Preparation and conformational analysis of severely hindered β-diketones. Dipole moment determinations and theoretical calculations.

00404020#91 $3.00+.00 0 1991Pergamcm Pms p

Tewah.e&on Vol. 47,No. 32,pp. 6511-6520.1991 Rited in GreatBrimii

M.

MXENO-M@AS*, A. -, A. m,

C. JAI& M.E. LLORIS,J. MARCUE& A.C. SIANI,A. VAUXBERA

Departmentof Chemistry.UliversitatAut&oma de Barcelona. Dellaterra.08193~Barcelona. [email protected]. I. DERN&DEPkUENT?B, M.F. REY-STOLLE,C. SAJAXY Dapartamento de QufmicaFisica I. Facultadde CienciasQufmicas. UniversidadComplutense. 2804~Madrid.Spain.

(Received in UK 8 May 1991)

Key words:p-Diketones; Preparationfrom CoUI) complexes; Conformational analysis;Dipolemoments;MolecularMechanics Abstract.-Severalp -diketonesbearing bulb substituentsat the intercarhonyl positionshave been prepared by alkylationof the cobalt(I1)canplexesof the unsubstituteddiketones. Agreement between experimentaland calculateddipole moments is fairly good thus rendering MolecularMechanics a safe tool for the conformational analysisof the title molecules.The most populatedConformations have been evaluatedfor nine P-diketonesbearingl- and 2- adamantyl,E-butyl, cyclohe+, 1-phenylethyl and bsnxhydrylgroups as well as two methyl groups and onemethylplus onel-adamantylgroups. lXrRODl_JCTION. Confonnationalanalysis of open-chain compounds is limited by their inherentconformational complexity.The advent of MolecularMechanics(m)2 has opened new possibilitiesin this field. Calculationsalone are not sufficient, however,and, if possible,cross checkingwith experimentaldata (NMR methods or others) is highly desirable. In a preliminarycotmnmi cation3we reportedthe useful combinationof ?1Mand dipole mcment

determinations to

evaluate the

conformational preferences of

3-(l-

adamantyl)pentane-2,4-dione (1) and 4-(1-adamantyl)-2,2,6,6-tetramathylheptane-3,5-dione (2). Since then, the combinationof dipole moment determinationsand theoretical calculationsof differentdegreesof complexityhas becane quite popular.4The idea is not new, however.Thus, in 1978 Allingerand coworkersstudiedthe conformations of 3,3dimethylpentane-2,4-dione (8) by a combinationof dipolemomentsand MM.' We now want to present an extensionof our work to severalp&ketones

bearing

secondaryand tertiarysubstituentsat the intercarbonylpositionas well as to some disubstituted ,+diketones.All tbe compoundsstudiedexist only in the diketo forms,no keto-enoltautcmersbeing detectedby NMR analysis.

6511

6512

M.MORENO-MANAS etal.

PREPARATIONOF coMpouNDG.Ihe studiedcmpounds are: 1, 2, 3-(2-adaraantyl)pentane-2,4dione (3), 3-m-butylpentane-2,4-dione(4), 3-cyclohexylpentane-2,4-dione (5), 3-(1(7), 8, and 3-(1phenylethyl)pentane-2,4_dione (6), 3-benzhydrylpentane-2,4+iione adammtyl)-3*thylpentane_2,4_dione (9). Ccqounds 1-7 were prepared by alkylationof the cobalt(I1)complexes of the uosubstituteddiketoneswith the correspondingalkyl halides.Compmnds 16a, 26a, 66b and 76b have been

previouslydescribedby us. Compound9 was preparedby a similar

alkylationof the copper

complex of 3methylpentane-2,4-dione.7 The disubstituted

diketme 8 was preparedby conventiooal methods.* Chummethod of alkylationhas proveduseful for the preparationof diketones3 snd 4 bearing a secondaryor a tertiarysubstituent. Other secondarysubstituents could not be introduceddirectly,probablybecause free radicalsare intermediates in the cobalt (and copper) method8 and secondaryfree radicalsdo not form easily. Compound 5 was, however,preparedby cobaltmediatedalblation with 3-branocyclohexene to afford 3-(2cyclohexen-l-yl)pentane-2,4_dione (lo),which gave 5 upon hydrogenation under palladia on charcoalcatalysis.This method is useful

for the introduction of secondaryradicals

at the intercarbonyl position,as evidencedby a similarsequenceleadingto 3-acetyl-4methyl-5-hepten-2-one (11)and 3-acetyl_4methyl-2-heptanone (12).

