Stereospecific ring opening of cyclopropanes : a route to functionalized medium sized rings

Stereospecific ring opening of cyclopropanes : a route to

functionalized medium sized rings

Citation for published version (APA):

Loozen, H. J. J. (1976). Stereospecific ring opening of cyclopropanes : a route to functionalized medium sized rings. Technische Hogeschool Eindhoven. https://doi.org/10.6100/IR144464

DOI:

10.6100/IR144464

Document status and date: Published: 01/01/1976

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

providing details and we will investigate your claim.

STEREOSPECIFIC

RING OPENING

OF CYCLOPROPANES

·

A ROUTE TO FUNCTIONALIZED

MEDIUM SIZED RINGS

STEREOSPECIFIC RING OPENING

OF CYCLOPROPANES

A ROUTE TO FUNCTIONALIZED MEDIUM

SIZED RINGS

PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE TECHNISCHE WETENSCHAPPEN AAN DE TECHNISCHE HOGESCHOOL EINDHOVEN, OP GEZAG VAN DE RECTOR MAGNIFICUS, PROF. DR. P.VAN DER LEEDEN, VOOR EEN COMMISSIE AANGEWEZEN DOOR HET COLLEGE VAN DEKANEN IN HET OPENBAAR TE VERDEDIGEN OP

DINSDAG 26 OKTOBER 1976 TE 16.00 UUR

DOOR

HUBERT

LOOZEN

DIT PROEFSCHRIFT IS GOEDGEKEURD DOOR DE PROHOTOREN

PROF.DR. H.H. BUCK EN

Aan mvn ouders

Aan Fia

CONTENTS

In troduction

Chapter

I

Chapter IJ

Ch-zpter JIJ

Solvolytic ring expansions

I-1 Mechanistic backgrounds of the sal-volysis of cyclopropyl derivatives I-2 Silver perchlorate promoted ring

ex-pansions of some geminal dibromo-cyclopropanes

I-3 Experimental sectien References and notes

7 12

Non-solvolytic ring expansions 25

II-1 Silver tosylate promoted ring enlarge-ment of geminal dibromocyclopropanes II-2 Silver nitrate promoted ring

expan-sions of geminal dibrornocyclopropanes II-3 Influences of nucleophilicity and

ring size in the aleoholysis of sorne dibrornocyclopropanes

II-4 Structure assignments II-5 Experimental section

Raferences and notes

Stereosele~tive synthesis of gerninal endo-iodo-exo-bromo-cyclopropanes

III-1 Introduetion

III-2 Iodination of endo-lithio-exo-bromo-cyclopropanes

Chapter IV

Summary

Samenvatting

III-3 Silver fluoride as an effective

catalyst in ring expansions of geminal dihalocyclopropanes

III-4 Experimental section References and notes

Bishomoaromatic interaction in the disrota- 67 tory ring opening of cyclopropyl carbenoids

IV-1 Introduetion

IV-2 Lithiation of geminal dibromocyclo-propanes

IV-3 Backgrounds of the ring opening of cycloprópylidenes

IV-4 Behaviour of cyclopropylidenes attached to polyenes

IV-5 Experimental section References and notes

94

96

Curriculum Vitae

98

INTRODUCTION

The important influence of steroids in the dynamics of the living organism has been accepted now for several decades. An important and fascinating achievement in chemistry was the elucidation of the metabolic pathway by which steroids are built up from the simple precursor acetic acid.

--N

_y~~oP

₂

_06-3 Geranyl- ~ pyrophosphate ~

---

----""-, Farnesylpyrophosphate Squalene Lanosterol

In the biogenesis of steroids, especially the conversion of squalene into the tetracyclic larrosterol constitutes a problem of outstanding interest. In this polycyclization reaction only one tetracyclic product is formed from an a-chiral precursor. It was assumed that an enzyme folds the squalene molecule in such a conformation that the cyclization results in one stereo-isomer only (the tetracyclic lanosterol). Experiments under non-enzymatic conditions showed that the methyl substitution pattern in squalene determines the formation of carbenium ion centers during the cyclization and therefore is respon-sible for the formation of products having a five-membered C-ring. This contrasts with the formation of a six-membered C-ring exclusively in the enzymatle reaction. The stereo-chemistry of larrosterol determines the mode of folding for 2,3-epoxysqualene (first product formed before the cyclization to lanosterol). From model studies it can be deduced that a chair-boat-chair folding must be involved (leading references have been cited below1_- 4

) .

