alkyl aluminum der 00 camp

8/12/2019 Alkyl Aluminum Der 00 Camp

1/100

ALKYLALUMINUM DERIVATIVES AS OXOPHILES IN ORGANICSYNTHESIS: STRUCTURE AND REACTIVITY STUDIES

BYCURTIS R. CAMPBELL

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THEUNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA1991


2/100

To my parents, with thanksfor their never endingsupport and encouragement


3/100

ACKNOWLEDGMENTS

The author would like to express his appreciationto Professor Merle Battiste for the guidance providedin this work, the late night philosophicaldiscussions, and the occasional attitude adjustment.My gratitude is also expressed to the other facultymembers that made life a little more interesting andlively during my stay at UF. Special thanks go toProfessor Jim Deyrup who taught me many things in andout of the classroom. A debt of gratitude is owed toDr. Radi Awartani for those special lab skills thathe has passed on to me. My appreciation is alsoextended to Dr. Dave Powell and his staff for theirassistance and special handling of those dissertationmaking, gotta have 'em yesterday mass spectra .Merle's Perles, both past and present, deserve aunigue thanks for the unigue outings and labconversations embarked upon. Finally, greatest thanksand best wishes go to Kristen for just being there andputting up with me.

111


4/100

TABLE OF CONTENTSPAGEACKNOWLEDGMENTS 11*

ABSTRACTCHAPTER I INTRODUCTION CHAPTER II VINYL OXIRANE SYNTHESIS VIA THEHORNER-WITTIG REACTION 14CHAPTER III STRUCTURE AND REACTIVITY OFALUMINUM ENOLATES 33CHAPTER IV ALKENE FORMATION VIA TRIMETHYL-ALUMINUM ACTION OF KETONES 49CHAPTER V SUMMARY 59CHAPTER VI EXPERIMENTAL 61General Experimental .* .' 61Apparatus and Technique 62Reagents and Solvents .' .* 62BIBLIOGRAPHY 88BIOGRAPHICAL SKETCH 91

iv


5/100

Abstract of Dissertation Presented to the GraduateSchool of the University of Florida in PartialFulfillment of the Requirements for the Degree ofDoctor of PhilosophyALKYLALUMINUM DERIVATIVES AS OXOPHILES IN ORGANICSYNTHESIS: STRUCTURE AND REACTIVITY STUDIES

BYCurtis R. Campbell

December 1991Chairman: Merle A. BattisteMajor Department: Chemistry

A new synthetic approach for the more volatilealpha-vinyl oxiranes required for subsequent studieswith organoaluminum reagents has been developed. Thisthe Horner-Wittig reaction of diphenylphosphinoyl-methyl lithium with a, p-unsaturated ketones, e.g.2-cyclohexen-l-one, is the key step in the sequenceproducing stable crystalline intermediates, incontrast to previous schemes in which volatile liquidsare realized at each stage. These solid intermediatescan then be stereoselectively epoxidized with MCPBA togive a penultimate crystalline precursor to thedesired vinyl oxirane. Treatment of this 2,3-oxidoHorner-Wittig intermediate under basic conditionsresults in the loss of the diphenylphosphinoyl groupas the water soluble phosphinous acid, thussimplifying final workup and isolation of the volatile


6/100

vinyl oxiranes. The synthetic sequence as presentedproceeds with good to excellent yields.

The reaction of diethyl-carbo-tert-butoxymethyl-alane (Rathke alane, RkeAl) with vinyl oxiranes hasbeen studied in order to determine the necessarystoichiometry for optimum yields of the producthydroxy ester. Several NMR studies were carried outin an attempt to elucidate the structure of the RkeAlin tetrahydrofuran (THF) solution. A mechanism forthe reaction of RkeAl with vinyl oxiranes is proposedbased on the reactivity and structural investigations.Reactions of the RkeAl with aldehydes and ketones werealso demonstrated to proceed with good yields.Structural variations of the ester moiety of the alanewere investigated in order to expand the reagent'sapplicability to organic synthesis. Preliminarystudies on chiral induction via the RkeAl typereaction are described.

Formation of alkenes through the thermal reactionof trimethylaluminum with ketones, an overlookedreaction, has been demonstrated. The possiblesynthetic utility of this reaction may be seen in theformation of 2-methylcamphene in one step from camphorin high yield. Reaction conditions are described foralkene isolation for several ketones. A mechanism forthe formation of these alkenes is offered.

VI


7/100

CHAPTER I

INTRODUCTIONThe field of organometallic chemistry has

undergone a tremendous surge in popularity in researchover the last guarter century. One of the main goalsof this research has been the development andapplication of organometals as synthetic reagents forselective organic transformations. In particular,investigations into the application and scope oforganoaluminum compounds has garnered significantinterest due to their application in natural productsynthesis and their somewhat unigue properties. 1 Asearly as 1955 aluminum alkyl species were noted toexhibit unusual behavior. 2 Aluminum alkyIs, R 3A1,react with carbonyl compounds in an analogous fashionto that of Grignard reagents; however, only a singlealuminum-carbon bond reacts in an additive fashion,leaving the remaining two alkyl substituentsdeactivated. This deactivation is associated withbond formation in the product between aluminum and theoxygen atom of the former carbonyl group, but theprecise reasons for the greatly reduced reactivity of


8/100

the remaining aluminum bound alkyls is still open tospeculation.

In general, organometallic compounds may act aseither a Lewis acid or as a nucleophile in theirreactions with carbonyl compounds and oxiranes. Froman antithetic standpoint oxiranes are seen to havegreater synthetic potential than carbonyl compoundsowing to their greater flexibility as startingmaterials for advanced syntheses since twocarboncenters are available for substitution, ratherthan one. In addition two different pathways areavailable for the reaction of saturated oxiranes withorganometals: 1) direct addition resulting in twopossible isomeric alcohols; 2) rearrangement to eitherthe aldehyde or ketone and subseguent addition to theresulting carbonyl group. The pathway taken can bedependent on the metal atom involved and thestructural (steric) environment of the oxirane sites.For example, nucleophilic ring opening at the leasthindered site of an epoxide is generally the mainreaction seen with alkyl cuprates, 3 whereastrialkylaluminums yield predominantly the additionproduct resulting from ring opening and addition tothe more substituted site. 4 Also seen in some ofthese reactions are products resulting fromrearrangement and subsequent addition. Reactionsinvolving trialkylaluminum reagents appear to be


9/100

sensitive to the choice of reaction solvent andconditions as demonstrated by the reaction of styreneoxide with trimethylaluminum as shown in Figure 1-1. 5

(CH 3 ) 3A135C -OHPh

(C 2H5 ) 2 CH 353%

(CH 3 ) 3 A135CC 6H14 ph

CH 3OH

86%

Figure 1-1

An additional illustration of the condition dependentreactions of trimethylaluminum is presented later inthis work (Chapter IV)

A common nucleophilic addition to epoxides isthrough the malonic ester synthesis. The resultingproduct, a hydroxy ester, can be further manipulatedenroute to a desired target molecule. A major problemassociated with this approach is the rather harshconditions necessary for the reaction to proceed, e.g.refluxing, alkaline ethanol solution. In order toalleviate this situation, a concentrated research


10/100

effort has been dedicated to the exploitation of metalmediated enolate/anion transfer reagents. Workthroughout the field has been focused in the two mainareas of reactivity and selectivity. Prostaglandinsynthesis is itself a very active area of chemicalresearch. It follows that a breakthrough in metalmediated anion addition could likely evolve from thisfield. Fried demonstrated a useful synthetic approach

Et 2AlC:=CR

R - -C 6H13 ; -CHC 5H11OCH 2Ph

Figure 1-2

shown in Figure 1-2 involving the opening of cyclicoxiranes with various alanes while working in the areaof prostaglandin synthesis. 6 These reagents wereprepared in toluene by addition of diethylchloroalane(Et 2AlCl) to various lithium acetylides to give thealkynylaluminums 2. Yields of these reactions rangedfrom 30-80% of the anticipated trans-cycloalkanols. 6

Somewhat later (1976) Danishefsky demonstratedthe first examples of the reaction of an aluminumester enolate with oxiranes. 7 This acetate eguivalentwas prepared in an analogous way to that of Fried'


11/100

aIkynylaluminums. Lithio tert-butylacetate, prepared

OBuI

Li C=CH 8 % ,o-\.^^fkA OHEt 2AlCl Et 2Al[CH 2 C0 2 Bu t ]

6

68 %R.T.6 hr

Figure 1-3

as described by Rathke, 8 on treatment with Et 2AlClafforded the diethylcarbo-tert-butoxymethyl alane(Rathke alane or RkeAl) reagent as a toluene solution.The resulting aluminum enolate gave a 34% yield of thetrans-hydroxy ester 5 when reacted withepoxycyclohexane (Figure 1-3) ; however, whenepoxycyclohexane was reacted with the Rathke lithiumsalt in the absence of the Et 2AlCl, a yield of only 8%of 4 is realized. Through manipulation of thereaction temperature and time the yield of thealuminum enolate reaction was ultimately increased to


12/100

68%. 7 However, after a failed attempt at using thismethodology on an oxirane in a steroidal ring system,Danishefsky discontinued investigations in this area.

This research group became involved in the areaof organoaluminum chemistry in 1982 when Dr. MeleanVisnick utilized the Rathke alane in the synthesis of() -anastrephin. 9 Visnick realized only one regio-and stereoisomer of 8 in a 24 % yield from thereaction of the Rathke alane with the vinyl epoxideshown in Figure 1-4. This reaction was carried out intoluene as were all previous examples involving theacetate eguivalent. Visnick increased the yield ofthe reaction to 87 % by a solvent change totetrahydrofuran (THF) after conducting a solvent studyto investigate the reactivity of the reagent. 9

+ RkeAl6

SeveralSteps

() -anastrephin

Figure 1-4


13/100

A previous investigation into the reactivity ofan alkynylalane with 3 , 4-epoxycyclopentene showed thatthe solvent exerted a marked effect on the resultingproduct distribution (Figure 1-5). 10 A non-polar

+ 2

R = -Bu

THFc rz cr oh oh

,..i0H y \w*C=CRPhCH.*\\ r?\T*\

10 11 jo C zz. CR

PhCH,

13 14Figure 1-5

reaction medium (toluene) resulted in rearrangement ofthe epoxide to the enone 13 followed by the additionof the alkyne to give alcohol 14. Changing thesolvent to a 1:1 mixture of toluene and THF eliminatedproducts resulting from the rearrangement pathway.The conclusion arrived at is based on the oxophilicityof the aluminum atom. A polar solvent such as THFwould satisfy the aluminum atom's oxophilicity andresult in better solvation of the reagent as well aspromote dissociation of the aluminum dimer (Figure1-6) . The lack of such a polar nucleophilic medium


14/100

THF(R3A1) 2 ^ * 2 R 3A1 THF

15 16

Figure 1-6

requires the reagent to become associated as dimersand larger aggregates thus reducing its nucleophiliccharacter while not seriously affecting its catalyticactivity for epoxide rearrangement. This isconsistent with previous reports and applications oforganoaluminums in hydrocarbon solvents such astoluene. While Visnick's solvent study on the Rathkealane echoed these results, regardless of the reactionsolvent, nucleophilic attack occurred exclusively atthe allylic position with no evidence of any productsresulting from rearrangement of the oxirane 9 .