1 2 i z i 9

R1 Ml? t-&l Me Me Me Me Mf? Me Me

R3 H H H H &:ohexyl H 1-Phenylethyl H H Eenzhydryl Me Me 1-Adamanty1 Me

R2 1-Uamantv1 1-Adama& 2-Adamanty1

Co(acac)z +

2X-R -

MeCOCHRCOMe ?.I!

6 Co(acac)z +

R=2-Adamantyl,2 R=t-Bu, 4

+$$

W'd

)

3

2

0

Co(acach +

2

&

)

ji,&W'_

s s

6513

Severelyhinderedb-diketones

DIPOLE MMBD?S. Dipolemcments mere determinedby the Debye method, in cyclohexanefor all ccmpoundsand also in benzene for ccqounde 1 and 2. Two differentapproximations were followedin the evaluationof the experimental values:thecuggenheim-sbalthmethodg and the Halverstadt-KuQler method.1oIn the lattercase the contribution of the atomic polarizationwas consideredto be 5% and 15% of the electronicpolarization. As can be seen frcm Table 1 the three approachesshovedno significant differences. An MM2 analysiswas undertakenfor compounds1-9. Allinger's THRCRETICAbCACCULILTIONS. MM2(77) force field11p12togetherwith all MY2(85)parameters l3 were used throughthis work. One torsionalenergy surfacewas calculatedfor each product.The rotationof all other bonds producingnon-equivalentconformerswas undertakenby the one-bonddrive techniqueon each previouslyobtainedminimmn.Torsionalenergy surfaceswere obtained driving both R-CGC-CO (R = Me or m-Ru) every moleculefrom -180Q to +18OQ in

dihedralangles (~1

snd ~2)

presentin

159 steps.In the one-bonddrive rotations,10Q

steps were used for all bonds. All the conformersobtainedby MM2 calculations were used as startingconformations l4 under the MNDO approach.15Finally,a set of to be optimizedby the M3PAC program dipolemomentswas computedby using the formulafor the bulk dipolemoments: p = (2% ri2)l". The individualpi and ni for each conformerwere evaluatedas follows: a) With the MM2 program using three differentdielectricconstants:1.5 (vacumQ value routinelyused in MM2 calculations), 2.02 (cyclohezane) and 2.28 (benzene).The finalbulk dipolemcmtents so calculatedare gatheredin colons 1, 2 and 3 of Table 1. b) With the MNDO program with one single self consistencefield (1scF) and with full optimization. The calculatedbulk dipolemomentsare in cohz~s 5 and 6. c) By mixing the molar fractionsobtainedby MM2 calculations with the individual dipole moments (p i) originatedfrom the 1SCP calculationson the MNDO program over geometriesfrom M2. Column 4 of Table 1 containsthe bulk dipole mcments computedby this approach. RESULTS AND DISCUSSION.Table 1 contains the calculateddipole mcments for the six differentapproachesconsidered,as well as the ones determinedexperimentally by the methods of Guggenheim-smithand Halverstadt-KmKler. The root mean square (rms) statisticalcriterion has been adopted to decide which is the best computational approach.Predictionsby MM2 using a dielectricconstantof 1.5 are the best (rms = 0.25). This rms value shouldbe consideredas fairlygood, takinginto accountthat all but one of the compoundsstudiedhave low dipolemcmentsso that the relativeerrorsin measurementscan be high. The worst approachwas that based on full optimizationunder the MNDO program (nns = 1.19). Our discussionwill thereforebe based on the data of column1 (MM2 with a dielectricconstantof 1.5).

M.MORENO-MAT% etal.

6514

It is interestingto note that dipolemmnsntsdeterminedin cyclohexane(dielectric constant2.02) are in better agreementwith the MM2 calculations using the standard1.5 value correpond_iqtovacuum. Cur previous studiesoncompound~1andZ~weremadeusing older MM2 parameters.16The calculatedvalues for dipole moments with dielectric constants2.02 and 2.28 were very close to the values determinedin cyclohexaneand benzene, respectively.The values calculatedfor dielectricconstant1.5 were not as we preferredto use the more modern close to the experimentalfigures.Nevertheless, parametersin the presentstudy (Wl2(85)).