\::'---

~~

...

---·~

l,

'

0 )

...

,/'

In order to gain more insight in the influences of conformation on the nature of the products originati~g from cyclization re-actions, the synthesis of molecules which contain a part of the squalene chain in a fixed conformation, was started.

n

=

2,3 R:: H,C~

Preparatien of nine- or ten-membered rings (n=2 and 3 respec-tively) containing two trans double bands and an unsaturated C-5 side chain at an allylic position seemed to be an attrac-tive approach for the aims above. At this stage attention was focused on the typical chemistry of medium sized rings.

In a relatively simple straight forward sequence such a molecule, containing a nine-membered basic skeleton, was made accessible in the following way5

• . H Br

c:x·~

AgClOt.

---

_5% aq. acetone tosy !chloride

~H

~···oH

---

pyridine, 0° 8r - 5%-H:!SDt_-Clfs OH Buli, THF;-1.5°

~H

~···oros Br 2,1. ,l. -trimethyl-oxa-zoline, Buli,THF. -70"

However, several fundamental problems remained unclear in such an approach. Reduction of bramine from a trans double bond with retention of contiguration proved to be rather cumbersome. Purthermare the control of stereochemistry was lost completely in the reaction of the oxazoline with the 2-bromocyclonona-1,6-dien-3-yltosylate, whereas the eluci-dation of the stereochemistry with the aid of pure spectros-copie techniques seemed to be hampered by the fact that we are dealing with a rather scarcely explored area of these

particular cyclic systems. Best outlook for the final achieve-ment of a straight forward synthesis seemed to be incorpora-ted in the development of new synthetic methods in this field, which allow the facile and stereocontrolled introduetion of

interesting functionality and which provide a profound in-sight in the elementary nature of reactivity in cyclic struc-tures.

Rq-erences

1. W.S. JOHNSON, Acc. Chem. Res.,

l•

1 (1968)

2. E.E. van TAMELEN, Acc. Chem. Res., ~. 152 (1975)

3. D. ARIGONI, Pure Appl. Chem. (IUPAC),

!l,

219 (1975)

4. E.E. van TAMELEN, Acc. Chem. Res.,

l•

111 (1968)

CHAPTER I

solvolytic nng expansions

I-1 Meohanistic backgrounds of the solvolysis of

cyclopropyl derivatives

One of the most used ways in constructing medium sized rings consists of rupture of the bond between the bridge-head carbon atoms of bicyclic systems1_{• In this way cyclic}

structures containing n+3 carbon atoms can be obtained from bicyclo[ n. 1. O] al kanes.

The property of a variety of cyclopropyl systems to undergo a- rapid solvolytic ring opening to allylic derivatives is commonly used as the basic element of such ring expansions. The presence of a good leaving group at Cn+ 3 of the cyclo-propane part of the molecule is a prerequisite.

As such monohalo- and geminal dihalocyclopropanes represent a favourite class of compounds for effecting such ring expansions; the more so because they are readily accessible.

x

"'0<-~

.

Rt.

Solv

The arrangement of the substituents attached to the double bond is, of course, dependent upon the mode of ring opening of the cyclopropyl cation.

Since the announcement of Woodward and Hoffmann's rules of conservation of orbital symmetry2_• 3_{, several profound}

investigations concerning the characteristic elements of the solvolytic ring opening of cyclopropane derivatives appeared in the literature4 13

, Solvolytic rate measurements of a series of substituted cyclopropyl halides and cyclopropyl tosylates were indicative for the following mechanistic features.

1. Ring opening proceeds in a disrotatory manner, as can be predicted from Woodward-Hoffmann rules for a two electron system (cyclopropyl cation).

2. Substituents which are arranged cis to the leaving group rotate inward and those which are trans to the leaving group rotate outward.

This latter condition implicates that the departure of the leaving group and the ring opening praeeed concertedly and not as separate processes.