Prompted by the favorable results obtained byVisnick, further studies were deemed necessary toinvestigate the scope of the Rathke alane in the ringopening reactions of a, p-unsaturated epoxides. Dr.Mapi Cuevas carried out a study designed to probe thescope, regiospecif icity, and synthetic application ofthe Rathke alane. 11

Cuevas' results on the applicability of theRathke alane to cyclic and acyclic a, p-unsaturatedoxiranes are summarized in Table 1-1 and 1-2,respectively.w As seen in Table 1-1 the 5- and


15/100

Table 1-1 Reactions of Vinyl Oxiranes with RkeAl

Oxirane

19


16/100

10

Table 1-2 RkeAl Reactions on Acyclic Vinyl Oxiranes

Oxirane Products Yield %

^X730

CC^Bu'22

3:1


17/100

11

reagent. The question of the location of the metal,whether bonded to oxygen or carbon, was investigatedalong with the mechanism of reaction. Results ofthese inquiries are detailed herein and by Cuevas. 11

In addition to the above studies, Cuevas furtherdemonstrated the application of organoaluminum speciesto advanced synthetic techniques. Two formalsyntheses of interesting molecules were demonstrated,that of cis-jasmone, 17., a* an advancedintermediate, 18, (Figure 1-7) that has been

cis-Jasmone, 17

CHO

11-Deoxy-prostaglandin PG series

Figure 1-7

elaborated by Corey to prostaglandins ll-deoxy-PGE 2and ll-deoxy-PGF 2 . 1 ' 12


18/100

12

As demonstrated above by Visnick and Cuevas thea, p-unsaturated epoxide is a very powerful startingmaterial for the construction of a variety of naturalproducts. The current preparation of these buildingblocks involves the alkaline epoxidation of cyclica, p-unsaturated ketones followed by a Wittigolefination. 13 The preparation suffers from problemsassociated with the isolation of the final volatileproduct from the Wittig olefination reaction mixturein the final step of the seguence. Therefore, a moreefficient synthesis was desired for preparation ofthis series of oxiranes. The first section of thisdissertation will deal with the development of a newsynthetic seguence for vinyl oxiranes and discuss itsadvantages over the previous preparations.

Promising more insight into the still young fieldof organoaluminum chemistry, additional studies wereundertaken to probe the scope of the reaction of vinyloxiranes with Rathke alane and other organoaluminumreagents with the view to provide insight into themechanism of such reactions. In addition to the workon unsaturated epoxides, the application of the Rathkealane and structural variations of it will bedemonstrated on several aldehydes and ketones. Asmentioned previously, a study of the reaction oftrimethyl aluminum with various ketones at elevatedtemperatures was undertaken as a natural outgrowth of


19/100

13

the above investigations. This brief study will lookat the reaction and its condition dependent productsin an attempt to demonstrate the utility of thereaction beyond current applications.


20/100

CHAPTER II

VINYL OXIRANE SYNTHESIS VIA THE HORNER-WITTIGOLEFINATION REACTION

As a result of an evolving study of the scope andmechanism of the highly selective ring openingreactions of vinyl oxiranes with organoaluminium

reagents, 9, 11,14

a need arose in our laboratoriesfor the development of a convenient and efficientsynthesis of the more volatile members of this seriesof epoxides including the cyclic exomethylenederivative 24.. The standard route to the methylene

39

'OH

H 2 2

Wittig

40 24

Figure 2-1

oxirane involves alkaline epoxidation of therespective 2-cycloalken-l-one followed by Wittig

14


21/100

15olefination as shown if Figure 2-1. 13 While theyields associated with the epoxidation step are in the80-90% range, the isolated yields from the olefinationstep were lower and often unacceptable (25-70%) . Themajor isolation problems stem from the surprisingvolatility and partial miscibility with water ofoxirane 24 .

In an attempt to alleviate the isolation problemsencountered in the current method we chose to explorean alternative pathway illustrated in Figure 2-2.

oHO /Sj|'P(Ph) 2

39LiCH 2 P (0)Ph 2 ^J^ MCPBA 24

41

Figure 2-242

This route was suggested by the earlier investigationsof Warren and coworkers on the utility of thediphenylphosphinoylethyl group in a two-stepolefination procedure. 15 The initially attractivefeature of this alternative scheme was the expectationof stable crystalline solids for intermediates 41 and42 . Subseguently, the methylene oxirane 24. could begenerated on demand from the stockpiled epoxide 42.using potassium hydride (KH) under non-aqueousconditions and workup. Ultimately, isolation would


22/100

16depend on the individual properties of the productepoxides: i) distillation directly from the reactionsolvent or ii) utilization of the oxirane as asolution of the reaction solvent. Although modest inscope as a pilot approach, we were also aware of thepotential synthetic bonuses of such a scheme. Forexample, the hydroxy 1 directed syn-epoxidation of theallylic alcohol 41 would, in a more substitutedsystem, lead to an alternative diastereoselectivitynot available through current means. Additionally,our method potentially offers a convenient three stepprotocol for the synthesis of chiral vinyl epoxidesvia enantioselective epoxidation of thediphenylphosphinoylmethylallylic alcohol (e.g. 41)

The first reaction attempted was the that of

0 c+ Li + CH 2P(0)Ph 2 THF

43 44

Figure 2-3

45

lithiomethyldiphenylphosphine oxide with cyclohexanonewhich resulted in the formation of a white crystallinecompound that was identified as l-(diphenyl-phosphinoyl)methylcyclohexanol 45 from its XH NMRspectrum (Figure 2-3). 16 Encouraged by this result,


23/100

17the synthesis of the vinyl oxiranes was undertaken.Treatment of 2-cyclohexen-l-one with the lithium saltgenerated from methyldiphenylphosphine oxide andn-butyl lithium in THF at 0 C resulted in exclusive1,2-addition to afford the crystalline alcohol 41 ingreater than 95% yield. Ohler and Zbiral havereported the synthesis of 41 under differentconditions in somewhat lower yields. 17 Epoxidation ofallylic alcohol 2 with m-chloroperbenzoic acid (MCPBA)in methylene chloride gave excellent yields (>90%) ofa crude oily epoxide mixture from which a single,crystalline epoxide was isolated as the predominantproduct on chromatography or trituration of the oilwith pentane (94.8% yield). We tentatively assignedthe ci s-structure to this product on mechanisticgrounds, 18 literature precedent, 19 and the results ofa Nuclear Overhauser Effect (NOE) differenceexperiment. A 14 % enhancement of the C-2 methineproton signal of 42. was noted upon irradiation of themethylene protons adjacent to the phosphinoyl group.By contrast the C-2 vinyl proton for the unsaturatedalcohol 41 showed a lesser enhancement of 10 %. Theminor product, presumably the trans-isomer of 42., wasnot obtained in sufficient quantities forcharacterization.

To verify the role of the hydroxyl group inMCPBA epoxidation of 41, we next examined theepoxidation of the methyl ether 46 prepared by


24/100

18O-methylation of 41 with potssium hydride (KH) andexcess methyl iodide (Mel) as shown in Figure 2-4.Treatment of 46 with MCPBA yielded an intractible

2. Mel

41 46

Figure 2-4

mixture of products which resisted crystallization orattempted purification. Proton and 13C-NMRexamination of the oil afforded evidence for twoepoxides in ca. 1:1 ratio. This result is consistent

with at least a significant fraction of hydroxyldirected ci s-epoxidation for alcohol 41 .

The epoxide 42. demonstrated an unexpected andsurprisingly facile ring opening of the oxirane bytraces of water on attempted recrystallization from anethyl acetate-hexane (75:25) mixture to form a singlecrystalline compound. NMR and mass spectralexamination confirmed the1- (diphenylphosphinoyl)methyl-l , 2 , 3-cyclohexane triolstructure 47 for the homogeneous white solid depositedduring the recrystallization attempt (Figure 2-5) . Ofthe four possible diastereomeric structural candidates


25/100

19

HO

42

EtOAc/Hex

Figure 2-5

IIPPh 2OH

47

for 47 (Figure 2-6), two, the cis, cis (48.) and trans,trans (51) hydroxyl configurations of C-2 and C-3, canbe eliminated on the basis of the proton coupling data

48 49

Figure 2-6

for the less shielded C-3 methine hydrogen (ddd, J =11.50, 8.78, 4.5 Hz) which revealed two diaxialcouplings. Assuming the bulky (Ph) 2 P(0) CH 2 - group islocked into the equatorial position, it is then clearthe C-3 hydroxyl must also be in the equatorialposition. Likewise, the C-2 hydroxyl must beequatorial in order for its methine proton (d, J =


26/100

208.77 Hz) to experience the observed diaxial coupling,thus ruling out structure 50. To confirm theassignment of structure 49 the triol was converted toits acetonide 52 (Figure 2-7) which, on NMR

o. *

49 +CH OCH, HOTsX C 6H 6

reflux

Figure 2-752

examination, revealed the identical downfield eightline multiplet (ddd, J = 11.64, 9.06, 4.0 Hz).

Final confirmation of the structures of 41 and 42.came in the form of single-crystal X-rayexamination. 20 The crystal structure obtained for 42afforded final proof of the cis relationship thatexists between the hydroxyl group and the oxiraneoxygen as supported by the above data and discussions.Both compounds 4_1 and 42. exist in the half chairconformation as expected; however, the bulkydiphenylphosphinoylmethyl group occupies thepseudo-axial position. Molecular models of 4.1 suggestthat in order to accomodate hydrogen bonding the bulkydiphenylphosphinoyl group assumes the pseudoaxialposition to relieve severe steric crowding with the


27/100

21cyclohexene ring protons. Additionally, both alcohols41 and 42. contain an intramolecular hydrogen bondbetween the hydroxyl and phosphine oxide groups, incontrast to published structures of similarcompounds. 21 Curiously, crystals of both 41 and 42occupy unit cells of the same dimension.

Elimination of the diphenylphosphinoyl group from42 was initially achieved by treatment of theepoxyalcohol with KH or sodium hydride (NaH) inN,N-dimethylformamide (DMF) at 60-65 C according toliterature precedent. 150 The desired methyleneepoxide was obtained as a solution (50:50) in DMF;however, attempts to separate the volatile oxiranefrom DMF by vacuum distillation were unsuccessful.Due to the nature of the subsequent reactions to becarried out on the epoxides, DMF contamination was notacceptable. Therefore, a change of reaction solventto THF was deemed advisable as subsequentorganoaluminum reactions are normally carried out inTHF. Treatment of 42. in THF under the aboveconditions afforded the epoxide as a solutioncontaining ~25% THF as determined by 1H-NMRintegration.

The synthesis of the methyl substituted oxirane56 was accomplished in a similar fashion to 24 exceptfor a slight modification in the first step (Figure2-8) . Addition of the diphenylphosphinoyl group to3-methylcyclohex-2-enone required lowering of the


28/100

22

53

44THF

54

54 MCPBACH 2 C1 2

55

55 KHTHF

Figure 2-8

56

reaction temperature to -3 0 C in order to avoidsevere side reactions that were encountered at areaction temperature of 0 C. Operation at the lowertemperature resulted in the isolation of a singlecrystalline solid, 54, in 97 % yield. Epoxidationwith MCPBA was carried out in the same manner as thatof 41 to form a single epoxide, 55, in good yields(>85 %) . In an analogous fashion to 42, epoxide 5_5readily opened to give the crystalline triol 57


29/100

23depicted in Figure 2-9. Single-crystal X-ray analysisshowed this compound to exist in the chair

55Silica Gel

Figure 2-9

57

P(Ph).

conformation with the bulky diphenyphosphinoylmethylgroup in the eguatorial position and hydrogen bondedintramolecularly to the C-l tertiary hydroxyl group.Additionally, intermolecular hydrogen bondinginvolving all three hydroxyl groups exists in anetwork throughout the crystal lattice. The X-raystructure also demonstrates the 1, 2-cis-3- transrelationship between the hydroxyl groups and themethyl group occupying an axial position. Finally,the methylene oxirane 56 was generated under the sameconditions as 24., but due to its lower volatility itwas isolated (65 % yield) as a pure compound by vacuumdistillation through a Vigreaux column.