TABLE 1. Calculatedand experimental dipolemomentsfor ccqxnmds l-9 ----------Calculated

-Experimental-

Ita>

Z(b)

3(c)

463

s(e)

(j(f)

G&d

1

2.52

2.80

2.88

1.81

2.46

3.98

TiTi6)(j1&$4](j) $:;9](j))

2

4.06

4.78

4.95

2.73

3.7l

3.91

4.18 (4.38)(j)$+?6j(j) :??32+j)

3

2.39

2.56

1.46

1.80

3.67

2.01

1.97

4

1.90 2.57

2.84

2.94

2.03

2.32

3.99

2.48

2.45

2.41

5

2.12

2.48

2.65

1.62

2.16

3.26

2.09

2.06

1.99

6

2.28

2.71

2.85

1.69

1.93

3.68

2.06

2.02

1.94

7

2.80

2.45

2.24

3.45

2.27

2.22

2.13

1.95

3.15 2.06

3.28

8

2.11

1.56

3.46

3.61

2.42

2.40

2.36

2.84

2.82

2.82

2.02

2.92

2.94

2.80

2.77

2.71

0.25

0.49

0.60

0.92

0.39

1.19

C~

9 Y_Tl&

m(h)

&i)

1.88

aIW2 program, dielectricconstant 1.5. blW2 program, dielectricconstant 2.02; 'MM2 the molar fractions obtained by MM2 program, dielectric constant 2.28. kxing calculations with the individualdipolemcmentsoriginatedfrom the 1SCF calculations on the MNDO program ver geometriesfrom MM2. eMNDO programwith one singleself consistent ield (1SCF); PMNDO program with full optimization; gGuggenhei&mith method; method, the contributionof the atomic polarizationwas considered L lverstadt-Kunler method,the contributionj;i to be 5% of the electronicpolarization;Halverstadt-Kumler the atomic polarizationwas consideredt parenthesesvalues determinedin benzene;

Severelyhinde~d~diketones

6515

The Ml2 progrm has predictedsets of conformerscorrespond@ to energy minima. All the conformerspredictedcan be groupedinto pairs of enantianers.In other words, no confornwrspossessingagmnetryplaneshave been detected.Also no spmretryaxes are present in the predicted conformers.This means than neither entropy of mixing correction(+RTln2for u

pairs) nor correctionsof entropy for symmetry (-RTlns_, 2

being the order of the symmetryaxis in one given conformer)need to be introduced.5htropy differencesbetween conformers possessing the same synmetry (none in our predicted conformers) should be very small since in such situations the major contributions to entropy arise from groups of atawl

In other words, the differences

in conformational energiescalculatedby MM2 can be consideredequal to the differences in Gibbs free energies. The confonnational featuresfor each conformationcorresponding to aminimhave been describedin terms of dihedralangles d1 (c~-c&c~-c~)anda (c~-c~-G$-c~) and the angle between the two dipoles (gL). In Table 2 one enantioisaner of each pair is representedand the correspondingrelative energies, population (overall for both enantioisomers),r*l1 and UJ 2 (for the enantioisomershown) and a! (for both enantioiscmers) are given.

TABLE 2. Relativehergies Wal/mol), population(X), dihedralanglesu'l (Cl-C2-Ct~~ andu, (c2-C3-Ut-c5), angles between dipoles r( (0) and calculateddipolemanen pair is units>2 (W.2) for all conformers.Only one conformerof each enantioisomeric represented. The populationrefers to the contribution of both enantioisomers lA

Kcalfmol 0.00 46.03 x

@l

“2 r

o.:o” E 27.89 26:03

2A

2B

3A

0

12 12.78

4:-z 8% 5::: 52.96 -54.60 -97.98 53:74 63:97 -107.65 78.32 63.74 114.59 133.22 81.30 76.48 65.09 168.77 172.80 46.27 28.54 169.63 168.85 30.27 0.34 1.15 0.64 0.54 5.51 4.72 4.99 36.90

