2 !- ryv* *.czHe *, czHq. * Pr(o) tetramethylplatinacyclopentane (1). We have explored ,^ ,'1.. / tf7 "c products / (1?"/ol (88 (5%) the influence of mercury on the of decomposition L 4/ 2 3 %) of 1.. The products of decomposition in solutions con- Pfl{ ( ^-a q : -6"t2 ( taining no mercury were a mixture of 2,2,3,3-tetra- - 1 colorlesssolulion ) -methyl- -butylcyclopropane, I H9(O) ,4\ methylbutane and 1 L-tert in lan oa 2 ( r0o o/.)

(24) H.; Whitesides, G. M. Am. Chern. Soc. 1984, 106, mixture is brown-black at its conclusion and shows variable Renmey, R. "/. 81-85. densities of platinum mirror deposited on the walls of the (25) Phosphines, as a group, are strong poisons for heterogeneous flask. Olefins which would be expected to be products of catalysis: Maxted, E. B. Adu. Catal.l95l, 3, 129-177. Baltzly, R. Ann. platinum-catalyzed ring opening of the cyclopropane N. Y. Acad. Sci. 1967, 31-45. the (26) Although no insertion of carbon monoxide into platinum-carbon (2-t ert -butylbut- 2-ene, 2-t er t-butylbut- 1- ene, and others) bonds was observed in this case,carbonylation of M-C bonds could be were occasionally observed as products and became major an important side reaction in other systems. See, for example: Collrnan, 'Principles products if platinum dioxide was intentionally added to J. P.; Hegedus, L. S. and Applications of Organotransition Metel Chemistry"; University Science Books: Mill Valley, CA, 1980. the reaction mixture. L822 Organometallics,Vol. 4, No. 10, 1985 Whitesideset al.

o N tf, I vcrticol oris o

(L o o ;* f lcxiblc tubing condenser

(\,l ?f) ,3,- thermomctcr -

(olumrnum ro

corh plu0 (boorlng) l2O n n+2O

Trme(mrn) Insulotton Figure 4. Reactionof CODPT(CH3)2.ll, with H2(1.3 atm) in n-heptane(5 mL, 27oC): O. 15mg of CODPT(CH.r)z;O, 15 mg of CODPT(CH3):and 3.5 g of mercurl'(0).n = 1500;o, 15mg of CODPT(CH"):and i0 mg of controlled-poreglass (see Experi- mentalSection); r, 15mg of CODPT(CH.):,10 mg of controlled poreglass, and 3.5 g of mercur\'{(-)).n = 1030.

duction was completell'suppressed,and the small quan- reoctton tube tities of ethaneand ethyleneobserved essentially disap- peared. The only significant h1'drocarbonproduct was 2. (c.o.2.5" In hngth) The solution at the conclusionof the reaction was clear and colorless. Tsflon plug Although the interpretation of these facts remains un- (bcorlng) clear in detail, we propose that the formation of tetra- methylbutane in these reactions is either entirely or in major part a heterogeneousreaction catalS'zedb1'colloidal hsol Ing or bulk platinum(0). This heterogeneousprocess can be eliminated or suppressedby the addition of metallic mercury. RBaction of ( 1,S-Cyclooctadiene)dimethylp latin um- (II) (ll) with Dihydrogen. We havealso examined the influence of mercury(0) on the heterogeneous,platinum- Figure 3. Apparatusused for thermaldecomposition of 1 in the catalyzed hydrogenation of (d i olefin ) dialkylplatinum (I I) presenceof mercury(0).To achieveintimate contact of the so- complexes(eq 6).za'zsIn the absenceof added Hg(O),an lution with the surfaceof the mercurybead, the samplesolutions positioned ttt, werepositioned perpendicular to a shaft ca.20o from ('r',"^" ( the vertical in sealedreaction tubes and rotated at 60 rpm. -.*o + + ? cH4 (6) Rotation causedthe mercurybead to roll from one end of the L>, ri t_> reactiontube to the other and ensuredefficient mixing. tl Variability in product yields, and the obvious require- ment that hydrogen atoms be transferred to the tetra- initial irreproducible induction period is followed by up- methylplatinacyclobutane ring to produce tetramethyl- take of H2 and precipitation of platinum metal (Figure 4). butane (3), suggestedthat platinum-catalyzed heteroge- This autocatalytic processis entirely suppressedby ad- neous reactions played an important role in the reaction. dition of mercury(0). Decomposition of I containing perdeuterated triethyl- When the hydrogenation of 11 is carried out in the phosphine ligands established that significant but irre- presenceof controlled-porosity glassbeads, the induction producible quantities of deuterium were incorporated into period is much shorter and the rate of hydrogenation in- both 2 and 3 (up to approximately LSVofor 2 and up to creasessubstantially. In the presenceof added mercury(O), 45% of a combination of dr and d2 material for 3). Within the rate of reaction decreases,but hydrogenationcontinues, the limits of GC detection, no tetramethylcyclobutane or albeit at a much slower pace. Addition of mercury(0) to 2-methylpropene was produced when decompositions were the hydrogenationof 11 in the presenceof controlled-po- carried out in alkane solutions; tetramethylryclobutane was rosity glassbeads after 50% conversionof 11 to products formed in l7% yield when the tri-n-butylphosphine ana- confirms this result: hydrogenation is slowed, but not loguewas decomposedin methylene chloride at 118 oC.27 halted, by the presenceof Hg(0) (Figure 5). Addition of a small bead of mercury to the reaction Although the courseof this reaction is discussedelse- solution before decomposition and vigorous mixing of the where.28'2ethe essentialfeature is that reaction of 11 and resulting mixture during decomposition using the appa- H2 proceedsrapidly in the presenceof bulk platinum(0). ratus in Figure 3 resulted in a qualitative changein the The role of Hg(0) thus appearsto be to poison initial traces mixture of products obtained (eq 5). The reaction pro- of Pt(0) produced by unknown mechanisms.soThe failure ceeded somewhat more slowly, tetramethylbutane pro- (28) McCarthy, T. J.; Shih, Y. S.;Whitesides,G. M. Proc. Natl. Acsd. Sci. U.S.A. 1981.78. 4649-4651. (27) Sowinski, A. F. Ph.D. Thesis, MassachusettsInstitute of Tech- (29) McCarthy, T. J. Ph.D. Thesis,Massachusetts Institute of Tech- nology, Cambridge, MA, 1980. nology, Cambridge, MA, 1982. HornogeneousReactions of Organoplatinum Compounds Organometallics,Vol. 4, No. 10, 1985 1823

Tobls I. TherEolyris ot Bir(cyclopertyldiEethylpho.phine)diE€thylpletiDuDr ln the Prorence and Abeence of Mercury(0) rel yields (%) of products(isotopic compositns) organometallic componentso methane ethane ethylene = lc-CsHgP(CH3) 2l 2Pt(CHJ 2 95 (dtl do = 0) 4u 4 (d1lds 0) = lc-CsHsP(CH3)2l2Pt(CHs)2,Hg(0) 74 (dtldo = 0.14) 2o 24 (d1ldo Q) = = Ic-C5HeP(CHJs]2Pt(CHe)2,Hg(0), (CD3)2Hg 16 (illdo = 0.28,dtldo 0)' 76d 8 (dtldo 0.14) = ggd = Ic-C5HeP(CHs)2]2Pt(CHs) z(CDg)zHg l0 (d4lds= 0.76,drldo 0)" 2 (drldo 1.11) (CHJzHe 65 (drldo = 0.35) 1b 34 (drldo = 91 (CHr)zHg, Hg(O) 77 (dLldo= 0.52) 1a 22 (drldo = 0) oC. tonly oTherrnolyzedin sealedtubes containing c-CsDr2fo! 4 h at L8t CrH, was obseNed. "Of the methane ptoduced, l2Vo was ds. dOnly C2D6was obeerved. 'Of tho methane produced,ca. 47o wae de.