Further attempts were made to demonstrate theusefulness of intermediates such as 41 to differentepoxidation technigues and possibly isolate the


30/100

24trans-isomer of 42. The first reagent studied wasmagnesium monoperoxyphthalate (MMPP, Figure 2-10)

.

Mg 2+ 6 H2

58

Figure 2-10

This reagent is marketed as a replacement for thecommercial grade of MCPBA (80-85 %) that is no longeravailable due to hazards associated with itspreparation. MCPBA is now available in only a 50-55 %grade from the chemical supply houses. MMPP is soldas its hydrated salt containing 80 % of the activeoxidant. Successful oxidations of several differentsubstrates and systems have been accomplished withMMPP and it was shown to be a suitable replacement forMCPBA 2 2 Oxidations carried out with MMPP arenormally run in halogenated or alcoholic solventsunder phase transfer catalysis (PTC) conditions.Reactions with MMPP were run on the methyl ether 46and 2-cyclohexenol. The reaction of 46 was carriedout in two-phase methylene chloride/water solution


31/100

25under (PTC) conditions with tert-butylammonium bromideat 0 C. Monitoring of the reaction by thin layerchromatography (TLC) revealed no product formationafter 20 hr of stirring so the reaction was warmedslowly to reflux. A product more polar than thestarting material was detected in a small amount byTLC. Preparatory TLC was used in an attepmt toisolate this new compound; however, liberation of thecompound from the silica media was not possible,indicating an extremely polar species or one that isdifficultly soluble such as the triol 49. If any ofthe epoxide did form it would follow that it furtherreacted to open the oxirane to the diol.

The reaction of MMPP with cyclohex-2-enol wascarried out at room temperature in is o-propanol underPTC conditions. Previous reports have demonstratedthe epoxidation of cyclohexene under similarconditions in 85 % yield after 7hr (Figure 2-11). 22

58, iPrOH/H 2

R.T. , 7 hr, 85 %

59 1Figure 2-11

This precedent would predict that the reaction of MMPPwith 2-cyclohexenol would afford the epoxy alcohol ingood yield. After stirring for 17 hr at room


32/100

26temperature resulted in no reaction, the solution waswarmed to reflux for an additional 2 hr with no effecton the progress of the reaction (Figure 2-12)Further literature search revealed no previousexamples, successful or otherwise, demonstrating the

58, iPrOH/H 2NO REACTION

60

Et.NBr, R.T.

Figure 2-12

utility of MMPP on allylic alcohols. One well knowntechnique for the epoxidation of allylic alcohols isthe Sharpless reaction.

The metal-catalyzed epoxidation bytert-butylhydroperoxide was first reported as a usefulsynthetic tool for the epoxidation of olefinicalcohols in 1973. 19 Since this time much work hasbeen dedicated to expanding the scope andenantioselectivity of the now so-called Sharplessepoxidation. Success of this epoxidation reactionwould then open up the possibility of achieving thestereoselective oxidation of chiral olefins throughthe use of tartrate esters in the reaction medium. 23

Figure 2-13 illustrates one example noted in theoriginal communication of the procedure by Sharpless


33/100

27which was the oxidation of 2-cyclohexenol bytert-butylhydroperoxide in refluxing benzene in thepresence of a vanadium catalyst (vanadyl acetylacetonate, VO(acac) 2 ) which resulted in very goodyields and excellent isomeric purity (98 % synaddition) . Unfortunately, our product epoxide 42

60

tBuOOHV0(acac) 2

C 6H 6reflux

OH

98 : 261 62

Figure 2-13

would not survive under such harsh conditions due toits facile ring opening to the triol 49. A laterreview by Sharpless and Verhoeven on these oxidationreactions noted that the heating of reactionscatalyzed by vanadium was unnecessary and, in fact,proceed readily at, or below, room temperature. 24Encouraged by this, a room temperature solution of 41in methylene chloride was subjected to a toluenesolution of tert-butylhydroperoxide in the presence ofVO(acac) 2 catalyst. The reaction was checked by TLCafter 5 hr to show product formation along withunreacted starting material. No notable change afteran additional 15 hr of stirring prompted the additionof a second eguivalent of the oxidizing agent.


34/100

28Continued stirring for a further 48 hr resulted in theconsumption of all starting olefin and a reduction inthe R f value of the product. Workup was carried outas described in the original literature and the

42

tBuOOHVO(acac) =-PhCH 3/CH 2 Cl 2

Figure 2-14

49

product was isolated as a yellow-tinted solid thatprecipitated upon washing with water. Spectralinvestigation by 1H NMR showed this compound to be thering opened epoxide 49 (Figure 2-14) . Apparently, 42is extremely sensitive to any sort of acid present inthe reaction solution or in the reaction workup. Thesuccess found in the MCPBA oxidation must becontingent on the buffering effect of the sodiumbicarbonate present in the reaction mixture.

A further example of the possible synthetic usesof compounds similar to 4_1 and 54 is demonstrated bytheir ability to undergo both 0- and C-alkylation.The O-alkylation was demostrated in the formation ofthe methyl ether 46 shown above in Figure 2-4. Theability to be alkylated on the carbon alpha to thephosphinoyl group was noted in an attempted


35/100

29preparation of 46. Upon generation of the anion of 41with BuLi at 0 C a deep red solution resulted when aslight excess of base was added. Quenching of thisreaction mixture with excess Mel resulted in theisolation of a mixture of the 0- and C-alkylatedproducts and recovered starting material. Subseguentgeneration of the dianion with BuLi and guenching withMel afforded two diastereomers (3:1 ratio) of theC-alkylation product in good yield (86 %) with noO-alkylation product observed (Figure 2-15) . The

2 eq BuLi41

Li

Z^ , Ph\Li Ph

Me

Mel'S IIPPh-64

63Figure 2-15

resultant C-alkylation product can be subjected toconditions necessary for elimination of thephosphorous group to yield the ethylidene sidechain.The yield and ratio of these E- and Z-alkenes could beof synthetic interest.

Warren has demonstrated the use of lithioethyl-diphenylphosphine oxide on benzaldehyde in theselective synthesis of Z-1-phenylpropene following the


36/100

30

H

Ph H

EtP(0)Ph 2 Ph^Me

THF

65HO 1 ^PhHerythro

66Figure 2-16

NaH

DMF Ph Me

67

elimination of the diphenylphosphinoyl group as shownin Figure 2-16. 15b The initial addition to thealdehyde results in the formation of the erythroisomer (IRS, 2SR) as the major product, 78 %, whichupon elimination of the diphenylphosphinoyl groupaffords the Z-alkene in 75 % yield. The E-alkene isaccessible via oxidation of the isomeric alcohols tothe ketone, followed by the sodium borohydride (NaBH 4 )reduction to primarily the threo isomer (89 % of themixture) , which gives the E-alkene upon elimination.An investigation was undertaken to obtain the threoisomer in a more direct route, thereby providing acomplimentary method to Warren's Z-alkene synthesis.

Benzaldehyde was subjected to lithiomethyl-diphenyphosphine oxide in THF at -78 C and allowed towarm to room temperature to yield the anticipatedaddition product (Figure 2-17). 25 Treatment of thesubstituted ethanol with 2.0 eg of BuLi resulted in


37/100

MeP(0)Ph265THF

HO

Ph

68

31

PPh,1. BuLi

2. Mel

Figure 2-17

the formation of the dianion which was quenched by theaddition of excess Mel. Spectral examination of thecrude reaction mixture revealed two diastereomerspresent in a ratio of approximately 5:1. Separationof these isomers by flash column chromatographyresulted in the isolation of the major component as apure compound that XH NMR confirmed to be the threoisomer as assigned by Warren. 15b The minor componentwas not obtained in sufficient quantities from thechromatography for accurate characterization; however,the peaks present in the 1H NMR of the mixturebelonging to the minor isomer are in good agreementwith those published for the erythro isomer. 15b Themethod described herein allows isolation of theintermediate alcohol responsible for the complimentaryalkene to that of the Warren method. This success hasspawned continued work in the area of alkylphosphinoyladdition to carbonyl compounds and subsequent


38/100

32

alkylation alpha to the phosphine oxide and itsapplication to natural product synthesis. 26


39/100

CHAPTER III

STRUCTURE AND REACTIVITY OFALUMINUM ENOLATES

The earliest examples of aluminum mediated anionsin oxirane ring opening reactions employed for thesynthesis of prostaglandins reported the use of thealane species in an 8.5 molar excess. 6 Danishefsky'sreport of the initial success of the Rathke alanenoted a 2.5 molar excess ot the organometal. 7Likewise, Visnick found the optimal stoichiometry tobe a 2.3 molar excess of the Rathke alane in reactionswith vinyl oxiranes. 9 The work described by Cuevaswith the Rathke alane on cyclic and acyclica, B-unsaturated epoxides was accomplished with aworking stoichiometry of a 2.5-3.0 molar alanylexcess. 11 With the general acceptance of a necessityof 2.0 equivalents of the organoaluminum species asimple mechanism of coordination to the oxirane by oneequivalent and subsequent delivery of the anion by asecond equivalent may be proposed as shown in Figure3-1.

33


40/100

34

7S V^\ .._ HO N re[CH 2 CO,R] AlEt, H C0 2R72Et 2AlCH 2 C0 2R71

Figure 3-1

In the case of the vinyl oxiranes this couldexplain the regiospecificity demonstrated in attack atthe allylic site due to stabilization of the chargebuild up upon weakening of the carbon-oxygen bond. Amechanism of this sort would allow for the reaction toproceed with a single equivalent of the enolate and acatalytic excess of the Lewis acid (Et 2AlCl) . Aseries of experiments were undertaken to test thisassumption and determine the necessary stoichiometryof the reaction. The three key reactions investigatedsubjected 3-methylene-l, 2-oxidocyclohexane, 24, tovarying amounts of the Rathke alane; the results ofthese reactions are collected in Table 3-1. Thestandard reaction was carried out with 3.0 equivalentsof the Rathke alane 6 to determine the yield underestablished conditions. Thereafter the remainingreactions utilized 1.5 equivalents of the alane 6. Afinal attempt to establish a cooperative competition


41/100

Table 3-1 Stoichiometric Studies on RkeAl

35

THF24 + 4 + RkeAl ~

6

Rxn 6 (eg) 4 (eg)

*^O0j* %

Yield (%)

3.0 75

1.5 48

1.5 1.5 3.4

reaction between the Rathke alane (6, 1.5 eq) and theRathke lithium enolate (4, 1.5 eq) was examined todetermine if the reaction is indeed catalyzed by thealuminum species. The results of these experiments(Table 3-1) clearly support the need for more than asinqle equivalent of the aluminum species in order forthe reaction to proceed with good yields and dismissthe assertion that the aluminum is only acting in acatalytic fashion. In view of the observedrequirement for at least two equivalents of the alanespecies one might envisage a dimeric species as thereactive intermediate.