5B 5A 6c 6D 6A 1.19 Kcal/mol 0.00 0.E 0.76 0.93 x 75.74 10.10 4E 8.54 34.79 11.35 -64.03 112.32 -60:94 72.86 106.56 60.11 "1 -77.49 -64.66 -72.69 65.98 -63.47-109.91 d2 167.68 34.08 174.12 170.56 31.43 36.86 d 5.00 4.86 4.87 0.25 0.52 0.60 r 9B 9C 9A 0.02 Kcal/mol 0.00 32.88 31.70 1kZ % 42.36 22.76 -128:30 0'1 56.42 86.93 29.57 32 157.81 158.36 87.30 or 1.94 3.99 0.82 Y

7A 6;% 61:82 69.06 177.41 0.30

4A 4:z 36:30 80.64 168.74 1.23

4B 4c 0.31 0.34 27.41 26.12 50.97 -55.07 68.91 116.66 174.56 47.45 0.45 4.73

8A 0.Z 33.58 8t.E -59.68 -65:90 111.67 -65.91 39.26 174.40 0.21 4.76

6516

M.MORENO-WASTE

al.

From the above data some conclusionscan be drawn, but before we go into greater detail it shouldbe noted that all the measureddipolemcmentsexcept txo (c-s ad

2

9) are lower than or equal to the contributionof single a ketone carbonylgroup

(about2.5D). This means that the preferredconformations(for all compoundsexcept 2) are those in which minimizationof dipole repulsionis attained.The discussionto follow will be presented for only one enantioisomerof each pair. Only conformers populatedto a largerextentthan 10% will be considered(5% of each enantioisuner). The geometricdescriptions for all conformersdiscussedbelow are gatheredin Table 2. a)Compoundsl, 4 and9 Compounds1 and 4 possessingtertiaryradicalsat C3, exhibitvery similardipole moments (calculatedand measured).The conformations predictedare practicallythe same for both compounds:A and B having low dipole manenta accounting for 74% of the populationand a third conformation(C), exhibitinga dipole momentmuchhigherthan mixture.From these data it could be Z.SD, accountingfor 26% of the conformational concluded that the tert-butyland the 1-adamantylgroups are equivalent from the conformational point of view. &mpound 9 presentedmore difficultiesin the calculationspossiblybecause the congestionof groups (two vicinalquaternarycarbon atoms)reducesthe mobilityfor the rotationaround the e

bonds. It presentsa family of conformers(A,B) showinglow

dipolemomentsaccountingfor 65% of the total,in which the dipole repulsionis low as well as

one

conformer(C) accountingfor 20% and exhibitinga high dipole manent.

Thus, the conformational behaviourof 9 is similarto that of 1 and 4. b)Compounds3and5 These compoundsbear a secondarygroup at the intercarbonylposition and their behaviouris similar.Both measureddipolemanents are close to each other. Compound5 has one stable conformer(A) which has a low dipolemoment, accountingfor 76% of the conformational mixture.The balance is made up of differentconformations having high dipolemoments.One of them (B) is populatedto the extentof 10%. Diketone3 exhibits similarbehaviour:one low dipole moment conformer(A) represents87% of the mixture, and ahighdipole manent conformer(B) is presentto the extentof 13%. c)Compounds6and7 Both present similar behaviour,but compound 7 has local symmetry around the substituentat C3. It exhibitstwo highly populatedconformers.The low dipole moment conformer (A) accounts for 65% of the total and the high dipole moment one (B) is somehowless populated(34%).Since there is no local synmetryat the substituentat C3 in compound6, the conformersare not enantioisomers but diastereoiscaners. Two of them (A,B) representing76% of the conformationalmixture have low dipole moments, and correspondto conformerA in 7. The other two conformers(C,D) are populatedto the extentof 20% and have high dipolemoments.

6517

Severelyhindcred&diietones

d)Compound8 Only one

conformerwith a populationabove 5% is predictedfor 8 and that is

conformer(A) (80%).It exhibitsa low dipolemanent.Our calculatedbulk dipolemoment is much lower than that calculatedin the earlierwork byAllinger5 and-

closer to

t of the MMmethod since then. the experimental value,a reflectionof the improvemen e)