of mercury(O) to halt the reaction of l1 and H2 when the 'tr,o hydrogenation is carried out in the presenceof glass beads l'o lo may result from the inability of Hg(O) to poison the Pt(0) ? buried in small-diameter pores. Platinum(0) on the outer 9 surface of the beads is, however, still susceptible to poi- o soning by mercury. Thus, in this system, the successof o -J mercury(0) in poisoning a heterogeneousreaction seems o.5 to depend on physical details (distribution of platinum in N FO pores). f mercury(O) . J-\./r-- (1,5-Cyclooctadiene)di- + Thermal Decomposition of = oddsd L I neopentylplatinum(II) (12). We have examined the o I ollorod lo worm thermolysis of (1,5-cyclooctadiene)dineopentylplatinum(Il) oC ca to 25 (12) in the presenceand absenceof mercury(0) in an effort 0.o to suppress the troublesome heterogeneotxldecomposition 24 of this compound.2e Solutions of 12 in benzene-d6were (min) decomposedthermally in sealedglass tubes in the presence Time and absence of mercury using the off-axis stirring tech- Figure 5. Poisoningof Pt(0)/glasscatalyst by mercury(0). nique (Figure 3). Before thermolysis, these solutions were Reactionof 15 mg of CODPT(CHg)2,11, with H2 (1.3atm) in (5 oC presenceof 8.6mg of Pt(0)/glass pale yellow (as is crystalline 12) and the mercury was bright n-heptane mL) at 0 in the catalyst(see Experimental Section). Mercury(0) (9.0 g) wasadded and shiny. After thermolysis, the inside walls of the tubes after7.5 min, not containing mercury were covered with a black solid and the reaction solution was brown and contained a suspended to the observed hydrocarbon products. black precipitate. In contrast, the inside walls of the tubes Thermal Decomposition of cis-Bis(cyclopentyldi- containing mercury were clean and the reaction solutions methylphosphine)dimethylplatinum(II). Thermal was clear and colorless. The mercury was covered with a decomposition of organoplatinum compounds at high dusty film. Thermolyses conducted in the absence of temperature in deuterated hydrocarbon solvents often mercury produced neopentane which was composed of results in incorporation of small amounts of deuterium into L4Vo neopentane-dl. Thermolyses conducted in the hydrocarbon products derived from the platinum complex. *activation' presenceof mercury produced only 4To neopentane-d1; This apparent of solvent carbon-hydrogen dineopentylmercury was also produced (eq 7; yields are bonds might occur either homogeneouslyor heterogene- ously.31,32In order to explore whether the observedhy- + N€oponione COD + Pr(O) representeda homogeneousprocess, o/o drocarbon activation -L 85 % tOO /\ d,/d^= o.16 we examined the thermal decompositionof cis-bis(cyclo- (C0D)Pr. - I -U \ , (7) pentyldimethylphosphine)dimethylplatinum(Il)s (selected \, C.Do, Hg(O) ' -- * a2 ----t Neopenlone + coD +H9 becauseit had useful solubility in hydrocarbon solvents) t3t .c, 6 h in alkane solvents. The cyclopentyldimethylphosphine >/g7 22% loo"a "h resist metalation: formation -ld./d^. -u O.O4 ligand would be expected to of a platinacyclopropaneby activation of a C-H bond a expressedas percent conversionsince 5% 12 (without Hg) to phosphorus is uncommon,s4and metalation B or 7 to and 25To 12 (with Hg) remained at the end of these re- actions). We suggestthat the transfer of deuterium from (31) Clarke, J. K. A.; Rooney, J. J. Adu. Catal. 1976,25, 125-183. benzene-d6to the neopentyl moieties in the absenceof Somorjai, G. A. "Chemistry in Two Dimensions: Surfaces"; Cornell added mercury is due to a heterogeneounreaction catalped University Press: Ithaca, NY, 1981;Chapter 9, pp 479-515. by colloidal or bulk platinum(O). Mercury is apparently (32) Activation of a C-H bond of a methyl group bonded to platinum able to suppress but not to stop entirely this transfer. would produce a platinum carbene,e.g., LzPt(CHt(CHJ(H) or L2PI(CH2) (after reductive elimination of methane). Although similar platinum- Thus, either the strategem of adding mercury(O)to sup- methylene specieshave been postulated as intermediates in heterogene- press heterogeneousreactions is only partially successful ous platinum-catalyzed reactions, the only soluble platinum-carbene complexes isolated thus far contain a heteroatom-stabilized carbene in this case, or there is an independent, homogeneous 'Organometallic moiety. See: Belluco, U. and Coordination Chemistry reaction which incorporates deuterium from benzene-d6 of Platinum"; Academic Press: London, 1974; pp 282-292. Hartley, F. into neopentane. The observation of dineopentyl- R. "The Chemistry of Platinum and Palladium";Wiley: New York, 1973; mercury(Il) as product establishesthat transmetalation pp 101-102,348-350. (33) Although formation of three-membered metallocycles by other can occur under these conditions. Thermal decomposition routes, such as reduction of metal-halide bonds, is a well-known reaction, of the dialkylmercury(Il) compound may also contribute only one caseof phosphine metalation to produce a platinacyclopropane has been reported. See: Bresciani, N.; Caligaris, M.; Delise, P.; Nardin, G.; Randaccio, L. J. Am. Chem. Soc. 1974, 96, 5642-5643. (30) The initial traces of platinum(O) may result from a glass- or (34) Metalation of tricyclopentylphosphine by platinum has been ob- light-catalyzed reaction or from a homogeneousprocess of which we are served, but whether metalation results from oxidative addition of a ligand unaware. C-H bond or from a radical reaction is unclear. See ref 55. L824 Organometallics,VoI. 4, No. 10, 1985 Whitesides et al.