In fact Fried 63 has proposed a mechanisminvolving an alane dimer derived 6-membered ring


42/100

36

transition state in the reaction of alkynylalanes withepoxides to form the trans substituted alcohols asshown in Figure 3-2. It would follow that the Rathke

OR


43/100

OBu74

Figure 3-3

37

carbon-oxygen bond thereby increasing the positivecharacter of the allylic carbon. This electrondeficient carbon is then attacked by the reverse endof the opened dimer 74 which then delivers the enolatemoiety. One may view this mechanism in colloquialterms as the ice-tong mechanism. In view of thepostulation of a dimeric structure for the RkeAl, aninvestigation into its structure in solution wasdeemed necessary.

A sample of the Rathke alane was prepared in theusual manner as described in the experimental section.Neat dimethylaluminum chloride rather than the hexanesolution of the Et 2AlCl was used in an attempt to aidein the simplification and interpretation of the NMRspectra. The reagent solution was warmed to 0 C andthe solvents removed in vacuo to give the aluminum


44/100

V

38

3CQO\ c

3pao

\& \ W4->Ww

IIP -I

/

o'' 3

3oaO


45/100

39

salt which was then dissolved in THF-d 8 and cooled to-78 C. The resulting spectra, recorded at varioustemperatures, were sufficiently complex to indicatethe presence of more than one species or perhapsunsymetrical dimers or oligomers. 27 The absence ofsignals in the vinylic region of the 1H NMR wouldsuggest the lack of a true enolate structure similarto that reported by Rathke for the lithium salt oftert-butyl acetate. 8 Spectral evidence obtained onthe lithium salt in benzene-d 6 included two doubletsat 3.14 and 3.44 ppm and the absence of a signal inthe IR spectrum between 1650 and 2 000 cm 1corresponding to a carbonyl stretch. Figure 3-5illustrates the enolate structure of 4. supportive ofthis evidence.

0 Li+AK2 OBu4

Figure 3-5

Researchers in the area of aluminum enolates havedepicted many different structures for these specieswithout citing experimental data or literatureprecedent. Japanese workers have used aluminum


46/100

40

enolates in aldol condensation reactions where theydepict an O-metallated species with no substantiatingevidence. 28 Mole and coworkers have reported the

Bu* 0AlMe 2 Bu c MeMe 3 Al \ / \ /> H He h 0AlMe 2 11 11Figure 3-6

isolation and characterization of the Z- (76) andE-enolates (77) shown in Figure 3-6 from the reactionof trimethylaluminum on mesityl oxide. 29 Furtherinvestigations revealed that the Z-enolate 76 existedin the dimeric form and the E-enolate 77 was made upof dimers and trimers. The lack of a monomer icspecies in this study encouraged us with respect toour proposal of the dimeric nature of the Rathkealane; however, the exact structure of the dimer wasstill in question.

A report on the X-ray diffraction study of theReformatsky reagent generated from tert-butylbromoacetate was published in 1983. The structurearrived at by the workers was that of a dimericspecies with each metal atom in the environment of twooxygens, a bromine, and a carbon atom (Figure 3-7). 30


47/100

41

THF

Br ZnOBu*

Zn Br

THF

All of the bonds in the non-planar species are oftypical single bond lengths. With this information onhand we have greater confidence in the suggestion ofthe dimeric species 74. for the Rathke alane and theassociated reaction mechanism shown in Figures 3-3 and3-4, respectively.

Side products observed in these aluminum enolatereactions include the self-condensation products ofthe ester enolate and chlorohydrin and glycolformation. In carrying out reactions with the Rathkealane it is imperative that all reagents are driedprior to use; otherwise, any water present in thesystem will quench the enolate and subject thestarting material to the unreacted Et 2AlCl resultingin the formation of the ring opened products as shownin Figure 3-8. Intentional treatment of the vinyl


48/100

24

II

AJ


49/100

43

43

43

+ RkeLi4

+ RkeAl6

81

81

C0 2Bu

Figure 3-9

cyclohexanone to 81 after 20 min at -60 C whereas 1.2equivalents of the Rathke alane afforded 76.7% of theadduct after 30 min under the same conditions (Figure3-9) . In this instance the aluminum may be acting asa catalyst and activating the carbonyl group foraddition through a 4 membered transition state, Figure3-10.

Several para-substituted benzaldehydes weresubjected to 1.2 equivalents of the Rathke alane inthe presence of 0.4 equivalent excess Et 2AlCl in THFat -60 C (Figure 3-11). The various substituentswere chosen to determine if the rate of the additionreaction was dependent on any electronic factors inthe molecule. All of the aldehydes studied resultedin similar yields of 60-70% with the reaction beingessentially complete in under 10 min.


50/100

44

,[Al(Et),CH ? CO.Bu t ]

OBu'Me Al R, R_ -2

82 ORI

m . - [Al(Et) 2 CH 2 C0 2Bu t ]

OBu

/ \83

Figure 3-10

^ x 2 A84

84

Explorations into variations of the enolate 6. toexpand its utility for the synthetic chemist was thenext logical step in these studies. The first changewas from an acetate to a propionate equivalent in theform of the tert-butylpropionate. The aluminumpropionate enolate (MeRkeAl, 95) was prepared in thesame fashion as the Rathke alane and used in the samestoichiometric proportions. The reaction of theMeRkeAl 95 with 3-methylene-l, 2-oxidocyclohexaneresulted in the expected addition products (Figure


51/100

45

HO H

RkeAl6

C02Bu

X = H;CH.

OCH 3 'CI;

NO.

8587899193

X = H;CH 3 '

OCH 3 CI;

NO.

8688909294

Figure 3-11

3-12). The goal of obtaining a diastereoselectiveaddition was not realized as a 68% yield of 96 wasobtained in a 1.2:1 ratio of isomers.

24 + Et2Al[CH3CH C02Bu ]-60 CTHF

9596

Figure 3-12

An interesting and quite possibly moresynthetically useful adaptation of the enolate may befound in the variation of the alcohol portion of thealanyl ester. Thus acetylation of a chiral alcohol orone having sufficient bulk to induce severe stericbias could result in the formation and isolation of


52/100

46

predominantly one stereoisomer (diastereomer orenantiomer) from the Rathke type reactions. The bulky

OH DMAPUH + (CH3 C0 2 )0 /^ ^^ OAcEt 3 N98

97 99

Figure 3-13

adamantyl group in 1-Adamantyl acetate, prepared from1-adamantanol and acetic anhydride as shown in Figure3-13 ,was examined as a steric biasing agent to helpinduce stereoselectivity during acetate delivery. Asecond consideration was the identification of theproducts by GC/MS through a parent mass peak that isnot present in the tert-butyl ester due to facile lossof the tert-buty group as isobutylene or as C 4H 9 0.The aluminum enolate (AdmAl, 100 ) was generatedanalogously to RkeAl and its reactions were run underthe standard Rathke alane conditions with variousfunctionalities (Table 3-2) . The yield of the 3- andT-hydroxy esters were good to modest with theexception of styrene oxide (3_6) . Unfortunately, uponanalysis of the reaction mixture by GC/MS, no parentmass peak was realized due to the facile loss of theadamantyl cation; therefore, without further


53/100

47

Table 3-2 Reactions of Adamantyl Acetate Alane with CarbonylCompounds and Oxiranes

X

+ AdmAl100

43

HO /^ CC- Adm

101

85

56

36

Ph H102

CO nAdm

C0 2AdmOH

CH103

Intractable Solid

purification the esters were hydrolyzed to theirrespective carboxylic acids for characterization.

The reaction of the AdmAl with styrene oxidefailed to give any addition products even after


54/100

48

extended reaction times and warming to roomtemperature. Reaction mixture analysis by capillaryGC revealed peaks attributable to the styrene oxide,1-adamantyl acetate, and 1-adamantanol. The presenceof the adamantanol indicates that some form ofreaction has taken place in order to liberate it fromthe ester. Continued monitoring of the reaction viaGC showed a steady decline in the styrene oxide signaland growth of the alcohol signal; however, no responsewas detected for any sort of addition product leadingus to believe that the aluminum species facilitatedpolymerization of the epoxide to a nonvolatilespecies. This was surprising owing to the fact thatstyrene oxide reacts with the Rathke alane in a 64 %yield to form the two isomeric addition products in a4:1 ratio favoring attack at the benzylic site. 11


55/100

CHAPTER IVALKENE FORMATION VIA TRIMETHYLALUMINUMACTION ON KETONES

The reactions involving alkylaluminum compoundswith carbonyl compounds are well documented, 31 ' 32and some of these reactions and their anomalies weredescribed in Chapter I. in particular, Mole hasdemonstrated the ability of trimethylaluminum (Me 3Al)AMe 3Al Me Me

2 eq .X^\104 105

Figure 4-1

to act as an exhaustive methylating agent in thepresence of tertiary or benzylic alcohols and anassortment of ketones (Figure 4-1). 33 ' 34 Prior tothis, no method existed for the direct exhaustivemethylation of carbonyl compounds. Yields of thegem-dimethylation product in the work reported by Molerange from 30% to complete conversion. The reactionswere carried out with a 2-3 mole excess of Me 3Al in a

49


56/100

500

sealed reaction vessel under various solvent,temperature and reaction time conditions. Athree-step pathway for this reaction has been proposedby Mole as shown in Figure 4-2. A persistent sideA Me 3Al M OAlMe.

104 106

/ \Me 3A1 u \,/106 k. Me

107 105 +

R R'107

.0Me 2Al AlMe 2

108Figure 4-2

reaction noted in many of the examples investigated byMole is alkene formation which is believed to resultfrom elimination after initial methyl addition. 34 Thecontribution of alkene formation to the productdistribution ranges from a trace amount to as much as50%, the major component. Herein we report conditionsthat allow the isolation of alkenes as thepredominant, if not sole, reaction product.


57/100

51

Entry into this area of research was gainedthrough an attempted methylation of (+) -d-camphor,109 . as illustrated in Figure 4-3, to obtain thetertiary alcohol 110 for use as a chiral auxilliary in

RM

109 R = Me 110M = metal

Figure 4-3

the Rathke alane investigations. Initial attempts atpeparation of the desired alcohol through the

traditional means of a Grignard (MeMgBr) reaction oraddition of methyl lithium gave less than satisfactoryyields of ~50% conversion to 110 . Starting ketone wasalways recovered despite the use of a large excess ofthe organometallic reagent. Treatment of camphor withexcess trimethylaluminum in refluxing hexane resultedin a -60% conversion to the carbinol 110 after 4hours. Still, considerable camphor was recovered fromthis reaction. Alternatively, a somewhat greateryield of ~70% was realized upon subjecting the camphorto excess Me 3Al in refluxing toluene for 5.5 hr. Themost profitable conditions discovered depend on


58/100

52

pretreatment of an ethereal solution of ketone 109with Me 3Al (2.0 eq, 1 hr) followed by MeMgBr (5.0 eq,3.5 days) which resulted in an -80% transformation to110. The question remained as to whether the twoorganometals react to form an alanate (R 4A1 ) as theactive methylating agent or if the Me 3Al acts in thecapacity of a Lewis acid and activates the carbonyltoward nucleophilic addition. A subsequent search ofthe literature uncovered similar results using LiAlMe 4in an ether solution as reported by Ashby's group in1974. 35 Our attempts to carry out the LiAlMe 4reaction in toluene resulted in no reaction even afterprolonged heating, further demonstrating the effect ofsolvation on organoaluminum reagents. When thereaction of Me 3Al in toluene shown in Figure 4-4 was

PhCH,109 + Me 3Al *-

Figure 4-4

allowed to continue for 2 4 hr a hydrocarbon product,2-methylcamphene, 111, was isolated in 81% yield.Noteworthy is the fact that the solvent had largelyevaporated and the pot temperature had increased


59/100

53

overnight. This result prompted the investigationsinto the possible synthetic utility of thistransformation as discussed herein.