Ccwound2 The conformationalbehaviour of 2 is differentfrom those describedabove. Its

determined and calculateddipole moments are much higher than 2.5D. The tendency, already visible in 1, 4 and 9 is

clear here. This molecule has very strong steric

congestionand the best conformersare, therefore,those in which this is minimizedat the expense of the dipole repulsion which is now of lesser importance.The most populatedconformationis that having a highdipole manent (A, 54X), canparedwith the low dipolemoment one (B) which accountsfor only 46% of the conformational mixture. In sumwy,

there are basicallytwo differenttypes of conformersin diketonesl-9:

one haa low dipolemoments and the angle d higher dipole manents and angles d

rangesbetween157 and 178*; the secondhas

between 28 and 88'. In the second type of

conformations the groups at Cl and C5 (Me or t-Bu) are far fran the bulkiersubstituent As indicatedby the experimentaldipole moments, the contributionof these

at Q.

conformersto the populationincreasewith increasingbulk of the substituentsat C3. -9 r i+2.01D and 2.09D for secondaryradicals (canpowds 3 and 5) and 2.48D snd 2.5OD for tertiaryradicals (compounds4 and 1). If additionalstrain is introducedat positionsCl and C5, as in compound2 (r = 4.18D), conformationspossessinga high degree of dipole-dipole repulsioncan predominate.

Conformers

JLowDipoleMomentConformers

HighDipoleManent

lA, lB, ZB, 34, 48, 4B, 5A, 6A,

lC, 2A, 3B, bC, SB, 6C, 6D, 78, 9C

6B, 78, 86, 9A, 9B

6518

M.MoRENo-WAS~~~~.

a) Dipole moments: the solventsused for calibratingthe dielectriccell and for all the measurements(dielectricconstant,specificvoluw and refractiveindex)were frcm Carlo FrbaRPE andwere dried before use over Merck 41molecular sieves.&

dielectric

measurementswere performedon a WIW Model DK 06 Multidekameter, at a frequenq of 2.0 MHz. The cellused was made of silveredPyrex glass and was calibratedat the working temperature,25.OOf 0.02 QC using liquids of well known dielectricconstants(i.e., benzene,tolueneand cyclohexane).18*1g The differencesin refractiveindex of solutions and solventwere measuredat 546 mn in a Brice phoenix2000V differential refractometer, calibratedwith aqueousKCl. For specificVolga determinations an Anton Paar DMA 55 digitaldensimeterwas wed, with distilledwater and air as calibratingsubstances. The temperaturein the measuringcell was regulatedto 25.CO~O.01 QC. a) Preparationof comnounds3, 4, 5 and 12 3-(2-Mamantyl)pentane-2,4-dione, 3. A mixture of 2-brcewadamantane (1.075g, 5.0 moole) and

Co(acac)~

(0.645 g, 2.5

mnole) in 1,1,2,2-tetrachloroethane (1.5 ml) was heated at 185QC for 90 minutes in a closed reactor. The mixture was partitionedbetween dichloranethane and 10% Hcl. The organic layer was washed,dried and evaporated.the residuewas treatedovernight with active charcoal in chloroformat rocm temperature.The mixture was filteredand the solventevaporated.The residuewas chromatographed throughsilicagel with mixturesof hexaw-ethyl

acetate to

give

recovered 2-braooedamantane,tracea of

2-

acetonyladmnantane, 2-adamant+ acetate (190 mg) snd 3 (270 mg, 21 X); m.p. 76-77.5*C (ether-petroleum ether);Ik(kBr):1694 cm-'; PMR(CDC~~):6 1.43-1.90(m, 1410, 2.18 (s, 6H), 2.63 (broadd, J 12.2 Hz, lH), 4.24 (d, J 12.2 Hz, 1H); MS: m/e 234(&l, 2), 43(100). Caled. for Cl5H22O2:C, 76.88;H, 9.46. Found:C, 76.55;H, 9.46. 3-t-Butylpentane-2,4-dione, 4. A mixture of Co(acac)2(1.29 g, 5 mnole), t-butyl iodide (1.85 g, 10 mnol) and chloroform(2 ml) was heated at 1COQC for 4 h in a closed reactor.The mixture was partitionedbetweendichloromethane and 10% HCl. The organiclayerwas washedwith water and eventuallywith aqueous sodium thiosulfate,dried and evaporated.The residuewas taken up in chloroformand treatedovernightwith activatedcharcoal.The mixturewas filteredand the solventevaporatedto give 236 mg (15%) of diketone4 as an oil; bp 350~ (oven temperature)/O.Ol ~@g; INfilm): 1721, 1697 cm-'; PWcDc13): 6 1.08(s,9H), 2.20(s, 6H), 3.65(s,lH), these data are coincidentwith those previouslyreported;20; MS: m/e 15704+1,ll), 48(23),47(22),43(100). 3-(2-Cyclohexewl-yl)pentane-2,4-dione, 10. A mixture of Co(acac)2(12.90g, 0.05 mole) and 3-brcmocyclohexene (9.73 g, 0.06 mole) in chloroform(20 ml) was heated at 1OOQC for 4 hours in a closed reactor.After cooling, the mixture was partitionedbetween hexane and 10% hydrochloricacid. The