phosphorus would produce a (seemingly) strained com- Table II. Thermal Decomposition of pound.35 t ran s - ((C 'De)3P )2Pt (C D2C (CD3)s)Cl ( I 4 ) in Cyclohexaneo Thermolysis of cls-bis(cyclopentyldimethylphosphine)- (cD2)2c- dimethylplatinum(Il) in cyclohexane-d12in the absenceof (CDs)sbl d mercury to low conversion produces methane, ethane, and Hg (mg) NpD-H + % -114r)2.-i196 pobsd ethylene (eq 8, Table I). Thermolyzed solutions are pale solv added Npt-D Np;-D" yellow and completely homogeneous in appearance. CoHt, 6.8 t4 L.74+ 0.07 gaseous CoHrz 740 7.0 15 1.78+ 0.07 Methane, the major product, contains negligible '/ deuterium. CeDt, \ 98 1.71+ 0.03 When a bead of mercury is present, the thermolyzed CrDtt 590 i.2 96 1.80+ 0.09 solutions are colorless and the mercury looks dirty. More oThe volume of solution was ca. 0.5 mL. The concentration of importantly, a significant amount (L2To) of the methane LD2PINpDCI was 0.029-0.042 M. Reaction solutions were ther- oC produced is monodeuterated (eq 8). We believe that this malized at 133 until -25Vo starting material remained (-15 h). bRelative quantities of 1,1-dimethylcyclopropane and neo- 9- coDrz pentane were determined by GC. The experimental accuracy of -;; t'o the yield of 1,l-dimethylcyclopropane produced is estimated at *4Vo. Thus, all variations listed here are within experimental er- (8) []ercH.r,1rr,,"rr,/ ror. "Experimental accuracy estimated at *3%. Npo-D is neo- -...-i-!oo-lo t' pentane containing a deuterium distribution compatible with re- 13 -t cHo + cH3D oc placement of the NoD-Pt bond of 14 by NpD-D on decomposition. r8r = -9 : % NpD-D (NpD-D/(Npo-D + NpD-H)) x 700%; NpD = CDzC- 'activation' (CDs)s. dThe thermal decomposition of l4 occurs by a mechanism apparent of deuterated hydrocarbon solvent which is half-order in 14.55 This kinetic order is consistent with is not a homogeneousprocess, but a mercury-promoted the postulated mechanism for the decomposition of 14 in which side reaction. reversible phosphine dissociation precedes the rate-limiting step. The mechanism of this side reaction has not been de- The error limits for these half-order rate constants are reported at fined. One possibility involves a mercury-platinum in- 95% confidence levels. terchange reaction yielding dimethylmercury. Di- methylmercury could then decomposeto methyl radicals to surface effects is notorious.4l'43'45-50 In the presenceof 13, ethane (derived from dimethyl- and generate CH3D by deuterium atom abstraction from mercury) becomesthe major product. We do not presently solvent. Although we cannot flumly exclude such a process, understand the mechanistic basis for this change in we have not observed dimethylmercury as a reaction product distribution. Metallic platinum is knownsl to product.s6 Moreover, the data in Table I suggestthat if catalyze the reaction dimethylmercury were formed under the reaction condi- tions, it would decomposeprimarily to ethane (seebelow). Pt(C) ------) + As ethane is only a minor (L-2%) product in the ther- RrHg R-Fi Hg(O) (9) molysis of cis-bis(cyclopentyldimethylphosphine)di- methylplatinum(Il) in the presenceof mercury, the for- The reaction of zerovalenttriphenylphosphine platinum mation of dimethylmercury under the reaction conditions and palladium complexeswith RrHgsz'sesuggests a plau- does not seem to be significant. sible mechanism (eq 10). Reaction of dimethylmercury platinum(II) The marked difference between the products of thermal with complexes, e.g., dichlorobis- (phosphine)platinum decomposition of di(methyl-d3)mercury in the presence complexes,is, however, quite slow.sa and absenceof 13 suggeststhat the chemistry involved in Lz*Rz + Xg(O) these systems is complex. In the absence 13, major of the (M= pr, n =3, L. pph3) products are methane and ethylene. The first product, + HeR2 | methane, is not unexpected;other workerss?'shave found LnM(O) ------+ R-Hg-M-R ilo) (n-2) methane to be the sole product of thermolysis of di- L l- methylmercury in the presenceof cyclopentane. We ob- LnM * Hg(O)+R-R served that much of the methane generated from the pph3) thermolysis of unlabeled dimethylmercury in C6D12is CHa, (lrt.pC, n.4, L= rather than CH3D. This result, as well as the relatively high yield of ethylene, suggeststhat some processother (42) Kallend, A. S.; Purnell, J. H. Trans. Faraday Soc. 1964, 60, than attack by methyl radicals on the solvent is important. 93-102,103-118. In fact, the detection of ethylene from the thermolysis of (43) Yeddanapalli,L. M.; Srinivasan,R.;Paul, V. J. J. Sci.Ind. Res., neat dimethylmercury has been used as evidence for rad- Sect. B 1954, 13,232-239;Chem. Abstr. 1955,49,4387b. ical-radical recombination mechanisms presently (44) Cunningham, J. P.; Taylor, H. S. J. Chem. Phys. 1938,6, 359-367. of un- (45) Yerrick, K.B.;Russell, M. E. J. Phys, Chern 1964,68,3752-3756. defined nature and surface-mediated heterogeneousre- (46) Ingold, K. U.; Lossing,F. P. J. Chem. Phys. 1953,27,368, actions.ro The sensitivity of this decomposition reaction 1135-1144. (47) Gowenlock, B. G.; Polanyi, J. C.; Warhurst, E. Proc. fi. Soc. London, Ser.A 1953,218,269-289. (35) The liquid phase of the reaction mixture (ca. 0.5 mL) was ana- (48) Rice, F. O.; Herzfeld, K. F. Phys. Colloid Chem. [95[, 55, "/. lyzed by gas chromatography (9-ft 20VoSE-30 column on 80/100 Chro- 975-985,986-987. masorb P). All components of the reaction mixture eluted before cyclo- (49) Waring, C. E.; Pellin, R. J. Phys. Chem. 1967,71,2044-2053. hexane-d12.Spiking the reaction mixture with dimethylmercury (L pL, (50) Rasuwajew,G. A.;Koton, M. M. Chem. Ber. 1933,66,1210-I2L3; 3 mg, 13 pmol, ca. 26 pM) produced an additional peak which eluted after Razuvaev,G. A.; Koton, M. M. Zh. Obshch. Khim. 1934,4,647452; CsDtz. Chem,Abstr.1935, 29, 3661. (36) Russeil,M. E.; Bernstein, R. B. J. Chem. Phys. 1959,30,607-512, (51) Sokolov,V. I.;Bashilov, V. V.;Reutov, O. A. J. Organomet.Chem. 613-€17. 1975, 97, 299-306. Sokolov, V. I.; Bashilov, V. V.; Reutov, O. A. J. (37) Weston, R. E.; Seltzer, S. J. Phys. Chem. 1962,66,2192-2200. Organomet.Chem.1976, 111, C13-C16. Bashilov, V. V.; Sokolov,V. I.; (38) Srinivasan,R. J. Chem. Phys. 1958,28, 895-898. Reutov, O. A.Izu. Ahad. Noufr SSSR, Ser.Khim. 1982(9), 2069-2089; (39) Ganesan,R.Z. Phys. Chem. (Munich) 1962,31,328-340. Chem. Abstr.1983,98, 53955v. (40) Laurie, C. M.; Long, L. H. Trans. Faraday Soc. 1955,51,665-672. (52) Rossel,O.; Sales,J.; Seco,M. J. Organomet. Chem. 1981,205, Long, L. H. Trqns. Faraday Soc. 1955, 51,673-679. 133-137. Rossel,O.; Sales,J.; Seco,M. J. Organomet.Chem.1982,236, (41) Long, L.H. J. Chem. Soc. 1956,3410-3416. Cattanach, J.;Long, 4t5-420. L. H. Trans. Faraday Soc. 1960, 56, 1286-1295. (53) Cross,R. J.; Wardle, R. J. Chem. Soc. A 1970,840-845. HornogeneousReactions of OrganoplatinumCompounds Organometallics,Vol. 4, No. 10, 1985 1825