-x (exol

..!:)-k, :--1~

"A

From these considerations it becomes clear that the ring expansion of ais-bicyclo[n.1.0lalkanes may lead to cyclic

olefins, containing n+3 carbon atoms, exhibiting a cis or a trans geometry, depending on the original orientation of the leaving group. -x ( endo)

_______

,_

Solv

~.,H

~Solv

_______

,..,.. Solv

When the exo group is lost the final geometry will be trans. Endo leaving results in a cis arrangement. This solvolytic behaviour of cyclopropanes, annulated to cyclic structures, has received wide spread experimental support1_~-28 _•

It has been established quite recently 29_- 35 _{that silver}

perchlorate assisted solvolytic ring opening of geminal dibromocyclopropanes, annulated to seven-, eight- and nine membered rings, proceeds with loss of the more accessible exo bromine atom. The products arising from such reactions should necessarily have the trans geometry.

A further remarkable selectivity in this ring opening reaction results from the fact that the nucleophile enters concertedly at the same side of the incipient allylic system from which the exo halogen atom is released, thus leading to one single diastereoisomer only. This latter phenomenon has been observed befare in the salvolysis studies performed by Parham et al· on 1,1-dichloro-2,3-dipropylcyclopropanes11

• Schematically this can be visualized as follows:

Ag+

r

2) Solv

It should be mentioned that the endo halogen atom is lost preferentially in those compounds where the expanded ring would become too small to accommodate a trans double bond. Then cis compounds are formed as a matter of course36

,

As an illustrative example the different rates of the

salvolysis of 7-endo-chloronorcarane, 7-exo-chloronorcarane and 7,7-dichloronorcarane are given:

~l

~\

H

k=1.4x1o- 6 sec- 1 k=4.5x10 -7 sec 1

In order to elaborate a well-set up synthesis, leading to the target compounds, mentioned in the introduction, the develop-ment of novel approaches for the introduetion of interesting

functionality with control of stereochemistry was required. As such an intensive examinatien of ring expansions in medium sized rings seemed to be a logical beginning.

I-2 SiZveP pepchZorate promoted ring expansions of

some geminaZ dibromocycZopropanes

As substrates for the silver ion catalysed ring expansion reactions the dibromides 1-4 were selected. They were readily accessible by the reaction of the corresponding olefins with dibromocarbene according to the original procedure of Doering

and Hoffmann37• H Br

(j:'

CD

H Br

~~

®

r-~r

CJ'~Br

These compounds, upon reaction with an excess of silver perchlorate, according to the method of Reeseet aZ.29 _,

afforded the trans allylic alcohols 5-8 as the main products.

cqc

.

cq<

.

~"

~OH '()H

-Br H Br Br

G)

(i)

G)

16

Whereas the ring opening of 1 and 3 afforded only S and 7 in a smooth reaction, the ring opening of 2 and of 4 gave besides the expected products 6 and 8 reproducible amounts of a side product (ca. 20% in both cases). These products could be separated by column chromatography. Their micro-analyses proved to be consistent with those of the concomi tantly formed trans alcohols. The possibility that these by-products might be the trans isoroers 9 and 10 could be ruled out easily. It has been established that both

trans-cyclonenene and trans-cyclodecene (which are optically active molecules) are optically labile at room temperature38

•39 •

The explanation for this behaviour is based on the fact that the olefinic hydragen can rotate through the loop of the ring with a relatively low activation energy:

~ E "' 20 kcal/mol

( Y l

...

oH

~

t: E n.11kcal/mol

@)

Br

From these values it can be seen that bath 6 and 8 exist at room temperature as two rapidly equilibrating diastereoiso-mers. This intra molecular isomerization process can be re-cognized by nmr. The 1H-nmr spectrum displays a double doublet at ê 4.02 (J=11 and 4 Hz) and a multiplet at ê 4.54 for the methine protons, and two double doublets at ê 6.03 (J=10 and

6 Hz) and at,ê 6.21 (J=10 and 7 Hz) for the olefinic proton. For the equilibrium 8 t 10 the activation energy is so small that at room temperature 8 and 10 are not distinguishable.

The spectrum displays only one double doublet for the methine proton at 6 4.17 10 and 5 Hz) and a triplet at 6 6.26

(J= 8 Hz) for the olefinic hydrogen. With all these conside-rations in mind the side products from the ring opening of 2 and of 4 must be the cis alcohols 11 and 12 respectively.