Severelyhinde~~-diketones

6519

organiclayerwaswashad, driedand evaporatedto afford a residueuhichwas takeninto chloroformand treatedwith activatedcharcoalovernight.The mixturewas filtered,the solvent evaporatedand the residue distilled (bp 97QC/O.O2Q~QH~)to give 10 (6.74 g 75%); IR(film):1723, 1699 an-l; PMR(CDC13):d 1.12-2.17(m, 6H), 2.22 (a, 6H), 2.813.23 (m, lH), 3.61 (d, J 10.3 Hz, lH), 5.35 (long range coupledd, J 9.3 Hx, lH), 5.78 (m, 1H); cMR(CDC13): d 20.4, 24.7, 26.4, 29.4, 29.8, 35.4, 74.5, 127.0, 129.7, 203.3, 203.6;MS: m/e 18O(M,l), 137(55),43(1CO). 3-Cyclohexylpentane-2,4-dione, 5. A mixture of 10 (315 mg, 1.75 Qswle) and a catalyticamount of 10% Pd/C in ethanol (25 ml) was shakedwith hydrogenat atmosphericpressure.After 2 hours the mixturewas filteredand the solventevaporated to afford250 mg (79%)of 5; m.p. 47-8QC (petroleum ether); IR(RBr):1727, 1694 cm-l; PMR(CDC13):d 1.10-1.80(m, llH), 2.16 (8, 6H), 3.50 (d, J = 10.9 Hx, la); MS: m/e 183(M+l,17), 101(52),97(24),43(100). Calcd. for CllH1802:C, 72.49;H, 9.96. Found:C, 72.39;II,10.12. 3-Acetyl-4-methyl-5-hepten-2-one, 11. A mixture of 4-brano-2-pentene (preparedaccordingto Ref. 21) (1.60 8, 0.011 mole), Co(acac)2(1.38 g, 0.005 mole) and chloroform(2 ml) was heated at 1OOQC for 2 hours in a closed reactor. The colour changed from pink to green. The mixture was partitioned betweenchloroformand 1N HCl. The organiclayerwas dried and evaporatedto afford a residue (2 g, nearly 100% yield) of spectroscopically pure 11 which was distilled (40-5QC/0.5nmHg) to afford pure 11 (58%); IR(film):1723, 1699, 970 cm-'; PMR(CEC13): d 0.98(d,J 7.7 Hz, 3H), 1.63 (d, J 5.1 Hx, 3H), 2.12 (s, 3H), 2.20 (8, 3H), 2.80-3.22(m, lH), 3.56 (d, J 10.3 Hx, lH), 5.06-5.77(m, W); m(CDCl3):

C 17.61,

18.77,29.34, 29.84,37.42, 75.60,126.14,132.31,203.56;MS: m/e 43(100);MS (chemical ionization): m/e 186(M + 18, 13). Calcd. for C-&602: C, 7l.39;H, 9.59. Found:C, 71.49;H, 9.83. 3-Acetyl-4-methyl-2-heptanone, 12. A mixture of 11 (0.50 g, 2.9 mnole) and 10% PalladiuQon charcoal (0.05 g) in absoluteethanol (25 ml) was shaked in hydrogenat atmosphericpressure.After 3 hours the mixture was filtered through celite and the solvent evaporated to afford 0.38g (78%) of 11 which was distilled(5OQC/O.O7mnHg). The distilledcompound(0.26 g, 53%) presentedIR(film):1722, 1698 cm-'; pMR(cDc13): d 0.6-1.9 (m, lOH), 2.2 (8, 6~), 3.60 (d, J 10.7 Hz, 1H); CMR(CRC13):s

13.94, 16.88, 19.65, 29.55, 29.61, 33.54, 36.68,

76.48,204.31,204.36. Calcd. for ClOHl8C2:C, 70.55;H, 10.66.Found:C, 70.36;H. 10.95.