A summary of the reactions carried out in thisstudy can be found in Table 4-1. The reactions are

TABLE 4-1 Reaction of Trimethylaluminumon Various Ketones

Ketone

113

115

118

PhPh

122

0026

Procucts in Decreasing Concentrati onCO 0>114

116

127

117

128 129

run with 4.0 eguivalents of Me 3Al in m-xylene at150-210 C for up to 30 hours in a simple refluxapparatus. At the end of the reaction time most of the


60/100

54

solvent had evaporated to leave a brown oil that waswashed with 10% hydrochloric acid (HC1) and extractedwith diethyl ether (Et 2 0) . The various productsrealized include tertiary alcohols, alkenes, andgem-dimethylated hydrocarbons. In accordance with theresults obtained, the ketones studied follow the samegeneral reaction pathway: methyl addition to form thetertiary alcohol followed by elimination to form thealkene product (s) . Exhaustive methylation productssimilar to those reported by Mole are also seen insome instances at higher reaction temperatures.

The most interesting and by far the mostsynthetically useful example demonstrated to date isthe reaction of Me 3Al and (+) -d-camphor as shown inFigure 4-4 to yield 2-methylcamphene. A small amountof the tertiary alcohol 110 is seen if the reaction isstopped prior to completion or not heated stronglyenough. No other products are seen by capillary GC inthe reaction mixture in concentrations greater thenone percent. Product determination was accomplishedby 1H and 13 C NMR and GC/MS. The one step isolationof 2-methycamphene is interesting because of thedifferent alcohols that can subsequently be achieved,by various forms of hydroxylation reactions, and theirpotential use in natural product synthesis.

The other ketones presented in this study formthe expected alkenes as the major product with the


61/100

55

only exception being a-tetralone. Table 4-1 shows thereactions and products obtained under the conditionsfound for optimum alkene formation. The reactionsreported in the table were run under two differentsets of conditions: 1) ~150 C for 30 hr and/or 2)~200 C for 15 hr. As demonstrated in the camphorcase, it appears that the reaction temperature is themajor factor in determining the product distributionof the reactions. For example, the reaction of2-indanone produced the methyl addition productexclusively at the lower temperature while at theelevated temperature only the one alkene product wasrealized.

The reaction of 1-indanone demonstrates theability of the resultant alkenes to rearrange underthe reaction conditions. This result is consistentwith the extreme reaction conditions present, hightemperature coupled with a strong Lewis acid. Thepresence of the isomeric alkenes was confirmed by 1HNMR and GC/MS. The diagnostic information, in thiscase, was gleaned from the vinylic region of the XHNMR which clearly depicts resonances for the twoisomeric alkenes.

A further example of this temperature effect maygive an insight to the route of the gem-dimethylationreaction of fluorenone reported by Mole. In our handsfluorenone gave as the major product


62/100

56

9-methylenef luorene at the milder conditions. Alsoseen in the reaction mixture were the alcohol anddimethylation products, 41% and 12% respectively. Itis believed that a longer reaction time would haveresulted in more alkene by elimination of the alcohol.At the same time more of the dimethylation productcould have resulted, this could possibly be alleviatedby a decrease in the reaction temperature. However,at the elevated temperature 9, 9-dimethylfluorene isthe dominant product with only trace amounts of thealkene and unreacted starting material present, thusdemonstrating the inability of the alcohol to surviveunder the reaction conditions.

This result prompts us to suggest that thedimethylation product seen in other studies resultsfrom the alkene as shown in Figure 4-5. In the caseswhere there are a-protons on either R or R'

,

rearrangement of the alkene can occur, as seen in thereaction of deoxybenzoin. Both the cis- andtrans-stilbene type structures and thel-methylene-l,2-diphenylethane structure are seen inthe product mixture. Interestingly, the dimethylationproduct was noted in both reactions but to a lesserextent, with respect to total alkene, in the highertemperature reaction. This would lead one to considerthat the intermediate responsible for dimethylationmust form prior to rearrangement of the alkene species


63/100

57

Me OAlMe.A Me Al Me OA3 i. \/R' R R'104 106Me MeCH, \ S\J Ai. AlMe,Me 3Al H0*^ N /^~A1^ jUMe.

- X07 >^ + 108'R'

112

Me Al3 fc 105 + possible rearrangement

Figure 4-5

in order to form the gem-dimethylation product.Support for this assertion comes out of the Allengroup. 36 They demonstrated the decrease in reactivity

HC=CH > HC=CHR > HC=CR > RHC=CHRFigure 4-6

of alkenes toward alkylation as its substitutionincreases as shown in Figure 4-6. 36


64/100

58

In stark contrast to the above results is thecase of a-tetralone. At either set of conditions themajor product isolated was that of the dimethylationreaction. Some alkene (23.6%) and 1-methylnapthalene(16.9%) were noted at the lower temperature; onlytrace amounts of these compounds were observed in thereaction mixture when subjected to the highertemperature conditions. It is believed that theisolation of the alkene in greater amounts is possibleby variation of the reaction time and, mostimportantly, reaction temperature.


65/100

CHAPTER V

SUMMARY

A new synthetic route to vinyl oxiranes that issuperior in some aspects to previous methods has beendemonstrated in the second chapter of this work. Thisroute utilizes the diphenylphosphinoyl group as ananchor to give stable crystalline intermediates ratherthan the volatile liquids encountered in otherschemes. The structure of 42 was assigned the cisconfiguration through mechanistic, chemical andspectral considerations, and ultimately through singlecrystal X-ray analysis. Yields of the vinyl oxiranes24 and 56 obtained by this sequence range from 55-65%.Although these non-optimized yields are somewhat lowerthan anticipated, it is believed that the advantagesassociated with this modified Horner-Wittig approachmerit consideration for the generation of the morevolatile vinyl oxiranes. Additionally, this work hasspurred continued studies devoted to the extension ofthe use of the diphenylphosphinoyl group as asynthetically useful tool for the synthetic organicchemist.

59


66/100

60

The work contained in the third chapter of thisdissertation, in conjunction with that of Cuevas, haspresented the use of aluminum enolate methodology inorganic synthesis. The scope and reactivity of thesereagents has been investigated yielding favorableresults and suggest continued investigation andexploitation of this methodology. Though the exactstructure of Rathke alane has not yet been proven and,therefore, the details of the mechanism of reactionwith vinyl oxiranes remain unclear, ample evidenceexists to support the suggested ice-tong pathway forring-opening. Continued studies in this area shouldinclude investigations incorporating chiral inductionagents on the aluminum atom as well as continuedefforts with the enolate.

The fourth section of this dissertationdemonstrates a use of the Me 3Al species as anadditionelimination reagent under thermal conditions.The key reaction demonstrated is the formation of2-methylcamphene from camphor in a single, high-yielding step. This previously overlooked applicationdemonstrates the need for continued expansion of theresearch into this field. Studies of more complexalkylaluminum reagents on structurally usefulskeletons, e.g. camphor, may yield incredibly powerfulsynthetic tools.


67/100

CHAPTER VI

EXPERIMENTAL

General ExperimentalMelting points were taken on a Thomas-Hoover

capillary melting point apparatus. Elementalanalyses were performed by the University of FloridaSpectroscopic Services. Proton and carbon NMR spectrawere recorded on either of two instruments, a VarianVXR XL-300 or a General Electric QE-300, unless notedotherwise. Proton chemical shifts were recordedrelative to the residual solvent peak (chloroform 7.26 ppm, unless otherwise noted). Carbon chemicalshifts are reported relative to the deuteriochloroformresonance at 77.00 ppm. Coupling constants arereported in Hertz (Hz) . Infra-red spectra were run ona Perkin-Elmer Model 1600 FT-IR spectrophotometer.Electron impact/ low resolution mass spectra wereobtained on a Finnigan MAT 4500 mass spectrometer at7 eV. A Finnigan MAT 95 spectrometer was used forhigh resolution electron impact and chemicalionization exact mass determination.

61


68/100

62

Apparatus and TechniqueAll glassware used for air-sensitive reactions

was flame dried under vacuum and filled with an inertatmosphere of either argon or nitrogen by successivepurging and charging using a dual manifold vacuumline. Standard syringe technique was used for theintroduction of liquid reagents and solutions to thereaction vessels. Purified samples were obtained bydistillation, recrystallization, or flash columnchromatography

.

3 7

Reagents and SolventsThe strength of the alkyl lithium reagents used

was determined by titration with 2, 5-dimethoxybenzyl-alcohol. 38 Tetrahydrofuran (THF) , hexane, toluene,and diethyl ether, when used as reaction solvents,were distilled from sodium-benzophenone. 3Diisopropyl amine and methylene chloride weredistilled from calcium hydride.1-fDiphenvlphosphinovl^methvlcyclohexanol f451

.

Cyclohexanone (1.17 g, 12.0 mmol) was added to astirred solution of the lithium salt 44 (1.2 eq, 14.4mmol) in THF (3o mL) at -78 C. The reaction wasstirred at dry ice temperatures for 15 min beforebeing allowed to warm slowly to room temperature wherea solid began to precipitate. THe solid was filtered,dissolved in methylene chloride (CH 2 C1 2 ), dried over


69/100

63

magnesium sulfate (MgS0 4 ), and solvents removed toyield a white powder 45 (3.32 g, 10.6 mmol) with am.p. - 163-166 C. rH NMR 6 7.75 (m, 4HAr ), 7.48 (m,6HAr ), 4.82 (br s, 1H 0H ) / 2.56 (d, J = 9.9 Hz, 2H C p),1.67 (m, 4H ring ), 1.29 (m, 6H ring ). 13C NMR 6 133.6(d, J = 98.1, C ipso ), 131.8 (d, J = 2.8, CAr ), 130.4(d, J = 9.6, CAr ), 128.7 (d, J = 11.8, CAr ), 72.2 (d,J = 6.0, C 0H ), 40.7 (d, J = 3.8, C P6 ), 39.7 (d, J =8.3, Ca _ 0H ), 25.4 (C ring ), 22.0 (C ring ). HighResolution Mass Spectrum (HR/MS) for C 19H 2 30 2 P -314.1430 f ound/ 314.1436 ca c ; Fragmentation (70 eV)315 (M + 1 [self C.I.]/ 26.1), 314 (M + , 35.2), 296(44.9), 271 (82.3), 258 (35.3), 215 (Ph 2P (O) CH 2 + ,base), 201 (Ph 2 P(0) + , 79.5), 91 (24.5), 77 (41.5).Anal. Calcd: C, 72.60; H, 7.32. Found: C, 72.15; H,7.39.l-(Diphenvlphosphinoyl)methylcyclohex-2-en-l-ol (41)

Methyldiphenylphosphine oxide (26.9 g; 0.124 mol)was dissolved in dry THF (100 mL) and cooled to 0 Cunder an Ar blanket. n-Butyl lithium (BuLi, 2.5 M,49.6 mL; 0.124 mol) was added dropwise via syringe toyield a bright yellow-orange solution which wasstirred an additional 15 min. Dropwise addition ofthe 2-cyclohexen-l-one (12.6 g; 0.131 mol) produced ablood-red solution, which after stirring a further 15min, was guenched by addition of water (50 mL) . The