®

~OH

~Br

®

From the back-ground hitherto known, there are two possibili-ties to explain the formation of such cis products. The first possibility, an isomerization of the diastereoisomerie mixture of 6 and 8 under the reaction conditions, could be ruled out easily. Stirring of 6 and 8 with silver perchlorate in aqueous acetone did not yield 11 and 12 respectively. Another possibi-lity is based on mechanistic considerations. In case that besides the exo halogen also the endo halogen is expelled in 2 and 4, then 11 and 12 would be formed:

AgCl04

---·

_{5o;. aq.acetone}

This explanation seems, for the moment, not unrealistic, because it has been demonstrated that

9-endo-bromobicyclo-[6.1.0)nonane actually undergoes, though very slowly, ring opening with silver perchlorate29_{• As such the ring expansion} of. 2 and 4 might be regarded as an ordinary competition of "exo leaving" VB. "endo leaving".

~

Table I

Salvolysis of geminal dibromocyclopropanes with AgCl04 in 5% aqueous acetone1

Substroles Products' Mp, °C 'field,% • _{NMR dato (CDCI 3), 6 values}

(j~·

_o::x

76 4,17 (dd, 1, methine H, J=9 and 6 Ez)

tlH 6 • 11 ( dd, 1 , o 1 e fin E, J = 11 and 5 Hz) G) @ Sr

dZ~

OH 11: 4.75 (m, 1, methine Hl,

CQ<

~r

72-742 6.18 (t, 1, alefin H, J=11 Hz) 73 6: 4.02 (dd, methine H, J=8 and 4 Hz) H (cis) 4.55 (m, methine H)

0

@72 Br @ 28 6.03 (dd, alefin H, J=10 and 6 Hz) 6.40 (dd, alefin H, J=10 and 7 Hz)

d

Cq<

4.16 (dd, 1, methine H, J=10 and 5 Hz) \ "Br ' 84-86' 65 5.37 (m, 2, alefin H) H 5.87 (t, 1, alefin H, J=7.5 Hz) @

cv

Br 12: 4.82 (t, 1, methine H, J=8 Hz)

(\)(

c::c

5.98 (dd, 1, alefin H, J=12 and 6 Hz)

d

Br _{48-so• 71 8: 4.42 (dd, 1, methine H, J=9 and 4 Hz)}

(cis)

6.23 (t, 1, alefin H, J=7.5 Hz)

OH

G) @77 Br @ 23

l Reactions were carried out with initial 1 molar concentrations of silver perchlorate. 2) Crystallized from petroleum ether (60/80). 3) Crystallized from diisopropyl ether. 4) Yields arebasedon crude isolated rnatorials (purity > 95~). 5) Satisfactory analytica! data were obtained for the new products

I-3 Experimental seation

General. Starting materials 1-4 were prepared by a

modification of the classica! procedure from the corresponding olefins and dibromocarbene•0 • Cyclononene, required for the preparation of 4 was obtained from 2 by conversion with methyl-lithium in ether to the cyclonona-1,2-diene and subsequent reduction with sodium in ammonia-1• The solvolytic ring

expan-sion reaction is exemplified with the salvolysis of 2 and 4. A survey of the products obtained in the salvolysis reactions of 1-4 is given in Table I.

2-Bromo-3-hydroxycyclonon-1-ene (6, 11)

To a solution of 5.64 g (0.02 mol) of 2 in 50 ml of 5% aqueous acetone was added a solution of 5.36 g (0.026 mol) of silver perchlorate in 25 ml of 5% aqueous acetone. The mixture was stirred for 1 hr at ambient temperature and monitored with tlc. After addition of 50 ml of saturated sodium chloride solution stirring was prolonged for an additional 5 min. The precipitate was filtered. The filtrate was diluted with 200 ml of water and extracted with ether. Upon washing, drying and evaporation of the organic phase 3.19 g (73%) of the product was left as a colorless oil. The product consisted of two components. Chromatography (silica gel; chloroform-2% methanol as eluent) afforded 0.89 g of the cis alcohol 11

(Rf 0.29) and 2.29 g of the diastereoisomerie mixture of trans alcohols 6 and 9 (Rf 0.22). The cis alcohol was recrystallized from petroleum ether: mp 72-74°; nmr (CDC1₃)

o

4.75 (m, 1, methine H), 6.18 (t, 1, olefin H, J= 11Hz). Anal. Calcd for C

9H15Br0: C, 49.31; H, 6.84. Found: C, 49.29; H, 6.80. 2-Bromo-3-hydroxycyclodec-1-ene (8, 12)

A solution of 5.92 g (0.02 mol) of 4 and 5.36 g (0.026 mol) of silver perchlorate in 75 ml of 5% aqueous acetone was stirred for 3 hr at room temperature. Work-up in a similar manner as described above afforded 3.21 g (71%) of the product.