Although the mechanistic bases for the products ob- the metals all affect the composition of the resulting served in eq 7 are obscure,the important conclusion from amalgam, PtHga is usually the final compositions. This the vantage of this paper is that the use of Hg(O) to sim- amalgam is stable to 400 oC, at which temperature de- plify the thermal decomposition of 13 is clearly unsuc- composition to PtHg2and PtHg is observed:mIn general, cessful: the reaction is rlore complex in the presence of we suspect that mercury will be most effective in poisoning Hg(0) rather than less complex. systemsin which amalgamation with mercury is possible. Thermal Decomposition of trans-Chloro(neo- Palladium, for instance, forms amalgams of various com- pentyl-d r, )bis (tri(cyclope ntyl- d 27'lphosphine)plati- positions Pd,Hg-; nickel forms NiHga and NiHgr.ut'6t num(Il) (LDrPtNpDCl, l4). The thermal decomposition Systems involving other transition metals which have not of cyclohexanesolutions of LDrPtNpDClssin the presence been observedto form amalgamswith mercury, but which of added mercury produces 1,1-dimethylcyclopropane-d1e, have a solubility in mercury (i.e.,cobalt, ruthenium, rho- neopentana-du, and neopentane-dr2in proportions nearly dium, and iridium)56 may be less amenable to this poi- identical with those observed from the thermal decom- soning scheme. In any event, mercury may adsorb on the position of 14 in the absenceof added mercury (Table II, surfaces of these metals, even if solution or amalgam eq 11, LD = (CsDe)sP). The rate of this thermal decom- formation is not possible. It is evident that the technique has limitations. In re- t-D ,coc(co) c-cD, L1 'Pr' 2 35 - 6 li ,D r Pt + c(co3)4 * (l) actions involving hydrogen abstraction from solvent during .Lo r33.c X-g,o O-gro cr, ci uo thermal decomposition of bis(cyclopentyldimethyl-

7il phosphine)dimethylplatinum(Il), it seemspossible that free methyl radicals derived from organomercury com- position seemsto be -\Vo faster in the presenceof added pounds may be involved. Both insertion of mercury into mercury, but this difference in rates is probably not ex- Pt{ bonds and transmetalation of mercury with platinum perimentally significant. Reaction solutions remain clear alkyls (driven by heat of amalgamation) seem possible and appear to be homogeneouswhether mercury is present reaction pathways. In other systems,in which the thermal or not, but reaction solutions decomposedin the presence decompositionsgenerating platinum(0) are rapid, it may of mercury become a pale salmon color as the reaction be impossible to suppress heterogeneousprocesses fol- proceeds. (Solutions decomposedin the absenceof added lowing from the initial nucleation of platinum colloids in mercury are colorless.) Reactions performed in the pres- solution. ence of added mercury give minor product(s) which in total This work leavesunresolved the question of the detailed constitute -\Vo (by ttP NMR integration) of the major mechanism of conversionof I to 2. We offer two specu- phosphorus-containing product LD2PtDCI. These products lative and schematicpathways for the reaction. One (eq are absent in the absenceof mercury. 12) involves an internal cyclometalation reaction. In this Mercury(0) apparently does not significantly change the courseof the reaction. Thus, the reaction which abstracts ,'T ------+i'1 .------*,,4------* p,ror * hydrogen (deuterium) from cyclohexane(-d,2) and forms t neopentane-dp appears to be a homogeneousreaction. r5 Becausethe reaction which leads to neopentaneis a rel- atively minor pathway for the decompositionof 14 (with equation, the other ligands present on platinum are not the major pathway being decomposition to dimethyl- indicated, but by analogy with previously studied sys- cyclopropane, probably via a 3,3-dimethyl- tems,le'62at least one vacant coordination site would platinacyclobutane intermediate), it is not practical to probably be required for this reaction to proceed. The investigate the mechanism of this reaction in detail. obvious difficulty with this schemeis the large strain which Whatever its mechanism, it probably does not reflect would appear to be involved in forming compound 15. The heterogeneoussolvent activation. conversionof 16 to 2 has excellent precedent in the thermal decomposition of 3,3-dimethylplatinacyclobutaneto di- methylcyclopropane.2l An alternative route Discussion for this transformation would involve a chain reaction with a Selective poisoning with mercury(0) is a useful but not platinum hydride of uncharacterized structure as the chain universally applicable technique for differentiating ho- carrier (eq 13).24 mogeneousand heterogeneousreactions involving tran- "Pt-H" Pr(o) "Pt- sition-metal organometalliccompounds. The technique, H H" when successful,probably involves formation of mercury + ------+ tt} -) + fl3) /--L /t amalgams. Mercury forms amalgamswith bulk platinum tr* Pr -l- N K I in a stepwise fashion giving PtHga, PtHgr, and PtHg H A species.s6-58Mercury may also adsorb on the surface ol I t7 2 platinum.se Similarly, mercury probably modifies the activity of colloidal platinum through amalgamation.28 Bxperimental Section Although the initial ratio of platinum to mercury, the General Data. All reactions requiring anhydrous or anaerobic experimental conditions,tr and the details of treatment of conditions were performed by using standard Schlenk or pressure