Financialsupportfrom DGICYT throughprojectPR87-0030is gratefullyacknowledged. We are also indebted to "ConselhoNational de Pesquisa"(Brazil)for a postdoctoral scholarshipto one of us

(A.C.Siani).

6520

M.MORENO-MANASC~~~.

REEmmcEsANDm shouldbe addressed l.- Towhomcorrespondenceconcerningcalculations 2.- Burkert,U.; Allinger,N.L. "MolecularMechanics".ACS Monograph177; Washington, D.C., 1982. 3.- Moreno-Ma?ias, M.; Gonz&ez, A.; Jaime, C.; Marquet, J.; HernBndez-Fumes, I.; Salom,c.; Bellanato,J. J. Chew Sot., Chem. Cumnm. 1987, 1706. 4.- a)

Saiz,

E.; Horta, A.; Gargallo,L.; Hern&dez-Fuentes,I.; AbradelO,

C.;

RadiC.

D. J. Chem. Res. (S) 1988, 280; b) Hen&dez-F&ntes, I.; Abradelo,C.; Domfngwz, C .i cs&y, A.G.; Plmet, J.; Cativiela,C.; Mayoral, A.; Gaset, A; Rigal, L. Heterocycles1989, 29, 657; c) Herrero,A.; Ochoa, C.; Stud, M.; Florencio,F.; Her&&z-Fuentes, I.; Abradelo,C.; De Paz, J.L.C. J. Org. them. 1989, 2,

5025;

d) Claratmnt,R.M.; Elguero,J.; Fabre,M.J.; Faces-Faces, C.; Hert'&dez-Cano, F.; Hern&dez-Fuentes,I.; Jaime C.; tipez,C. Tetrahedron1989, 45, 7805; e) LIIIILbrOSO, H .; Bertin, D.M.; Olivato, P.R.; Bonfada, e.; Madino, M.G.; Hase, Y. J.

~010

stnlct.

1989, 2l.2,113. 5.- Allinger, N.L.; DosewMifovi&, L.; Viskocil, Jr., J.F.; Thaaas Tribble, M. Tetrahedron1978,2, 3395. 6.- a) Gonzglez,A.; Marquet, J.; Moreno-M&m, M. Tetrahedron1986, 42, 4253; b) Manpet,

J.; Moreno-Ma&w,M. Synthesis1979, 348.

7.- Lloris,M.E.; Marquet,J.; Moreno-M&as,M. . 1990, 3I, 7489. 8.- MarQuet,J.; Moreno-Ma&s, M.; Pacheco,P.; Vallribera,A. Tetrahedronl&t. 1988, 29, 1465. 9.- a) Guggenheim,E.A. Trans. FaradaySot. 1949, 45, 714; b) Guggenheim,E.A. Trans. Far&y

SIX. 1951,47, 573; c) Smith,J.W. Trans.

FaradaySot. 1950, 5, 394.

lo.- Halverstadt,I.F.;Kmler, W.D. J. Am. Chem. Sot. 1942, 64, 2988. ll.- Allinger,N.L.; Yuh, Y. Quanta

ChemiStrY

Pr0@311

FXChaWe

1980, 12, 395.

12.- Allinger,N.L. J. Am. Chem. Sot. 1977,99, 8127. 13.- MM2(85)Parameters;Allinger,N.L., personalccmunication. 14.- Stewart,J.J.P.QuantumChemistryProgramExchanne1987, 455. 15.- Dewar,M.J.S.;Thiel,W. J. Am. C&em. Sot. 1977, 99, 1977. 16.- OriginalMM2(77)parameters:Ref. 11. 17.- BCTISOII,

s.w.;

Cruickshank, F.R.; Golden,D.M.; Haugen,G.R.; O'Neal,H.E.; Rodgers,

A-S.; Shaw, R.; Walsh,R. Chem.Rev.1969,3, 279. 18.- Petro,A.J.; Smith,G.P. J. Am. Chem. Sot. 1957,7J, 6142. 19.- Heston,W.M.; Smith,G.P. J. Am. Chem. Sac. 1950, 72, 92. 20.- Boldt,P.; Militzer,H. TetrahedronL&t. 1966, 3599. 21.- Mulliken,S.P.;J. Am. Chem. Sot. 1935, 57, 1605.