Chromatography over silica gel (chloroform-2% methanol) as eluent afforded 2.55 g of trans alcohol 8 (Rf 0.38) as a colourless oil: nmr (CDC1

3) ó 4.18 (dd, 1, methine H, J= 9 and 5 Hz), 6.24 (t, 1, alefin H, J= 9 Hz). The other com-ponent (0.75 g, Rf 0.44) was the cis alcohol 12 mp 48-50°

(petroleum ether); nmr (CDC1

3)

o

4.82 (t, 1, methine H, J= 8 Hz), 5.98 (dd, 1, alefin H, J= 12 and 6 Hz). Anal. Calcd for

c

10H17Br0: C, 51.50; H, 7.29. Found: C, 51.62; H, 7.31.

Re_[erences and notes

1. C.D. GUTSCHE and D. REDMORE, "Advances in Alicyclic Chemistry"; Carbocyclic Ring Expansion Reactions, Academie Press, New York, 1968

2. R.B. WOODWARD and R. HOFFMANN, J. Amer. Chem. Soc., 8 ' 395 (1965)

3. R.B. WOODWARD and R. HOFFMANN, "The Conservation of Orbital Symmetry", Verlag Chemie G.m.b.H., Weinheim,

1970

4. C.H. DE PUY, L.G. SNACK, J.W. HAUSSERand W. WIEDEMANN, J. Amer. Chem. Soc.,

JU,

400.6 (1965)

5. L. SKATTEB~L, J. Org. Chem.,

ll•

1554 (1966)

6. C.H. DE PUY, L.G. SNACK and J.W. HAUSSER, J. Amer. Chem. Soc., , 3343 (1966)

7. P.v.R. SCHLEYER, G.W. VAN DINE, U. SCHÖLLKOPF and J.PAUST, J. Amer. Chem. Soc.,~. 2868 (1966)

8. J.W. HAUSSERand N.J. PINKOWSKI, J. Amer. Chem. Soc.,

~. 6981 (1967)

9. W.E. PARHAM, K.S .. YONG, J. Org. Chem., ~. 3947 (1968) 10. J.A. LANDGREBE and L.W. BECKER, J. Org. Chem., 33, 1173,

(1968)

11. W.E. PARHAM and K.S. YONG. J. Org. Chem., ~. 683 (1970) 12. P.v.R. SCHLEYER, W.F. SLIWINSKI, G.W. VAN DINE,

U. SCHÖLLKOPF, J. PAUST, K. FELLENBERGER, J. Amer. Chem. Soc.,

Qi,

125 (1972)

13. W.F. SLIWINSKI, T.M. SU and P.v.R. SCHLEYER, J. Amer. Chem. Soc., 94, 133 (1972

14. S.J. CRISTOL, R.M. SEQUEIRA, C.H. DE PUY, J. Amer. Chem. Soc., !U_, 4007 (1972)

16. U. SCHÖLLKOPF, K. PELLENBERGERand M. PATSCH, Tetr. Lett.,

3639 (1967)

17. M.S. BAIRD and C.B. REESE, Tetr. Lett., 1379 (1967) 18. W. PARHAM and R.J. SPERLEY, J. Org. Chèm., ~. 924

(1967)

19. W.E. PARHAM and R.J. SPERLEY, J. Org. Chem., ~. 926 (1967)

20. C.W. JEFFORD and W. WOJNAROWSKI, Tetr. Lett., 199 (1968) 21. W.E. PARHAM, F.M. PARHAM, J.F. DOOLEY and M.K. MEILAHN,

J. Org. Chem., ~. 3651 q968)

22. D.T. CLARK and G. SMALE, Chem. Comm., 868 (1969)

23. M.S. BAIRD and CrB. REESE, J. Chem. Soc. (C), 1808 (1969) 24. M.S. BAIRD, D.G. LINDSAY and C.B. REESE, J. Chem. Soc.

(C), 1173 (1969)