(5a) The s5mthesisof LDrPtNpDCl, and a detailed mechanistic study (60) Briinnland, R.; Leffler, J. A.; Frildt, L Suen. Kem. Tidskr. 1957, of its thermal decomposition, will be described elsewhere: Brainard, R. 68, 380-386; Chem. Abstr. 1957, 51, 979a. L.; Miller, T. M.; Whitesides,G. M., in progress. (61) Whitesides,G. M.;Gaasch,J. F.;Stedronsky,E. R. J. Am. Chem (55) Jangg, G.; Dortbundak, T. Z. Metallhd. 1973,64,7\5-7t9. Soc. 1972,94,5258-5270. McCarthy, T. J.; Nuzzo, R. G.; Whitesides, G. (56) Hassan,M.Z; Untereker,D. F.;Bruckenstein,S. J. Electroanal. M. J. Am. Chem.Soc. 1981,/03, 1676-1678.McCarthy, T. J.; Nuzzo,R. Chem. Interfacial Electrochem. 1973,42, 16l-18l. G.;Whitesides, G. M. J. Am. Chem. Soc. 1981,103, 3396-3403. Nuzzo, (57) Sutyagina, A. A.; Vovchenko, G. D. Surf . Tech.l981, .f3, 257-284. R. G.; McCarthy, T. J.; Whitesides,G. M. J. Am. Chem. Soc. 1981,I0J, (58) Affrossman, S.; Donnelly, T.; McGeachy, J. ..I.Cqtal. lgZJ,29, 3404-3410. Whitesides, G. M. Pure Appl. Chem. 1981,5J, ZB7-292. 346-351. (62) Fieser, L. F.;Fieser, M. "Reagentsfor Organic Synthesis";Wiley: (59) Barlow, M.; Planting, P. Z. Metallkd. 1969,60, 292-297. New York, 1967;Vol. I, p 1179. 1826 Organometallics, Vol. 4, No. 10, 1985 Whitesideset al. bottle techniques. Lab Glass septa for use in the pressure bottles 2,2'-azobis(2-methylpropionitrile)71 was connectedto the reaction were extracted with toluene until the washings were colorless and flask with Gooch tubing. The reaction flask was heated to 90-95 then washed with hexane, methanol, and finally water. The septa oC, and portions of the AIBN were added to the hot heptane over were dried by heating them to 110 "C in a slow stream of argon 15 h. The reaction mixture was filtered to remove a small amount and were stored at room temperature under argon before use. of insoluble polymeric material and allowed to cool. The product Olefins were removed from alkanes by stirring over concentrated crystallized from the reaction mixture and was separated by sulfuric acid for 48 h, followed by a washing with a saturated filtration. It was recrystallizedfrom 325 mL of ethanol to give sodium bicarbonate solution and water. Some batches were also 128g (0.94mol,72Vo) of the dinitrile 6 as colorlessprisms: mp oc lH treated with concentrated aqueouspotassium permanganate so- 157-160 (lit.70(165 "c); NMR (CDCU 6 1.53(s) (lit. 6 1.45 lution. All were filtered through alumina and/or distilled under (s));IR (KBr) 3034(s), 2980 (m) 2926(m), 2278(s),1482 (s), 1407 inert gas immediately before use. Diethyl ether and THF were (s),1206 (s), 1170 (s), 1133 (m),928 (m),355 cm-l. distilled from disodium benzophenoneunder argon. l-Methyl- 2,2,3,3-Tetramethylsuccinic Acid Anhydride (7).to The cyclopentene (Chem. Samples Co.) was distilled from calcium dinitrile 6 (128g,0.94 mol) wasslowly dissolved in a mixture of hydride (Alfa Products,40 mesh) under argon. Dihydrogen was 195 mL of water and 455 mL of concentrated H2SOa. The re- oC Matheson U.H.P. (99.9997o).Hexamethylphosphoric acid tri- sulting mixture was gradually heated to 100-110 at which point amide (HMPA) was distilled from CaH2under reduced pressure. a vigorousexothermic reaction occurred. The temperature of the oC. Dimethylacetamide (DMA) was distilled from CaH2 under reduced reaction mixture rose quickly to ca. 150 After the initial oC pressure and redistilled from molten sodium under reduced reaction had subsided,the mixture was heated at 130-150 for pressure. Tosyl chloride was recrystallized according to the an additional 2 h. The reaction mixture was allowed to cool to procedure of Fieser.63 Dichlorodiphenylstannane, dichlorodi- room temperature and was then extracted with three 400-mL methylstannane,dichlorodiphenylsilane, dichlorodimethylsilane, portions of ether. The combined organic extracts were washed and dichlorodiphenylgermanewere obtained from Alfa and used with 400 mL of water and then with 40 mL of saturated aqueous as received. Dichlorophenylphosphine was distilled prior to use NaHCO3. The organicphase was dried over anhydrousNa2SOa and stored in a flask equipped with a Teflon stopcock. Magnesium and filtered. The solvent was removed under reduced pressure. turnings were supplied by J. T. Baker. Triply distilled mercury Benzene(250 mL) was added to the residue and then removed was used in poisoning experiments without further purification. under reducedpressure. The residuewas dried in vacuoto give Melting points were determined in open capillary tubes on a t34 g (0.85mol, 9L%) of the anhydride 7 as a white solid: mp oC73);rH Thomas-Hoover apparatus and are uncorrected. Chemical shifts 151-153oC (subl) (lit. mp 147,70145-146,72 152,68140 are reported relative to internal MeaSi. Yields of Grignard NMR (CDCI3)6 1.27(s). reagents were determined by using the Eastham method.oa 22,3,3-Tetramethylbutane-1,4-diol. To a 1-L, three-necked cis-Dichloro(1,5-cyclooctadiene)platinum(II),s cis-dineopentyl- round-bottomed flask equipped with a condenser,addition funnel, bis(triethylphosphine)platinum(II),1eand (1,5-cyclooctadiene) - and magnetic stirring bar was added 600 mL of THF. Lithium dimethylplatinum(II)66 and (1,5-cyclooctadiene)dineopentyl- aluminum hydride (LAH)r4 (18.6 g, 0.48 mol) was added in platinum(Il) le were synthesizedaccording to literature procedtues. portions with stirring. The resulting mixture was heated at reflux. Unlabeled dimethylmercury was purchased from Strem and used A solutionof the anhl'dride(55.7 g, 0.36mol) in THF (180mL) without further pwification. Perdeuterated dimethylmercury was was added dropwise to the refluxing suspension.After the ad- prepared from perdeuterated methyllithium and mercuric chlo- dition was complete,the reaction mixture was heated at reflux ride.67 . Neopentane, 2,3-dimethylbutane (internal standard), for 8 h. The reaction flask was immersedin an ice bath, and ethyl methylcyclopentane, l-methylcyclopentene, and n-octane were acetate (-25 mL) was cautiousll' added to destroy the excess separated by using a2-ft Hi Plate 600 SE-30 on Chromosorb W LAH. Water (75 mL) was added to ensurethat all of the LAH column (Analabs) in a temperature-programmed run. Methane, had beendestroyed (Caution). The reactionmixture was decanted ethane, and ethylene were analyzed on a 2-ft 4To Apiezon on into 500 mL of 2 N H2SO4with cooling to maintain its temperature 'C. alumina column in a temperature-programmed run. Analysesof between 10-15 The organic phase was removed, and the 1-tert-butyl-1-methylcyclopropane and 2,2,3,3-tntramethylbutane aqueousphase was extracted with two 100-mL portions of ether. were canied out by using a20-ft 4% SE-30 column at 60 oC. The combined organic extracts were washed with 100 mL of water Characterization of products was based on co-injection with au- and two 100-mL portions of saturated aqueousNaHCO3. The thentic materials and on GC/MS analysis. organic phase was dried over anhydrous MgSOa,filtered, and 2,2,3,3-Tetramethylsuccinic Acid Dinitrile (6;.es-zoTo a evaporatedunder reduced pressure. Dissolution of the residue 1-L, three-necked round-bottomed flask equipped with a magnetic in benzenefollowed by removalof the solventin vacuo gave45.1 stirring bar, reflux condenser,and thermometer was added 230 g (0.31mol, 86%) of the diol as a colorlesssolid: mp 206-207"C oc;23tH mL of heptane. A second flask containing220 g (1.3 mol) of solid 0it. 209.5-211.5 NMR (CDCI3)6 0.86(s, 12 H), 3.40(s, 4 H),4.55(br s, 2 H); IR (CDCI3)3250 (s) (br), 2980 (s), 2880 (s), 1460(s), 1365 (m), 1270(w), 1145(w), 1050(s), 1010 (m), 805(w) (63) Watson, S. C.; Eastham, J. F. J. Organomet. Chem. 1967, 9, 165-168. cm-1. (64) McDermott, J. X.;White, J. F.; Whitesides, G. M. J. Am. Chem. 2,2,3,3-T etramethylb utane- 1,4-ditosylate. The d itosylate Soc. 1976, 98, 652t4528. was prepared as describedpreviously.22 (65) Clark, H. C.; Manzer, L. E. J. Organom.et.Chem. 1973, 59, 1,4-Dichloro-2,2,3,3-tetramethylbutane (8a). Procedure 4ll-428. Kistner, C.; Hutchinson, J.; Doyle, J.; Storlie, J. C. J. Inorg. A. The dichloride (24.2 mol, 93%) was prepared by the 1255-61. 9,0.13 Chem.1963,2, g (0.a6 with the ditosylate (66) Marvel, C. S.; Gould, V. L. J. Am. Chem. Soc. 1922,44,153-157. reaction of 19.5 mol) of lithium chloride Gilman, H.; Brown, R. E. J. Am. Chem.Soc. 1929,5I,928-930. Gilman, in HMPA as described.22 H.; Brown, R. E. J. Am. Chem. Soc. 1930, 52, 33L4-3317. Procedure B. A solutionof the ditosylate(45.4 g,0.1 mol) (67) Thiele, J.; Heuser, K. Justus Liebigs Ann. Chem. L896,290,l-43. in 400 mL of dimethylacetamidewas prepared in a L-L, three- (68) Bickel, A. F.;Waters, W.A. RecI.Trau. Chim. Pays-Bas1950,69, necked round-bottomed flask equipped with a reflux condenser, 3L2-20. Bickel, A. F.; Waters, W. A. Recl. Trau. Chim. Poys-Bos 1950, thermometer,and magneticstirring bar. Lithium chloride (12.6 69,1490-94. The authors also report that the thermolysis of dimethyl- 2,2'-azobis(2-methylpropionate) afforded dimethyl tetramethylsuccinate. g,0.3 mol) was added to the solution. The reactionmixture was (69) Scheffold, R.; Liiliger, J.; Blaser, H.; Geisser,P. Helu. Chirn. Acta heated at 110 "C for 48 h and allowed to cool to room temperature. 1975,58,49-84. (70) 2,2'- Azobis(2-methylpropionitrile) (AIBN) is toxic when ingested orally or inhaled as dust and should be handled with caution. In the (71) Hudson, B. E,: Hauser, C. R. J. Am. Chem. Soc. 1941, 63, organism it reportedly decomposesto hydrogen cyanide, which is found 3156-3162. in the liver, blood, and brain: Rusin, V. Ya. Tr. Nauchn Sess.Leningr. (72) Rathke, M. W.; Lindert, A. J. Ant. Chem. Soc. 1971, 93, Nauchn.-Issled. Inst. Gie. Tr. Profzabol. 1958, 247-251 (pub. 1959); 4605-4606. Chem. Abstr. 1962, 56,2682t. In addition, attempts to dissolveAIBN in (73) Reduction of tetramethylsuccinic anhydride with LAH in diethyl acetone or heptane have resulted in explosions on occasion: Carlisle, P. ether gave a mixture of the desired diol and the lactone,whereas reaction J. Chem. Eng. News 1949,27,150, and ref 70. Upon decomposition, with sodium borohydride in THF afforded the lactone as the major AIBN emits toxic fumes of NO, and HCI: "Dangerous Properties of product. Industrial Materials", 6th ed.; Sax, N. I., Ed.; Van Nostrand Reinhold: (74) Whitesides, G. M.; Gutowski, F. D. J. Org. Chem. 1976,41, New York, 1984;p 339. 2882-2885. Homogeneous Reactions of Organoplatinum Compounds Organometallics, Vol. 4, No. 10, 1985 L827