70/100

64

reaction solution was extracted with methylenechloride (CH 2 C1 2 , 3 x 100 mL) , the organic layerscombined, washed with brine and dried over magnesiumsulfate (MgS04 ). Removal of the solvent left a whitepowder (35.98 g, 93 %) , m. p. (acetone) 152-154 C. 12*H NMR 5 7.80 ( m, 4 HAr ), 7.50 (m, 6 HAr ), 5.72 (d, J= 10.2 Hz, 1 Hvinyl ), 5.64 (dt, J d = 10.2, J t = 3.5, 1Hvinyi), 5.10 (br s, 1 Hon), 2.84 (dd, J = 15.2, 10.6,1 Hcp), 2.65 (dd, (J = 15.2, 8.6, 1 H CP ), 2.10 - 1.40(m, 6 H ring ). 13 c NMR 6 134.1 (d, J = 95.4 Hz,C ipso ), 133.9 (d, J = 95.2, C ipso ), 132.6 (d, J = 9.7,Cvinyi), 131.8 (d, J = 2.7, C Ar ), 130.5 (d, J = 9.5,C Ar ), 130. 4(d, J = 9.6, CAr ), 129.0 (s, C vinyl ), 128.8(d, J = 11.9, C Ar ), 128.7 (d, J = 11.9, CAr ), 70.2 (d,J = 5.3, C 0H ), 40.4 (d, J = 69.0, C p ), 37.6 (d, J =7-0, Ca0H ), 24.7 (s, Ca _ vinyl ), 19.0 (s, C ring ). HighResolution Mass Spectrum (HR/MS) for C 19H 21 2 P -312. 1272 found/ 312.1279 calc ; Fragmentation (70 eV)313 (M + 1 [self C.I.], 42.2 %) , 295 (base), 284(27.7), 215 (Ph 2P(0)CH 3 \ 85.5), 202 (Ph 2P(0)H\33.0), 91 (25.8), 77 (30.8). IR^m 1 ) : 3500-3100broad, 3063, 2980, 1438 (P-C) , 1182, 1167. Anal.Calcd: C, 73.08; H, 6.73. Found: C, 72.93; H, 6.76.1- (diphenylphosphinovl)methvl -? , 3-oxidonynlohexan-i -mI 3 I i

To a solution of 41 (4.00 g, 12.8 mmol) in


71/100

65

methylene chloride (CH 2 C1 2 , 40 mL) at 0 C was added,with stirring, a solution of m-chloroperbenzoic acid(MCPBA, 65 %, 5.34 g, 30.9 mmol) in CH 2 C1 2 (25 mL) atapproximately 1 drop per second. This solution wasstirred for 24 hr at which time thin layerchromatography (TLC) indicated no starting materialremained. The reaction was guenched with 0.32 Magueous sodium thiosulfate (NaS203, 40 mL) .The organiclayer was then washed with 1.0 M NaOH followed bybrine solution, dried (MgS04) , and concentrated invacuo to yield a yellow-tinted oil. This oil wastriturated with pentane to afford 42. (3.98 g, 12.1mmol, 94.8 %) as a white solid: m.p. 120-124 C ; 1HNMR 6 7.80 (m, 4 HAr ), 7.50 (m, 6 HAr )/ 4.39 (br s, 1H 0H ), 3.25 (d, J = 3.5 Hz, 1 H epox ), 3.17 (dt, J d =3.6, J t = 1.2, 1 H epox ), 2.75 (16 line m, 2 H C p),1.96-1.11 (4 m, 6 H ring ). 13 C NMR 5 133.9 d, J = 99.7Hz, C ipso ), 133.6 (d J = 99.7, C ipso ), 131.9 (3 lines,CAr ), 130.5 (3 lines, CAr ), 128.7 (4 lines, C Ar ), 70.9(d, J = 4.6, C 0H ), 58.3 (d, J = 9.6, C epox ), 54.8 (s,C epox ), 38.0 (d, J = 70.5, C P ), 34.4 (d, J = 5.8,Ccoh), 22.6 (s, Caepox), 16.6 (s, C ring ). HR/MS forCi 9H 2 i0 3P - 328. 1229 f 0und , 328 . 1219 c i c ; Fragmentation(70 eV) : 329 (M + 1 [self C.I.], 26.3 %) , 328 (M + ,8.7), 311 (12.8), 258 (24.1), 215 (Ph 2 P(0) CH 3 + , 51.1),202 (Ph 2P(0)H + ,base) , 91 (12.9), 77(33.5). IR (cm-1 ):


72/100

66

3600-3100 broad, 2987, 1437, 1166, 1119.Preparation and attempted epoxidation of 1-fdiphenyl -phosphinovl)methvl-l-methoxvcvclohex-2-ene (46) .

Hydroxyalkene 41 (0.80 g, 2.6 mmol) was dissolvedin 5 mL of dry THF and the solution cooled to 0 C.While stirring, potassium hydride (0.11 g, 2.8 mmol)was added in one portion and the resulting yellowsolution was stirred for an additional 15 min.Addition of excess methyl iodide (Mel, l.lg, 7.8 mmol)resulted in immediate disappearance of the yellowcolor. After further stirring (15 min) brine wasadded and the aqueous solution extracted withmethylene chloride (3 x 10 mL) . The organic layerswere combined, dried (MgS04) , and concentrated invacuo to yield a thick, yellow oil. % NMR of thecrude oil showed complete conversion to the methylether, 46; XH NMR 5 7.75 (m, 4HAr ), 7.45 (m, 6HAr )

,

5.83 (dt,J d = 10.4 Hz, J t = 3.9, H vinyl ), 5.68 (d, J =10.4, Hvinyl ), 2.99 (s, 3H0Me ), 2.70 (d, J = 2.1,H CP ), 2.66 (d, J = 1.6, H CP ), 2.05-1.55 (m, 6H ring );13 C NMR 6 135.1 (d, J = 99.7 Hz, C ipso ), 134.7 (d, J =99.6, C ipso ), 131.7 (C vinyl ), 131.6 (d, J = 2.9,Cvinyi), 131.0 (d, J = 2.7, C Ar ), 130.7 (d, J = 9.2,CAr ), 130.6 (d, J = 9.2, CAr ), 128.5 (d, J = 12.0,CAr ), 128.1 (d, J = 11.6, CAr ), 128.0 (d, J = 11.7,CAr ), 74.5 (d, J = 4.2, C 0H ), 49.9 (C 0Me ), 40.3 (d, J


73/100

67

= 69.9, C P ), 33.0 (d, J = 5.0, Cq-coh), 24.7(Ccx-vinyl)/ 19.3 (C r ing).

Without further characterization the crude methylether, 46, from above was treated with 1.5 equivalentsof MCPBA (55 %) in CH 2C1 2 at C with slow warming toroom temperature. Workup as in the epoxidation of 41yielded a crude yellow oil on solvent removal which on1H NMR examination (CDC1 3 ) revealed a rather congestedepoxide region (6 3.15-3.30) suggesting a mixture ofcis and trans isomers of the epoxide. The methoxypeak lies in the middle of this epoxide region,further complicating the interpretation. A 1:1solvent mixture of benzene-d 6 and CDC1 3 did little toresolve the region. The 13 C NMR is very complex withpeaks from the alkene 46 and what looks like twoisomers of the epoxide.1- ( Diphenylphosphinoy 1 ) methy1-1 ,2.3 -trihvdroxcyclo -hexane (49) .

Attempts to recrystallize (3:1 EtOAc: hexane) theepoxide, 42., with heating (55 C) resulted in theformation of a difficultly soluble white powder, m.p.183.5-187.5 C, and recovery of epoxide 42. as a yellowoil. Attempts to recrystallize the white solid wereunsuccessful. 6: ^-H NMR 5 7.85 (m, 2 H Ar ), 7.70 (m,2 HAr ), 7.50 (m, 6 HAr ), 4.94 (baseline roll, 1 H h) /3.76 (br s, 1 H 0H ) , 3.65 (ddd, J = 11.5, 8.8, and 4.5


74/100

68

Hz, 1 H C3 -oh), 3.29 (d, J = 8.8, 1 H C2 -oh), 2.78(overlapping dd, J = 15.3, 14.2, 1 H C p), 2.59 (dd, J =15.4, 8.1, 1 H C p), 2.02-1.21 (m, 6 H ring ). 13C NMR 5ipso carbons not seen, 132.1 (d, J = 3.7 Hz, CAr )

,

130.9 (d, J = 9.6, CAr ), 130.3 (d, J = 9.2, C Ar ),128.8 (d, J = 12.0, CAr ), 79.7 ( d, J = 5.3, C 2 -oh),75.3 (d, J = 4.8, Ci.oh), 71.2 (s, C 3 -oh), 40.0 (d, J= 68.9, C P ), 38.5 (d, J = 8.3, Cal - H)# 31.3 (s,CQ 3-oh), 18.9 (s, C ring ). HR/MS for C 19H230 4 P -346. 1330 f 0U nd/ 346. 1370 c | c ; Fragmentation (70 eV)347 (M + 1, [self C.I.]/ 49.8 %) , 328 (M - 17, 9.6),311 (7.6), 258 (21.2), 215 (Ph 2 P (O) CH 3 + , 23.7), 202(Ph 2P(0)H + ,base) , 77 (19.3). IR(cm- 1 ): 3401, 3248,2931, 1437, 1173, 1114, 1079, 979. Anal. Calcd: C,65.89; H, 6.65. Found: C, 65.53; H, 6.65.Formation of acetonide 52.

p-Toluenesulfonic acid monohydrate (5 mg) and2,2-dimethoxypropane (3.0 g, 29 mmol) were refluxedin 5 mL benzene for 10 min and allowed to cool underan Ar blanket. Triol 49 (0.45 g, 1.3 mmol), dissolvedin 10 mL of benzene and 2.5 mL of methylene chloride,was added and the reaction mixture was refluxed withstirring for 2 hr. TLC examination indicated a singlecomponent with a different R f than that of thestarting material. Solvent removal afforded a whitepowder (0.46 g, 1.2 mmol, 91.6 %) , m. p. (acetone)


75/100

69

164-168 C. 8: lH NMR 8 7.80 (m, 4 HAr ), 7.50 (m, 6HAr ), 4.74 (br s, 1 H 0H ), 3.93 (ddd, J = 11.6, 9.1,and 4.0 Hz, 1 H C3 -oh), 3.08 (d, J = 9.0, H C2 -oh), 2.52(dd, J = 15.2, 12.0, 1 H CP ), 2.52 (dd, J = 15.2, 8.2,1 H CP ), 2.07 (br d, 1 H ring ), 1.93 (br d, 1 H ring ),1.73-1.04 (m, including two methyl singlets at 1.45and 1.25, 10 H) . 13C NMR (ipso carbons not seen) 6131.8 (d, J - 2.8 Hz, CAr ), 131.7 (d, J = 2.2, CAr ),130.5 (d, J = 9.1, CAr ), 130.3 (d, J = 9.8, CAr ),128.7 (d, J = 11.5, CAr ), 128.5 (d, J = 12.0, CAr ),108.7 (s, C acetal ), 85.6 (d, J = 9.9, C ) , 73.7 (d, J= 2.0, C ), 73.1 (d, J = 6.3, C ), 37.8 (d, J = 65.6,C P ), 37.3 (s, Ca _ acetal ), 28.8 (s, Ca - 0H ), 27.0 (s,CH 3 ), 26.7 (s, CH 3 ), 19.8 (s, C ring ). HR/MS(chemical ionization) for C 22H 28 4P - 387. 1739 f un ,387. 1725 ca i c ; Fragmentation (70 eV) : 387 (M + l, selfC.I.), 347 (M - CH 3 C(0)CH 2 , 4.21 %) , 328 (5.1), 310(14.5), 258 (16.8), 215 (Ph 2 P (O) CH 3 + , 33.1), 202(Ph 2P(0)H + , base), 155 (13.8), 125 (12.3),77 (14.4). IR (cm- 1 ): 3330 (broad), 3074, 2932,1712, 1590, 1439, 1175, 1112, 1081.3-methvlene-1. 2-oxidocvclohexane (24)

.