25. D.B. LEDLIE and E.A. NELSON, Tetr. Lett., 1175 (1969) 26. S.R. SANDLER, J. Org. Chem. , _{~. 3876 (1966)}

27. S.R. SANDLER and P.S. SKELL, J. Amer. Chem. Soc., ~. 2024 (1958)

28. E.E. SCHWEIZER and W.E. PARHAM, J. Amer. Chem. Soc.,~.

4085 (1960)

29. C.B. REESE and A. SHAW, J. Amer. Chem. Soc., 2 2566 (1970)

30. C.B. REESE and A. SHAW, Chem. Comm., 1365 (1970) 31. C.B. REESE and A. SHAW, Chem. Comm., 1367 (1970)

32. D. DUFFIN and J.K. SUTHERLAND, Chem. Comm., 626 (1970) 33. M.S. BAIRD and C.B. REESE, Tetr. Lett., 4637 (1971) 34. G.H. WHITHAM and M. WRIGHT, J. Chem. Soc. (C), 1173

(1969)

35. _{C.B. REESE and A. SHAW, J. Chem. Soc., ( Perkin I) '} 2422 (1975)

36. P.M. WARNER, R.C. LA ROSE, R.F. PALMER,

c.

LEE, D.O. ROSS and J.C. CLARDY, J. Amer. Chem. Soc. , 9 7,

5507 (1975). See also references _{1 2 ' 1 3' 1} 4, _{26' 27}

and 28

37. W.E. DOERING and A.K. HOFFMANN, J. Amer. Chem. Soc.,

~. 6162 (1954)

38. A.C. COPE, K. BANHOLZER, H. KELLER, B.A. PAWSON, J.J. WHANG and H.J.S. WINKLER, J. Amer. Chem. Soc.,

!I,

3644

(1965)

39. G. BINSCH and J.D. ROBERTS, J. Amer. Chem. Soc.,

!I•

5157 (1965)

40. L. SKATTEB~L, Acta Chem. Scand.,

!l,

1683 (1963) 41. P.D. GARDNER and M. NARAYANA, J. Org. Chem., ~' 3518

(1961)

CHAPTER IJ

Non-solvolytic rzng expansions

II-1 SiZver tosyZate promoted ring enZargement of

geminaZ dibromoayaZopropanes

The reactions described in the preceding section are always performed by reacting halocyclopropanes, with or without silver salts, in solvents such as methanol, ethanol, aqueous acetone or acetic acid. Therefore they may be classi-fied as solvolytic ring opening reactions1_{• The role of the}

silver ion consists in facilitating the formation of a cyclopropyl cation by withdrawal of the halogen atom. Based on steric considerations it is self-evident that complexation will occur preferentially at the exo iite. The products ari-sing from such solvolytic ring opening reactions are generally allyl ethers, allyl alcohols or allyl esters when these re-actions are carried out in alcohols, aqueous solvents or acids respectively. With regard to the synthesis of the target com-pounds the question was raised if the scope of this reaction could be extended such that interesting functionality could be introduced at the allylic position in one step and with control of stereochemistry. The use of silver tosylate in a neutral solvent seemed to be a good choice2_{-~. The silver ion}

induces the ring opening whereby the intermediate allylic cation should be captured by the tosylate anion; thus giving a direct entry to allylic tosylates. The substrates 1-4 were used again as representative model compounds.

H Br

d'"'

~c

~IBr

CD

0

H Br

c:i"'

. H Br

c:Jt:'Bc

0

₀

Upon treatment of 1 and 3 with two equivalents of silver to-sylate in refluxing acetonitrile trans-2-bromo-3-tosyloxy-cyclooct-1-ene (13) and trans,ais-2-bromo-3-tosyloxycyclo-nona-1,6-diene (14) were formed respectively.

@

_®

The configurations of these tosylates could be confirmed easily by independent synthesis from the corresponding aleo-hals with tosyl chloride in pyridine. It is obvious that the formation of 13 and 14 underlies the samebasic features as the solvolytic reactions outlined in Chapter I. Again the ring expansion of 2 and 4 with silver tosylate afforded ring expanded tosylates. However, they proved to have cis double bands. Via synthesis from the corresponding alcohols they were identified as ais-2-bromo-3-tosyloxycyclonon-1-ene (15) and ais-2-bromo-3-tosyloxycyclodec-1-ene (16) respectively. The formation of exclusively cis products from the ring opening of 2 and 4 is rather curieus. They are nat formed by isomerization from the trans tosylates 17 and 18.