The reaction mixture was decanted into a mixture of 750 mL of The mixture stood at 0 oC for L2 h. Methanol (8 mL) was added, water and 200 mL of pentane. The organic phase was removed, and the solvents were slowly evaporated in a nitrogen stream at and the aqueouslphase was extracted with three 200-mL portions 0 oC. A light yellow, granular solid precipitated and was recovered of pentane. The combined organic extracts were washed in by decanting the supernatant and drying in vacuo. The crude successionwith three 100-mL portions of water and 100 mL of product (0.54g, 56Vo)was recrystallizedfrom ether (3 mL) and saturated aqueousNaCl. The organic phasewas dried over MgSOa methanol (2 mL) by preparing a saturated solution as above and and filtered. The solvent was removed in vacuo to give the chilling slowly to -78 oC: mp 90.5-95.5 oC; IR (KBr) 2990*2870 'C dichloride (bp 106-107 (I7 torr); 13.0g, 0.071mol). The lH (vs,br), 1452(m, sh), 1455(s), 1430 (br, sh), 1380(m, sh), 1375 NMR and IR spectra of the dichloride were indistinguishable from (m), 1363(m, sh), 1351(m, sp), 1251(br, w), 1240(sh, br, w), 1178 those obtained by procedure A.22 (w),1122(w), 1039(s), 1028 (sp, sh, m),765 (s,br),718 (s),630 rH l,4-Dibrom o-2p,3,}-tetrqm ethylb utane ( 8b ). The dibromide (br, m); NMR (CDCU 6 2.V1.4 (ca.16 H, br multiplet, {H2-), was prepared in 90% yield using the HMPA procedure described 1.4-{.85 (ca. 30 H, multiplet consistingof small s (1.4-1.0ppm) previously22except that LiBr was substituted for LiCL mp 5I-52 and larges (0.95ppm), {Hr);31P NMR 6 11.5(Un-p = 1798Hz). tH "C; NMR (CDCls)6 1.06(s, 12 H), 3.51(s,4 H); IR (14Br) Anal. Calcd for C26Ha6P2PtC,44.17; H, 8.53. Found: C, 46.11, 2970(s), 1470 (s), 1438(s), 1395(s), 1380(s), 1250(s),1125 (s), H, 8.76. The unacceptable elemental analysis for this compound 988 (s), 664 (s), 520 (s) cm-l. appearsto result from the use of impure CODPT(C8H1j. Despite (2p,3,3-T etramethylbutane- 1,4-diyl)bis(magDesium brom- severalrecrystallizations, a sharp melting point and acceptable ide) (9b). A 300-mL pressure bottle equipped with a magnetic analysiswere not obtained. The recrystallized compound prepared stirring bar was chargedwith magnesium turnings (8.0 g,0.33 mol), from this same precursor complex showed an impurity in its lH capped, and flame-dried under a stream of argon. After the NMR (1.3ppm (br s, ca.4 H)). apparatus had cooled, THF (ca. 50 mL) was added, followed by Bis (tri(ethyl-d s)phosphine) -3,3,4,4-tetramethyl- 1,2-dibromoethane (1 mL). The mixture was stirred vigorously platinacyclopentane was prepared in analogousfashion in 62To for 40 min to erxlure complete reaction of the dibromoethane. The yield from a pure sampleof (1,5-COD)Pt(CH2CMe2)2.It exhibited solvent and suspendedmagnesium salts were removed via cannula, the following properties: mp 95.5-97.0oC; IR (KBr) 2955(s), 2920 and the activated magnesium turnings were washed with 3 x 25 (s),2860 (s), 2800 (m),2220 (s),2190 (w), 2140(w), 2120(w), 2080 mL of THF. Fresh THF (80 mL) was added, and a pressure- (m), 1125(m), 1060(s), 1045 (m), 1038(m),891 (m),878 (m),790 adjusted dropping funnel equipped with a Leur trip and needle (s),M2 (m),632(m),622 (m); lH NMR (CD2CI,6 1.45(4 H, 1:4:1 *t" uq",2Jpr-H was attached to the pressure bottle. The funnel was charged with of = 64 Hz, PICH2-), 0'80 (12 H, br s, CCH3). a solution of 1,4-dibromo-2,2,3,3-tetramethylbutane(9.2 g, 35 1,1,2,2-Tetramethylcyclobutane (TMC ). An impure sample mmol) in 20 mL of THF. The solution was added at 75 oC over of TMC was prepared for use as a GLC standard by stoichiometric a 2-h period to the stirred suspensionof magnesium. Heating coupling of (2,2,3,3-tetramethylbutane-1,4-diyl)bis(magnesium continued for an hour after addition was complete. The mixture chloride) using silver(I) according to published procedures.Ts was then allowed to cool to room temperature. Aliquots (1 mL) Silver triflate (2.80g, 10.9mmol) and diethyl ether (10 mL) were were removed by syringe and titrated against 0.14 M 2-butanol combined in a 250-mL,three-necked flask that was equipped with in xylene6awith2,2'-biquinoline as the indicator. The concen- an 80-mL addition funnel and a Teflon-coatedstirrer. The so- tration of the di-Grignard reagent was 0.15 M (43% yield, as- lution of [Me2CCH2MgCl]2in THF (0.13M, 70 mL,9.1 mmol) suming the concentration of the di-Grignard reagent to be half was added over 70 min to the vigorously stirred solution at -5 the concentration of titratable organomagnesium moieties). oC. Reaction appearedto occur instantaneously: black solids Use of diethyl ether as solvent, or more concentrated solutions were depositedimmediately after addition began. The mixture of the dibromide, gavesignificantly lower yields. At concentrations was allowed to warm to ambient temperature over a L2-h period. above 0.15 M in di-Grignard reagent, a precipitate formed upon The solids were removed by filtration and the supernatant was cooling of the reaction mixture. Although the diGrignard reagent washed with two 30-mL portions of water. The solution was dried could be prepared in slightly higher yield (as estimated by ti- over magnesium sulfate and was subjected to careful distillation tration) from the dibromide than from the dichloride, (2,2,3,3- through a 10-in. stainless-steelspinning band. Fractions boiling tetramethylene-1,4-diyl)bis(magnesiumchloride) (9a) gavecon- below 98 oC at 760 torr were discarded.and the concentrated sistently higher yields of metallacycles. solution was distilled through a short-path apparatus. The product ( 1,5-Cyclooctadiene)-3,3,4,4-tetramethylplatinacyclop- fraction (98-120 oC) was found to consist of 2,2,3,3-tetra- entane (10). To a three-necked500-mL flask equipped with a methylbutane (3) and L,1,2,2-tetramethylcyclobutane( 1 :4.5, re- 250-mL addition funnel, a glassstopper, a No-Air septum, and spectively;ca. 0.16 g (16%) of TMC by tH NMR and GLC). It a Teflon-coatedstirrer was added (1,5-COD)PtCI2(2.24 g,6.0 was not possibleto purify the product by distillation; hencean mmol) and diethyl ether (50 mL). The mixture was chilled to elemental analysis was not performed. The impure colorless wax -50 oC, and a solution of 2,2,3,3-tetramethylbutane-1,4-diyl)- exhibited the following properties: mp ca. 50'C; IR (CCl, 2960 bis(magnesium chloride) (0.070M, 100 mL, 7 mmol) was added (br, s),2860(s), 1465 (m), 1450(m), 1384(m), 1379(m), 1367(m), dropwise over 3 h. The mixture was allowed to warm slowly to 1177(m), tr46 (m); rH NMR (CD2CI,D 1.60 (4 H, s,-CH2-),0.9? ambient temperature. Reaction appeared to commence at -25 (12 H, s, -CH3), 0.86 (ca.4 H, s, 3 -CHs). to -15 oC. The mixture was stirred for 12 h at 20 oC, and the 1-tert-Butyl-l-methylcyclopropane (TBC). The three- resulting red-brown solution was subjected to forced column membered ring was prepared by a Simmons-Smith reaction.T6 chromatography over silica gel/charcoal at 0 oC, and the column The zinc-copper couple was prepared by a reported procedure.?7 was washed with diethyl ether (ca. 200 mL). The pale yellow The zine-coppercouple (19.0g,0.19 mol) was suspendedin diethyl solution was concentrated to an oil under reduced pressure on ether (100 mL) in a 500-mL, three-necked flask that was equipped a rotary evaporator; the oil solidified after evacuation to 0.05 torr with a 250-mL addition funnel, condenser,and mechanical stirrer. for 30 min. The produce was recrystallized from a solution of Neat methylene iodide (74.0g,0.28 mol) was added rapidly to methanol (4 mL) and ether (15 mL) by solvent evaporation in the refluxing slurry. A crystal of iodine was introduced, and a a nitrogen stream at 0 oC, yielding white flakes (1.54g,62%): slow, mildly exothermic reaction ensued. The mixture was re- mp 97.3-99.0oC;IR (CCl4)2998 (w), 2940 (s),2870 (s), 2795 (m), fluxed for 12 h, giving a gray-black supernatant and a reddish 1475(m), 1455(m), 1440(w), 1428(m), 1375(m), 1368(s), 1353 suspension.A solution of 2,3,3-trimethyl-1-butene(7.7 g,0.080 (s),1338 (m), 1310(m), 1235(m), 1130(m), 1123(s),991 (m), 970 mol) in ether (40 mL) was introduced in a dropwise fashion over (m),855 (m) cm-t; tH NMR (CDCI3)6 4.79(4 H, 1:4:1utn,2Jpr_H 40 min, and the resulting mixture was stirred at reflux for 11 h. = MHz,CH:{m,2.25 (8 H, CH2C:),I.92 (4 H, 1:4:1utn,2Jpr_H *t', = 92H2, PICH2-),0.90 (12 H, J = L Hz, CCH3). Anal. Calcd for C16H2sPt:C,46.23 H, 6.79. Found: C,46.32:H, 6.91. (75) Simmons,H. E.;Cairns,T. L.;Vladuchick,S. A.; Hoiness,C. In Bis (triethylphosphine)-3,3,4,4-tetramethyl- "Organic Reactions";Dauben, W.G.; Ed.; Wiley: New York, 1973;Vol. platinacyclopentane (1). A solutionof impure (1,5-COD)PI- 20, pp 1-131. (76) Lambert, B.; (CH2CMe2)2(0.693 g, 1.66mmol) in (15 J. Koenig, F. R.; Hammersma,J. W. J. Org. Chem. diethyl ether mI-) was 1971,36,294t-2947. oC. (0.50 chilled to 0 Neat triethylphosphine mL, 0.39g,3.32 (77) This catalyst was found to give reproducibly 3.5 x 10-5mol of mmol) was added by syringe,producing a pale yellow solution. Ptlg of catalyst. 1828 Organometallics, Vol. 4, No. 10, 1985 Whitesides et al.