Epoxide 42 (3.0 g, 9.2 mmol) was dissolved in 5mL dry THF and warmed to 60 C with stirring. Sodiumhydride (0.26 g, 11 mmol), previously washed withpentane, was suspended in 2 mL of dry THF and the


76/100

70

hydride slurry added slowly to the warm epoxidesolution causing a color change from yellow to brownand evolution of hydrogen. Upon completion of hydrideaddition (20 min) the solution was allowed to stir foran additional 0.5 hr during which time a suspendedsolid formed. The solution was allowed to cool toroom temperature and an equal volume of water wasadded to dissolve the solid. The resulting singlephase solution was extracted with ether (3x5 mL)and the combined ether extracts were washed with 1 MNaOH and brine solutions. Drying (MgS04) and solventremoval in vacuo yielded a yellow tinted oil, 1.03 g(25% THF by XH NMR) 77 %. ^ NMR 6 5.23 (dd, J = 1.4,1.7 Hz, 1 H vinyl ), 5.11 (dd, J = 1.4, 1.5, 1 Hvinyl ),3.42 (d, J = 3.9, 1 H epox ), 3.38 (7 line m, 1 H epox ),2.27 (m, 1 H ring ), 2.03 (m, 2 H ring ), 1.83 (m, 1Hring), 1.59 (m, lH ring ), 1.40 (m, lH ring ). 13CNMR5 142.6 (C vinyl ), 116.1 (Cvinyl ), 55.1 (C epox ), 54.2(Cepox), 28.6 (C ring ), 24.0 (C ring ), 19.7 (C ring ).LR/MS fragmentation (70 eV) 110 (M + , 28.2 %) , 95(40.4), 81 (57.8), 67 (46.5), 53 (43.3), 41 (79.1), 39(base)l-(Diphenvlphosphinoyl)methvl-3-methvlcvclohex-2-en-l-ol (54).

To a cooled solution (0 C) of diphenylmethyl-phosphine oxide (13.9 g, 64.4 mmol) in THF (50 mL)


77/100

under argon was added a hexane solution of BuLi (2.5M, 25.8 mL, 64.5 mmol). The resulting golden yellowsolution was stirred for an additional 20 min and thencooled to -25 C, whereupon3-methylcyclohex-2-en-l-one (7.09 g, 64.4 mmol) wasadded dropwise to give a red-orange solution. Stirringwas continued at -25 C for 20 min before the solutionwas allowed to warm slowly to room temperature. Aftera total of 6 hr stirring at room temperature thereaction flask was opened to the atmosphere andstirring continued until color abatement.Concentration of the reaction mixture in vacuofollowed by addition of water resulted in theprecipitation of a white solid (20.2 g, 97.1 %) whichwas recrystallized from acetone to yield 17.9 g (54.9mmol) of 54: 126-129 C; XH NMR 5 7.69 (m, 4 HAr ),7.41 (m, 6 HAr ), 5.30 (s, 1 H0H ), 4.92 (s, 1 Hvinyl ),2.63 (dd, J = 15.0, 8.4 Hz, 1 H CP ), 2.50 (dd, J =15.0, 10.6, 1 H CP ), 1.72 (m, 4 H ring ), 1.43 (m, methylsinglet at 1.42, 5 H ,2 H ring and CH 3 ). 13 C NMR 5136.7 (s, C vinyl ), 133.9 (d, J = 99.6 Hz, C ipso ),133.6 (d, J = 98.1, C ipso ), 131.5 (d, J = 3.0, CAr ),131.4 (d, J = 3.1, CAr ), 130.3 (d, J = 8.2, CAr ),130.1 (d, J = 9.3, CAr ), 128.5 (d, J = 11.9, CAr ),128.4 (d, J = 11.8, CAr ), 127.5 (d, J = 9.8, C vinyl ),70.5 (d, J = 5.0, C 0H ), 40.6 (d, J = 69.0, C P ), 36.9


78/100


79/100

73

Hring)/ 1.05 (s, 3 Hmet hyi). 13 C NMR 5 134.2 (d, J =111 Hz, C ipso ), 134.1 (d, J = 114, C ipso ), 131.9 (3lines C Ar ), 130.5 (3 lines, CAr ), 128.6 (3 lines,CAr ), 70.8 (d, J = 3.0, C 0H ), 65.2 (d, J = 9.1,Cepox), 61.1 (s, Cepox), 38.9 (d, J = 70, C P ), 34.7(d, J = 5.3, Ca _ 0H ), 28.2 (s, Ca . epox ), 23.8 (s,Cmethyl), 16.9 (S, C ring ). HR/MS for C20H23O3P -342. 1392 found/ 342.1385 c a i c ; Fragmentation (70 eV) 342(M + , 5.1 %), 328 (1.7), 323 (16.5), 258 (20.1), 215(Ph 2P(0)CH 3 \ base), 202 (Ph 2P(0)H + , 61.0), 91 (8.1).IR (cm 1 ): 3363 (broad), 2938, 1437, 1157, 1119.1- (Diphenvlphosphinovn methvl-i . 2 , 3-trihvdroxv-3-methvlcvclohexane f57^ .

Upon chromatography of the above oxirane 55, alate eluting compound was realized in the form oflarge colorless crystals, m.p. (EtOAc) 166-168 C. lHNMR 6 7.82 (m, 2 HAr ), 7.69 (m, 2 HAr ), 7.48 (m, 6HAr ), 4.46 (d, J = 5.7 Hz, 1 H) , 4.13 (s, 1 H) , 3.52(d, J = 5.8, 1 H), 2.95 (dd, J = 15.3, 13.1, 2 H, H CPand H 0H ), 2.62 (dd, J = 15. 4, 8.5, 1 H CP ), 1.90-1.26(m, 6 H ring ), 1.24 (s, 3 Hmethyl ). 13 C NMR 6 133.7(d, J = 100.9 Hz, C ipso ), 132.7 (d, J = 99, C ipso ),131.9 (d, J = 3.3, CAr ), 131.8 (d, J = 3.1, CAr ),130.7 (d, J = 9.6, CAr ), 130.3 (d, J = 9.6, CAr ),128.7 (d, J = 12.6, CAr ), 80.0 (d, J = 6.1, C C2 -oh),75.2 (d, J = 5.3, Ccloh), 73.4 (s, C C3 -oh), 39.5 (d,


80/100

74

J = 69.1, C P ), 37.6 (d, J = 6.9, Cq-ci-oh), 36.8 )s,Cq-cs-oh), 23.5 (s, Cmeth yi), 18.7 (s, C ring ). HR/MS(chemical ionization) mass for C20H25O4P -361. 1571 f ound/ 361. 1569 c 1 c ; Fragmentation ( 70 eV)361 (M + 1, 99.7 %) , 342 (M - H 2 0, 22.0), 324 (14.9),271 (15.7), 258 (14.6), 243 (11.4), 215 (Ph 2P (O) CH 3 + ,56.7), 202 (Ph 2P(0)H + , base). IR (cm 1 ): 3460(broad), 3342 (broad), 2943, 1431, 1167, 1120. Anal.Calcd: C, 66.66; H, 6.94. Found: C, 66.63; H, 7.10.l-Methvl-3-methvlene-1.2-oxidocyclohexane (56)

To a solution of potassium hydride (1.54 g, 3 8.5mmol) in 20 mL of dry THF warmed to 60 C was addeda CH 2C1 2 (100 mL) solution of epoxide 55 (12.0 g, 35.1mmol) via cannula transfer. An immediate color changeof the solution from colorless to a brown accompaniedthis addition. Continued addition was carried out insuch a way as to minimize foaming due to gasevolution. After addition of 3b was completed thesolution was stirred for an additional 2 . 5 Hr and thencooled to room temperature. Cooling led to theprecipitation of a white solid. The reaction wasguenched by the addition of agueous K 2 C0 3 (1.1 eg) andthe agueous mixture extracted with ether. Thecombined organic layers were washed with brine anddried (MgS0 4 ) . Analysis of the etheral solution byGC/MS showed the presence of the desired methylene


81/100

75

oxirane 56 as the predominant component other thanTHF. Distillation (b.p. 30-32 C at 3 mmHg) provided2.82 g of lb (65 % yield): XH NMR 8 5.16 (br d, J =1.4 Hz, H vinyl ), 5.05 (dd, J = 3.2, 1.6, Hvinyl ), 3.21(s, H epox ), 2.23 (m, 1 H ring ), 1.94 (m, 2 H ring ), 1.53(m, 3 H ring ), 1.33 (s, 3 Hmethyl ). 13 C NMR 6 142.9(Cvinyi), H5.9 (C vinyl ), 62.2 (C epox ), 59.5 (C epox ),29.4 (C ring ), 28.3 (C ring ), 23.3 (Cmethyl ), 19.9(Cring). LR/MS Fragmentation (70 eV) 124 (M + , 9.6),109 (M - CH 3 , 21.1), 95 (16.8), 81 (50.4), 55 (30.9),43 (base)Preparation of the C-methvlation product 64.

One equivalent of BuLi (2.4 M, 0.5 mL; 1.2 mmol)was added to 41 (0.39 g; 1.2 mmol) dissolved in dryTHF at 0 C to yield a bright yellow solution. Oneadditional drop of the BuLi beyond 1 eq turned thesolution orange in color which darkened with continuedaddition up to 2 eq. This solution of the dianion wasstirred for 15min and then quenched with Mel (1.2 g, 7eq) to give a colorless solution. A white, waxy,solid was isolated (0.35 g, 86 % yield). The crude %NMR of the solid showed it to be the C-methylationproduct 64 .Preparation of tert-butvlpropioate.

Propionyl chloride (74.6 g, 0.806 mol) was slowlyadded to a stirred ethereal solution of tert-butanol


82/100

76

(65.2 g, 0.881 mol) and N,N-dimethylaniline (110 g,0.909 mol) to yield a pale blue solution that darkenedover time to a dark blue. The reaction was quenchedafter 30 hr by careful addition of water (30 mL) thenthe mixture was extracted with ether (2 X 15 mL) . Theorganic layers were combined and washed successivelywith 10% HC1 (6X5 mL) , saturated NaHC0 3 (3x5 mL)

,

and brine (2 X 10 mL) and dried over MgS0 4 .Distillation through a Vigreaux column afforded thetert-butyl propionate as a colorless liquid, b. p.117-120 C (lit. 119-121 C 40 ), 69.0 g, 65.7%. XH NMR5 2.23 (q, 2H) , 1.44 (s, 9H) , 1.08 (t, 3H) . 13 C NMR 5173.7 (Cester ) ( 79.7 (C 0R ), 28.6 (Ca -c-o), 27.9(Cme thyi), 9-00 (Cp-c-o).Preparation of 1-adamantvlacetate (99)

.