®

This could be checked easily by control experiments. Formation of 15 and 16 by expelling the endo bramine atom from 2 and 4 is highly improbable, as there is no reason that complexing of the exo bramine would become unfavourable. Conclusive evidence is obtained from the fact that

9-endo-bromobicyclo[6.1.0)nonane (19) is unreactive towards silver tosylate5

•

Obviously the ring opening of 2 and 4 leads to a strained transient trans cation which isomerizes rapidly to a ~is

cation befare reacting with the weakly nucleöphilic tosylate anion. In contrast, the ring opening of 1 and 3 leads ëxclu-sively to trans tosylates 13 and 14 because full development of a free trans cation would require a too unfavourable geometry. Therefore these reactions praeeed completely con-certed (a survey of the tosylates is given in Table II). The model description presented above seems to be quite "re-alistic, but it deserves to be supported by more experimen-tal evidence. Therefore a more profound investigation was undertaken.

Table II

React10n of gem,nal dibromoc_yclopropanes with AgOTos m acetonitrile

Substroles Produels l Mp,°C Yîeld.~ó 2 NMR dol a _{I CDCt3 ,,} b volues

Cl

IBr 6.10 (dd, 1, alefin H, J=11 and 5 Hz)

a::;>(

os 80-82 85 4.92 (t, 1, methine H, J=8 Hz) 2.42 (s, 3, CH 3)

CD

@

Br 3

2i

OT os _{6.04 (t, 1, alefin H, J=8 Hz)}

~IBr

(j-Br

_91-93 93 5.52 (dd, 1, methine H, J=10 and 5 Hz) 2.40 (s, 3,

®

H Sr 5. 85 (t' 1, alefin H, J=8 Hz) d / B r

ccx

5.27 (m, 2, H cis double bond) 89-91 89

4.85 (dd, 1, methine H, J=10 and 6 Hz) OTos _{2.39 (s, 3,}

0

@ Br

eX ..

C::XOTos 5.75 (rn, 2, alefin IJ and methine Ü)

108-111 81

2.41 (s, 3, CH 3)

Br

0

®

1) Satisfactory analytica! data were obtained (~ 0.3% for C and H).

II-2 Bilver nitrate promoted ring expansions of geminal dibromoeyelopropanes

The geminal dibromocyclopropanes 1-4 react rapidly with silver tosylate giving either trans or cis products. In order to prove this tendency a series of ring enlargement reactions with silver nitrate6_•7 _{was carried out under similar}

con-ditions as described for the silver tosylate promoted ring expansions. The substrates 1 and 3 gave the trans allylic nitrates 20 and 21 respectively.

The identity of 20 was proved by synthesis from the parent alcohols with acetyl nitrate in acetic anhydride-methylene chloride mixture8

• The structure of 21 could be established

by reduction with lithium aluminohydride in refluxing THF to trans,eis-2-bromocyclonona-1,6-diene (22) 9,10•

cq<"

0

cq<"

AcN0

3, 0

...

.

"OH AcfJ _CH2Ct2 "ON02

®

Br

@

Br

Q:,X"

LiAlH4, THF

Q:{<"

.

...

.

"ON0₂ reflu x "H

@

Br

@

Br

1

LiAIH4 ,0' ether

(9<"

' _{OT os}

@

Br

Upon reaction of 2 with silver nitrate in acetonitrile

ois-2-bromocyclonon-1-en-3-ylnitrate (23) was formed as sole product. This could be confirmed by nitrating the cis alco-hol 11 .

The ring opening of 4 with silver nitrate seemed, according to tic, to afford a single isomer. However, the 1H-nmr spec-trum revealed a 1:1 mixture of ois-and tPans-2-bromocyclodec-1-en-3-ylnitrate 24 and 25 respectively.

The isomers 24 and 25 were not separable by chromatography and could only be obtained by synthesis from their parent alcohols. The formation of both the trans and cis isomer 24 and 25 may be explained readily by assuming a competition between attack of the nitrate anion and isomerization of the trans cation into the cis cation. This contrasts with the re-action of 4 with silver tosylate which leads exclusively to the cis tosylate 16. Apparently the nitrate anion is a better nucleophile than the tosylate anion. The formation of only a cis nitrate 23, in the ring opening of 2, is probably the result of such a severe strain in the intermediate cation that isomerization occurs far more rapidly than attack of the nitrate anion. A survey is given in Table III.