in Figure 3 which permitted agitation of the Table III. Hydrogenolysis of apparatus illustrated 177 oC for 36 h. Analysis of the colorless cis -Dineopentylbis(triethylphosphine)platinum(II) (4) in reaction mixture at (2) as the the Presence of l-Methylcyclopentene (MCP)" solution by GC gave L-tert-butyl-1-methylcyclopropane only detectable product. 101 ftob€d(s{) for max of products from the thermolysis of hydrogenolysis of convn 7o Qualitative examination (as platinum/glass and PtO2) 4 of MCPb poison Pt cat.' present I in the presenceof added Pt(0) charcoal,glass, and water indicated that 2 and 3 were the major <1 none no 10.7 products in all instances. In mixtures containing suspendedPtO2, 14.0 100.0 none yes L0-30% yields of 2-t er t -butylbut- 2- ene and 2-t ert -butylbut- 1- ene 11.0 <1 Hg yes Yields of ethane and ethylene ranged from 0% 11.0 <1 CO yes were detected. (mol/mol presenceof Hg(O) to 5% in solutions con- Lt.2 <1 PPhs yes Pt) in the taining no additives to 36To (with glass)to 60% (with Pt/glass) oExperiments were carried out in n-octaneunder 2 atm of H2 at and to 87% (with charcoal). bDetermined HiPlate 600 SE-30 40 "C. by GLC analysison a2-ft Mercury Poisoning of the Hydrogenolysis of CODPT(C- ChromosorbW column (Analabs). heterogeneousplati- on "The H3)2. These experimentswere performed in 25-mL tubes capped num catalyst consistedof 10 mg of a 29To Pt on scintered glass with butyl rubber septa and crown caps. A 10 x 5 mm foot- catalyst. ball-shaped stirring bar, 15 mg of CODPT(CHg)2,and glass, mercury, or catalyst, as appropriate, were weighed into the tube were removed by filtration, and the supernatant was The solids which was then capped and taken through three cycles of evac- with two SGmL volumes of saturated ammonium chloride treated uation (steel spinning band column. and analyzed for CODPT(CHs)zby UV spectrophotometry (280 The fractions boiling at 103-104 oC at 760 torr were combined, nm (e 1700)). For the experiment shown in Figure 5, the catalyst grving a clear, colorless,low-melting wax (0.2 g, ca.3To): IR (CCl4) was activated with two conditioning runs involving the hydro- (s), 3010 (s), 2960 (br, s), 2870 (s), 1480 (s), 1460 (br, s), 1428 3080 genolysisof 15 mg of CODPT(CH3;2.Complex and mercury(0) (m), 1382(s), 1365 (s), 1188 (s), 1105 (s), 1092 (m), 1010(s), 930 were added to the reactor via syringe. In the reactor containing (m),855 (m); tH NMR (CD2CI2)d 0.98(3 H, s, CHr),0.93 (9 H, only CODPT(CH3)2,hydrogen gas,and solvent, a shiny platinum s, C(CHg)s),0.45 (2 H, multiplet, -CH2-),0.02 (2 H, multiplet, mirror formed over the surfaces (glassand stirring bar) exposed -CHr-). Anal. Calcd for CsH15:C, 85.63;H, 14.38. Found: C, to the solution. Growth of the mirror could be visually observed 85.48;H, 14.60. to commenceat the vapor/liquid/wall interface and grow from Preparation of Supported Platinum Catalysts.2s (1,5- there. After ca.50Vocompletion the mirror began to flake off Cyclooctadiene)dimethylplatinum(Il) (1 g) was reduced with from the surfaceof the tubes. In the presenceof 10 mg of porous dihydrogen (2 atm) onto controlled porosity glass (1.5 g) (Sigma glass (having a surface area of ca. 0.34 m2, about the same as the PG-120-200)from n-pentane solution at 25 The number of "C. surface area of reactor exposedto solution), a smokey black minor active platinum atoms per gram of catalyst was subsequently formed on the reactor walls exposedto solution, and the glass obtained by dioxygen-dihydrogen titration.T&ru A typical catalyst particles turned first brown and then black as platinum was loading was 29Eo platinum. deposited on them. The reactor containing CODPT(CH3)2and Concurrent Hydrogenolysis of 4 and Hydrogenation of mercury(0) was allowed to stir for 5 da1'sunder 2.3-atm pressure l-Methylcyclopentene. A 1.5 cm X 2.5 cm heavy-walledpressure of H2. The solution was clear; a mirror formed on the reactor walls. tube equipped with a Teflon coated, football-shaped, magnetic Mercury adhered strongly to this mirror. The CODPT(CH3)zwas stirring bar was charged with the 29Vo platinum on glass catalyst Iess than 50% consumed. In the reactors containing glass, (10 mg, 5.7 x 10-4mmol of platinum) and 4 (10 mg, 0.018mmol). platinum complex, and mercury, the glassparticles becameblack The pressure tube was capped, evacuated to 0.01 torr, charged from platinum deposition,but few, if any, of the particles were with dihydrogen (30 psig), and immersed in a thermostated absorbed into the mercury bead. The surface of the mercury bead ethylene glycol/water bath (lll,vlv) at 40 oC. A solution of becamedirty gray with black flecks. \\:hen mercury(O) was added n-octane (6 mL) which contained 2,&dimethylbutane (0.015 mmol) active Pt(0)/glass catalyst, most of the catalyst was absorbed (internal standard) and l-methylpentene (0.030mmol) was added, to an into the mercury bead. After lessthan 1 min of vigorous stirring and stirring was begun. Reaction samples (-5 pL) were taken very small amount (ca. 1 mg) of verl'fine catalyst particles periodically for GLC analysis. Reactions carried out in the only a 'loose'. Even after weeks of stirring, these particles presence of poisons followed the same procedure, except that small remained loose. The mercury(0) bead in this experiment was a quantities of the indicated materials were added: mercury(O)-a remained gray color. Although soft as a whole, the surfaceappeared small bead (-1 g,0.005 g atom) of triply distilled mercury was dull *crusty" and the bead held shapesand contours. added at the same time as the catalyst; carbon monoxide-the reaction tube was pressurized tn 7.5 psig before it was pressurized Thermal Decomposition of (1,5-Cyclooctadiene)dineo- with H2; triphenylphosphine-5 mg (0.018 mol) of triphenyl- pentylplatinum(Il) (12). A sampleof 10 mg of CODPT(CHz- phosphine was added at the same time as the platinum catalyst. C(CHJJ2 $2) (0.022mmol) and -400 mg of mercury in a 6-mm Observed first-order rate constants are summarized in Table III, o.d. Pyrex tube was dried under vacuum (<0.015 torr, 3 h) and together with qualitative observations concerning the poisoning then dissolvedin 0.30 mL of freshly distilled benzene-du.The of the heterogeneousreactions. solution was freeze-pump-thawed three times and sealedunder Thermolysis of Bis(triethylphosphine)-3,3,4,4-tetra- vacuum (0.015torr) so that the tube was not longer than 6 cm. methylplatinacyclopentane (l). An 11.5cm X 0.6 cm Vycor A similar tube was prepared which did not contain mercury. thermolysis tube charged with I (15 mg, 0.024 mmol) was evac- These tubes were attached to an off-axis stirrer (Figure 3) with oC uated to (0.02 torr for 24h. A small bead of mercury (-0.5 g) Teflon tape and stirred at ca. 15 rpm at 131 for 6 h. The tubes was subsequently added, and the tube was again evacuated for differed markedly in appearancedue to added mercury (see 2 h. After the addition of dry cyclohexane (100 pL) via bulb to Discussion). Tubes were cooledwith liquid nitrogen, opened,and bulb vacuum transfer. the tube was sealed and affixed to the analyzed by GC and GC/MS. 1,5-Cyclooctadiene,dineo- pentylmercury, and neopentane were identified by comparing the GC retention times of the products with those of authentic sam- (?8) For titration method see: Carballo, L.; Serrano, C.; Wolf, E. E.; ples. No 1,l-dimethylcyclopropanewas observedby GC. The Carberry, J. J. J. Catal.1978,52,507-514. relative amounts of neopentane-d6and neopentane-d1were de- (79) The mass spectra of the reference samples were redetermined termined by comparing the relative abundance of the mle 57 and from variations in the periodically to minimize errors which could arise 58 peaks of the product neopentane with the ml e peaks of au- performance of the GC/mass spectrometer. These variations were never 5814.7 %) and neopentane-d1 very large and would, at the most, only contribute l'27o efior to the thentic neopentane-do 67 1100%, determination of Vo neopentane-dl or To neopentane-d12. (57135%,58/100%).80'8r The relative concentrationsof 1,5- Organometallics, Vol. 4, No. 10, 1985 1829 H omoge ne ous Reac tions of Organo platinurn Cornpounds cyclooctadiene, neopentane, dineopentylmercury, and 12 from Table Iv. Mass spectral ofsMethane and Deuterated the thermal decomposition of 12 in the presenceof mercury were nD€ta lH determined by NMR integration to be 6 1.0,0.43,0.87,and mle 0.33, respectively, and in the absenceof mercury were determined l3 t4 l5 l6 t7 l8 t9 20 2l to be 0 1.0, 1.7, 0.0, and 0.05, respectively.s2 These relative concentrations were determined by comparing one or two well- cHou 1o.o 20.2 95.0 100.0 1.9 resolved proton resonancesfrom each compound: 1,5-cyclo- cHsD 5.5 9.5 22.5 82.r 100.0 r.4 62.0 100.0 octadiene(5.5? (s, CH),2.20 Ppm (s, CH2)), neopentane(0.90 ppm cH2D2" 6.0 10.0 31.0 44.3 39.8 100.0 1.0 (s)),and 12 (4.70(t, cH),831.33 (s, CHs)). cHDs 1.3 6.1 5.3 9.1 15.5 1.1 90.1 2.8 100.0 0.7 Cyclopentylmaguesium Bromide. A solution of the Grignard CDn 8.4 reagLnt in diethyl ether wffl prepared from cyclopentyl bromide oAnalyzedon a