To 20 mL anhydrous triethylamine was added1-adamantanol (10 g, 65.7 mmol) , N,N-dimethyl-aminopyridine (DMAP, 20 mg) , and acetic anhydride(14.8 g, 144.6 mmol) and the resulting mixture washeated with stirring to 90-95 C under an argonatmosphere for 30 hr. The reaction solution wasconcentrated on the rotary evaporator to give a yellowoil. The oil was taken up in ether and washed with 1N sodium hydroxide (NaOH) and brine and dried overMgS0 4 . Concentration under vacuum gave a clear oilthat crystallized upon sitting to give colorless


83/100

77

needles (12.33 g, 96% yield, m.p. 31.5-32.0 C, lit.32.5-33.5 C). 41 XH NMR (ppm) : 2.07, broad,irregular d, 6 H; 1.89, s, 3 H; 1.64, s, 3 H; 1.56,broad, irregular d, 6 H. 13 C NMR (ppm): 170.3,80.2, 41.3, 36.2, 30.8, 22.7 . Mass Spectrum: 194(M + , 2.25%), 134 (base), 95 (36.0), 92 (73.5), 79(23.42), 43 (41.5).Genera l procedure for the reaction of aluminumenolates on electrophiles.

The aluminum enolate reactions were carried outwith a 2 . molar excess of the organoaluminum species(3.0 eq alane to 1.0 eq of the electrophile) exceptwhere noted. Diisopropyl amine at 0 C was dissolvedin hexane and BuLi added to it dropwise to generatethe lithium diisopropyl amide which was then stirredfor 30 min. The dropwise addition of the ester(tert-butyl acetate, tert-butyl propionate, or1-adamantyl acetate) was carried out at -78 C and theresulting enolate was stirred for 30 min beforewarming to 0C and opening to vacuum to remove thereaction solvents. The lithium salt obtained waspumped on for 30-60 min before being dissolved in THFand cooled to dry ice/ acetone temperature. Thedialkylchloroalane was added dropwise keeping thetemperature below -60 C after which the electrophile(epoxide, ketone, or aldehyde) was immediately added,


84/100

78

dropwise. The reaction was generally allowed to stirfor one hour before being quenched via cannulatransfer into a rapidly stirring solution of 10% HC1and ice. The reaction mixture was extracted withdiethyl ether, washed with water and brine, dried withMgS0 4 and solvents removed in vacuo to yield thehydroxyester

.

Reaction of RkeAl 6 with 3-methvlene-l . 2-oxidocycl o-hexane 24.The oxirane 24 (0.179 g, 1.63 mmol) was reacted

under three stoichiometric ratios with the Rathkemetals; the results of these studies are collected inTable 3-1. The yield of the hydroxy ester 25 wasdetermined by capillary GC analysis against aninternal standard (tridecane) . 1H NMR 5 4.82 (s,IHvinyl), 4.68 (s, lH vinyl ), 3.27 (br s, 1H 0H ), 2.55(m, lHa _o H ), 2.41 (m, 2Ha . c=0 ), 2.18 (dt, J = 17.1,7.6, lH allylic ), 1.94 (m, 2H ring ), 1.70 (m, 2H ring ),1.39 (m, 2H ring ), 1.35 (s, 9H t _ Bu ). X 3C NMR 5 172.9(Cester), 148.3 (C vinyl ), 108.3 (C vinyl ), 80.5 (C 0R ),74.3 (C 0H ), 48.0 (C allylic ), 35.9 (C ring ), 34.8(C r ing), 34.1 (C ring ), 28.0 (C t _ Bu ), 24.4 (Ca - C =o).Rincr opening to form chlorohydrin 79 and glycol80 from oxirane 24.

The ring opened products (79, 80) were formedfrom 24. in two different ways: 1) reaction of 24(0.10 g, 0.93 mmol) with Et 2AlCl (0.54 mL, 0.93


85/100

79

mmol)in THF at -65 C for 25 min followed by anaqueous workup and 2) a THF solution of 24 (0.12 g,1.1 mmol) was subjected to standard dilute acid workupconditions (5 mL 10% HC1, -50 g ice). Both reactionswere extracted with ether (3 X 25 mL) , washed withbrine (1 X 10 mL) , and dried over MgS0 4 . The productswere separated by gradient elution flash columnchromatography (hexane to EtOAc) . The chloroalanereaction favored formation of the later eluting glycol80 (1:2.1) while the workup conditions produced moreof the chlorohydrin 79 (1.6:1). 79: lH NMR 8 5.24(s, lHvinyl ), 5.00 (s, lH vinyl ), 4.27 (d, J = 8.6 Hz,IHaiiyiic), 3.62 (m, IHo-oh), 2.2.48 (m, 2H ring ), 2.18(m, lH ring ), 2.04 (m, lH ring ), 1.76 (m, lH ring ), 1.53(m, lH ring ). 13 C NMR 8 143.9 (C vinyl ), 112.2(Cvinyl), 75.0 (C 0H ), 68.7 (C C i), 33.4 (C ring ), 31.4(Cnng), 23.3 (C ring ). 80: 5.02 (brs, lH vinyl ),4.86 (br s, lH vinyl ), 3.88 (d, J = 8.3, 1H C2 -oh), 3.36(m, 1Hci- h), 2.78 (br s, 1H 0H ) , 2.59 (br s, 1H 0H )

,

2.37 (m, lH ring ), 2.03 (m, 2H ring ), 1.76 (m, 2H ring ),1.49 (m, lH ring ), 1.32 (m, 2H ring ). 13 C NMR 6 148.6(Cvinyl), 107.2 (C vinyl ), 78.8 (C 0H ), 77.1 (C 0H ), 34.6(Cnng), 33.4 (C ring ), 25.2 (C ring ).Formation of hydroxvester 81 from cvclohexanone f43)

.

The hydroxyester 81 was formed from cyclohexanonein two ways: 1) by action of 1.0 eq Rathke's salt 4


86/100

80

to give 38.5% conversion (GC) and 2) reaction with 1.0eq of RkeAl 6 to give 76.7% conversion (GC) . Theprocedure followed was as outlined in the generalprocedure above. 81: XH NMR 5 3.63 (s, 1H 0H ) , 2.38(s, 2Ha _ c =o), 1.66 (m, 4H ring ), 1.47 (s, 9Hmethyl ),1.44 (m, 2H ring ), 1.29 (m, 4H ring ). 13 C NMR 8 172.0(Cester), 81.4 (C0R ), 70.0 (C 0H ), 46.3 (Ca-c-o), 37.5(C ring ), 28.2 (Cmethyl ), 25.7 (C ring ), 22.1 (C ring ).Reaction of RkeAl 6 with benzaldehvde (85)

.

Benzaldehyde (1.76 g, 16.6 mmol) was added to a-65 C solution of the Rathke alane (20.0 mmol) andexcess Et 2AlCl (6.6 mmol). No change was noted on thereaction mixture composition after 15 min reactiontime; the reaction was worked up as described in thegeneral RkeAl procedure above to yield, afterKugelrohr distillation (2.40 g, 10.8 mmol), 86: 1HNMR 5 7.38 (m, 5HAr ), 5.17 (br S, 2Ha . Ar & 0H ) , 2.62(br d, 2Ha -c=o), 1-43 (s, 9H t _ Bu ).Reaction of RkeAl 6 with anisaldehyde (89)

.

Anisaldehyde (2.27 g, 16.6 mmol) was added to a-65 C solution of the Rathke alane (20.0 mmol) andexcess Et 2AlCl (6.6 mmol). No change was noted on thereaction mixture composition after 15 min reactiontime; the reaction was worked up as described in thegeneral RkeAl procedure above to yield 90 after columnchromatography (4% EtOAc in CH 2 C1 2 ) : XH NMR 5 7.27


87/100

81

(d, J = 13.3 Hz, 2HAr ), 6.84 (d, J = 13.3, 2HAr ), 5.00(m, lHa . Ar ), 3.61 (br s, 1H 0H ) , 2.62 (irr t, 2Ha -c-o),1.43 (s, 9H t _ Bu ).Reaction of MeRkeAl 91 with 3-methvlene-l . 2-oxido-cvclohexane (24)

.

This reaction resulted in the formation of twoisomers in a 1.2:1 ratio (0.30 g, 68%); the spectraldata is listed with what is believed to be isomericsignals in brackets. 1H NMR 5 4.92 [4.79] (s,IHvinyi), 4.84 [4.71] (s, lHvinyl ), 3.83 (dq, lHa _ 0H ),3.71 [3.60] (t, lH allylic ), 2.66 (m, lHa _ c -o),1.80-1.40 (m, 6H ring ), 1.36 (s, 9H t _ Bu ), l.ll [0.94](d, J = 8.24 HZ, 3HMethyl ). 13 C NMR 6 175.2 [173.8](Cester), 146.5 [145.2] (C vinyl ), 113.6 [111.8](Cvinyi), 80.4 [79.8] (C r), 69.6 [69.0] (Ca _ 0H ), 54.5[53.4] (C allylic ), 40.6 [40.0] (C ring ), 32.5 [31.1](C ring ), 28.7 [28.2] (C ring ), 28.0 (C t _ Bu ), 22.5[21.9] (Ca _ c=0 ), 16.0 [14.1] (CMet hyl).Reaction of AdmAl 100 with cyclohexanone f43) .

The general aluminum enolate reaction procedurewas followed with 100 (5.45 mmol) on cyclohexanone(0.104 g, 1.02 mmol). Following workup, the crudereaction product mixture was dissolved in methanol (15mL) and aqueous NaOH (5 mL, 6 M) and warmed to agentle reflux for 3.5 hr. Upon cooling the reactionmixture pH was lowered to ~10 by addition of 10% HC1


88/100

82

and extracted with ether (2 X 15 mL) to remove the1-adamantanol. Continued acidification followed byether extraction (2 X 15 mL) , drying (MgS0 4 ) , andsolvent removal in vacuo to give a yellow oil loi(0.12 g, 0.41 mmol, 40%). 1H NMR 6 6.34 (baselineroll), 2.53 (s, 2Ha _ c=0 ), 1.68 (m, 4H ring ), 1.49 (m,4H ring ), 1.31 (m, 2H ring ). 13 C NMR 6 176.8 (C c =o),70.6 (C 0H ), 45.0 (Ca-c-o), 37.2 (C ring ), 25.4 (C ring ),21.9 (C ring ).Reaction of AdmAl 100 with benzaldehyde f85)

.

The general aluminum enolate reaction procedurewas followed with 100 (3.06 mmol) on benzaldehyde(0.115 g, 1.00 mmol). Following workup, the crudereaction product mixture was dissolved in methanol (15mL) and agueous NaOH (5 mL, 6 M) and warmed to agentle reflux for 3.5 hr. Upon cooling the reactionmixture pH was lowered to -10 by addition of 10% HC1and extracted with ether (2 X 15 mL) to remove the1-adamantanol. Continued acidification followed byether extraction (2 X 15 mL) , drying (MgS0 4 ), andsolvent removal in vacuo to give a yellow oil 102(0.11 g, 0.37 mmol, 37%). XH NMR 6 7.27 (m, 5HAr )

,

6.30 (baseline roll), 5. 06 (dd, J = 9.3, 4.1 Hz,lH benzyl ), 2.68 (m, 2Ha _ c=0 ). 13 C NMR 6 176.7 (C c=0 ),142.1 (C Ar ), 128.6 (CAr ), 128.0 (CAr ), 125.7 (CAr ),70.3 (C 0H ), 43.1 (Ca-c-o)


89/100

83

Reaction of AdmAl 100 with l-methvl-3-methvlene-1 r 7 -oxidocvclohexane (56)

.

The general aluminum enolate reaction procedurewas followed with 100 (5.45 mmol) on the vinyl oxirane56 (0.23 g, 1.8 mmol). Following workup, the crudereaction product mixture was dissolved in methanol (15mL) and aqueous NaOH (5 mL, 6 M) and warmed to agentle reflux for 3.5 hr. Upon cooling the reactionmixture pH was lowered to ~10 by addition of 10% HC1and extracted wit

alkyl aluminum der 00 camp

Documents