Hewlett-Packard5992A GC/y-q using a 6-ft (24.3g,1.00 mol)' -of. bobtained (50 mL, 69 g, 0.46mol) and magnesiumturnings column 4% Apiezon on 80/100 alumina. from Titration ofthe dark yellow supematant indicated a concentration MathesonGas Products. . From: "Index of Mass SpectralData"; of t.2 M (quantitative yield). Kuentzel.L. 8., Ed.; Americansociety for Testing and Materials: Cyclopentyldimethylphosphine. A three-necked500-mL Ann Arbor, MI, 1963. flask was equipped with pressure-adjusted dropping funnel, a argon inlet. stirring bar, and a Friedrich condensertopped by an The tube was reevacuated for at least an hour before C6Dt2 The reaction flask was charged with cyclopentylmagnesium (previously subjected to three freeze-pump-thary cycles) was bromide (1.2 M, 65 mL, 78 mmol); a solution of chlorodi- ,ou.rr.r- distilled into the tube. The contents of the tube were mL of diethyl methylphosphine (6.0 mL, 6.0 g, 63 mmol) in 75 frozen in liquid nitrogen and the tube was sealedunder vacuum, latter was added ether was plracedin the dropping funnel. The taking care to avoid a platinum mirror at the seal. -78 oC period. After 30 to the Grignard reagent at over a 1-h when dimethylmercury was involved in the thermolysis, the (colorless su- min at room temperature, the reaction mixture procedure was slightly modified. The argon-filled tube was re- h. The mixture pernatant, white precipitate) was refluxed for 2 moved from the manifold, charged with the necessaryPt(II) -78 oC quenched with 2 mL was cooled to before reaction was complex and/or Hg(0), and capped with a septum' The di- (50/50 v/v distilled of degassedaqueous ammonium chloride methylmercury (0.2 mL) was added by syringe, a1d th-econtents After the mixture water/saturated aqueousammonium chloride). were quickly reattached to the vacuum line. The solution was was filtered had warmed to room temperature, the supernatant allowei to thaw; it was subjected to one freezrpump-thaw cycle flask' The through a medium-porosity frit into an argon-flushed before solvent was distilled into the tube. It was sealed under with ether; whitelesidue in the reaction flask was washed twice vacuum, as describedabove. the rinses were filtered and combined with the supernatant. The tubes were heated in the vapors of refluxing o-dichloro- 6-in. Vigreaux 3tP Solvent was removed by distillation through a benzene(181 'C) for 4 h. Analysis by NMR indicated that The Vigreaux column column topped by a short-path still head. the only phosphorus-containingspecies present was the starting product as a was removed prior to fractional distillation of the material. Prior to gas analysis, the tubes were attached to one oC. yield 4.ffi NMR colorlessoil, bp 160-162 The was e 65Vo):^1H end of a brass valve by means of a rubber O-ring and a swagelok (d, H); 31PNMR (CoDe)6 1.5(Lr m,9 H),0'86 Jp-n= 2.3IJ2,6 fitting. The other end of the valve was fitted with an adapter (ceDo) 6 40.3. which contained two rubber septa. The valve was evacuated -eir tii (cyclopentyldimethylphosphine)dimethvl- through the septa (via syringe needle),the tube was broken by with a platinum(Il) (13). A 25-mL recovery flask equipped closing the valve, and the vapor phasewas sample4 yltt gas-tight (1,5-cyclooctadiene)dimethyl- l stirring bar was charged with syringe. The gaseorlsproducts were analyzed by GC/\S and GC, capped with a platinum (0.8232g,2.47 mmol). The flask was using a 6-ft 4% apiezon on 80/100 alumina column. Results are platinum septum and evacuated-back-filled with argon twice. The presented in Table I. To this compound was dissolved in 10 mL of diethyl ether. The isotopic composition of the methane produced was de- (1'0 coloiless solution was added cyclopentyldimethylphosphine termined by iomparing the mass spectral data of the product with The colorlesssolution mL, 0.955g,7.34 mmol) in 5 mL of ether. data obtained from authentic samplesof CHa, CHBD,CHD3, and by rotary evaporation; was stirred overnight. Solvent was removed cDa and the literature values for cH2D2.e Unlabeled methane pentane at -80 the residual white paste was recrystallized from was'obtained from Matheson Gas Products; methane-d1was oC yield of and dried over potassium hydroxide in vacuo. The prepared by quenching an aliquot of CH3MgBr, which had been oC, rH (CoDo)6 2.05 colorlessplates, mp 64-65 was 45Vo: NMR lr""ru-p..-p-thawed twice, with DrO. Methane-d3 and meth- (s, (m, 30 H);31PNMR (CeDo) (s,br, 2 H), 1.68 br,4 H), 1.55-{.95 ane-dnwereobtained similarly by quenching aliquots of CD3MgBr = Hz)' Anal' Calcd for 6 8.9 (s with Pt satellites, Jpt-p 1800 with H2O and D2O, respectively. Results of the GC/MS analyses Cr6Hs6P2Pt:C, 39.58;H,7.47 P, 12.76.Found: C, 39.86;H,7.47; pt.i".tted in Table IV. The relative abundance of the mf e P, 13.03. "."16, 1?, 18, 19, and 20 peaks of the methane product were then Thermal Decomposition of Bis(cyclopentyldimethyl- compared with those of CHa, CH3D, CH2D2,CHD3, and CDn. In Thermolyses were con- phosphine)dimethylplatinum(II). no casewas CH2D2observed. flame-dried under ducted in 5-mm o.d. NMR tubes that had been The ethane produced was either completely unlabelled or The vacuum and then allowed to cool in dry, oxygen-freeargon. perdeuterated, as determined by comparison with literature values with tubes were removed from the vacuum manifold and charged and C2D6.& (Mass spectral data for C2H5Detc. are not (ca. for C2H6 cis-bis(cyclopentyldimethylphosphine)dimethylplatinum(II) available, but in all casesthe mf e values obtained experimentally (ca. g) if necessary. 20 mg,41 pmol). Mercury(0) 0.3 was added were in excellent agreement with those of C2H6or C2D6.) The isotopic composition of the ethylene produced was determined of the mf e 28,29,30, 31, and 32 peaks with (80) Neopentane-d1 was produced using two different methods. (i) by comparison C2H3D, Addition ofOct in D2O (99 atom Vo\ to a solution of cis-dineopentyl- lilerature values of the corresponding peaks of C2Ha, le bis(triethylphosphine)platinum(Il) in dry methylene chloride produced C2H2D2(cis and trans), C2HD3,and C2Dn.e neopentane-dt. (ii) A sample of neopentytmagnesir4n bromide which had Thermal Decomposition of ttans-Chloro(neopentyl- quenched with D2O to produce neo- been dried under vacuum was d (tri(cyclopentyl- d zz) platin um (II ) produced samples of neopentane-dt, which 11)bis Phosphine) pentane-d1. Both methods (LDrPtNpDCl, 14). To 14 mg of LD2PINpDCI(0.017 mmol) and spectra. -gave similar mass NMR tube (81) The relativoconcentrations of the products of decomposition of 740-mgof mercury dried under vacuum in a 5-mm 12 in the presenceof mercury are probably accuratetn t'vo, while in the on a vacuum line was added 0.5 of cyclohexane-dsby vacuum absenceof mercury are probably only accurate to *20To. The hetero- transfer. The solution was freeze-pump-thawed three times and geneousnature of the solutions produced by thermolysis in the absence of added mercury probably caused this discrepancy in error limits. 1/r) (82) Platinum-is composedof one-third lebPt(spin and two-thirds neopentane-d12were other isotopes (spin 0). A characteristic coupling pattern is the 1:4:1 (84) Authentic samplesof neopentane-drrand platinum 'triplet". orenared by quenching samples of neopentylmagnesiumbromide-dt' - *Index under vacuum with (83) Juentiel, L. E., Ed. of Mass Spectral Data"; American tu."d to make-LDrPtNpDCl)which had been dried Society for Testing and Materials: Ann Arbor, MI, 1963. H2O and D2O, respectively. 1830 sealed under vacuum (<0.015 torr). The other three reaction authentic samplesof neopentane-d11rc\1n0%,66139%) and solutions shown in Table II were prepared in a similar fashion. neopentane-dp (65I Ll.4, 66/ 1007o).ao'ss Solutions were decomposed thermally in a constant temperature oil bath at 133 + 0.5 oC with no stirring. The extent of decom- Supplementary Material Available: Synthesisand char- position was determined by comparing the integrals of the 31P acterization of triethylphosphine-d15and several 3,3,4,4-tetra- NMR resonances of LDTPIDCI and starting material. The kinetics methylmetallacyclopentanes (bis(tri-n-butylphosphine)-3,3,4,4- at decompooition were half-order in LDrPtNpDCl.ss Solutions were tetramethylplatinacyclopentane; bis(triethylphosphine) -3,3,4,4- thermally decomposed for 15 h or until -25Vo of LD2PINpDCI tetramethylplatinacyclopentane; 1-phenyl-3,3,4,4-tetr amethyl- remained. After the volatiles were condensed by cooling with phospholane oxide; 1-phenyl-3,3,4,4-tetr amethylphospholane; liquid nitrogen, tubes were opened and sealedwith a septum. The 1,1 -d iphen yl-3,3,4, -tetramethylsilacyclopentane ; I,1,3,3,4,4- liquid phase of each tube was analyzed by GC and GC/MS. The hexamethylsilacyclopentane;1, 1-diphenyl-3,3,4,4-tetramethyl- relative amounts of neopentane-d11and neopentane-d12were stannacyclopentane; and 1,1-diphenyl-3,3,4,4-tetramethyl- determined by comparing the relative abundance of the mle 65 germacyclopentane)(11 pages). Ordering information is given and 66 peaks of the product neopentane with the rnle peaks of on any current masthead page.