~ Substrot es

d

_(Î)

~~

0

cY:

-~

w~

®

cY:

®

Table III

Reaction of geminal dibromocyclopropanes with AgN03 in aceotonltrile Praducts trotio)

~:02

@) Br ~NOz ~Br @ ~H ~~NO:z

@)

Br

WH

_@ 50 Br '6N0:2

cyç:o

2 @ 65H

c:x::02

@so

dNOz

@3S l Yield,0_{/o Time,hr} 71 78 62 83 3 72 0.5 NMR dato I CDCl3 I, 6 vol u es 6.37 (t, 1, alefin H, J•B Hz) 5. 76 (t, 1, mcthine H, J•B Hz] 6.33 (t, 1, alefin H, J=9 Hz) 5.86 (m, 1, methlne fl) 6.08 (dd, 1, olefln H, J•8 and 6 Hz] 5.35 (m, 2, cis double bond) 5.14 0•, 1, mcthine H]

~: 6.37 (t, 1, alefin H, J=8 Hz) 5. 21 (t, 1, methine H, J=7 Hz)

cis 5.82-6.32 (m, 2, alefin H and methine H)

~: 4.72-5.28 (m, 1, methine H) 5.36-6.03 (m, 2, alefin H)

cis 5.20-6.08 (m, 3, methine H + alefin H) 13_{c (ppm downfield to external TMS)}

trans: 86.8 methine C

cis 82.9 mothine C

With regard to the previously presented data it is clear that two factors control the cis or trans geometry of the products formed after ring opening, viz. the strain in the incipient trans cation (ring size) and the nucleophilicity of the anion. If the ring size imposes severe restrictions on the formation of a free transient trans cation, then the production of trans products is favoured, without regard to the nature of the nu-cleophile. Larger rings allow the formation of a transient trans cation, which can isomerize to a more stabie cis cation, thus leading to cis products. The extent to which such an iso-merization takes place is clearly dependent on the nucleophili-city.

II-J InfZuenaee of nuaZeophiZiaity and ring eize in the aZcoholysis of some dibromoaycZopropanee

It has become clear in the preceding section that the ring expansion products all originate from the same kind of a disrotatory ring opening of the cyclopropane ring. The extent to which cationic character may be ascribed to the transition state is clearly dependent on the ring size. The experiments carried out with a series of geminal dibromo-cyclopropanes show that 9,9-dibromobicyclo[6.1 .Oinonane (2) is the smallest ring in the series which obviously possesses the property of undergoing ring opening in a semi-concerted way i.e. via a more or less free transient cation. There is no reason to assume why the solvolytic reactions should exhibit a different behaviour towards nucleophilicity than the reactions performed under non-solvolytic conditions. So, a series of solvolytic ring expansions was carried out with 2 which clearly demonstrate the effect of nucleophili-city on the cis/trans ratio of the products formed.

As typical reagents for such an experiment the following alcohols were chosen with decreasing nucleophilicity in the order methanol > ethanol > iso-propanol > t-butanol. The reactions were carried out with initial concentrations of lM of silver perchlorate at 40°. The results are summa-rized in Table IV.

Table IV Aleoholysis of 9,9- dibromobicyclo[6.1.0]nonone OR

c:J-.,.

~

~··~aR

@)

@

Br R Cis,% Trans,%

0

CH3 10 90

(§)

C2H5 39 61

CS)

_iso-c3H7 53 47

@

i·C4Hg 64 36 S3

The results which are achieved with these solvolytic ring expansions are self explanatory. A rather drastic increase in cis products is observed with decreasing nucleophilicity of the solvent. Obviously the influence of nucleophilicity is in both solvolytic and non-solvolytic reactions an impor-tant factor in determining the cis/trans ratios. Alternative-ly the ring opening reactions of 9,9-dibromobicyclo[6.1.0) non-4-ene (3) and 8,8-dibromobicyclo[5.1.0)octane (1) should give trans products exclusively, even if solvolyzed in the solvent with the lowest nucleophilicity viz. t-butanol. Indeed, reaction of 1 and 3 with silver perchlorate in t-but-anol at 40° afforded only trans t-butyl ethers.

H Br

lj·~

CD

_____

,.,.

. AgCl04-t-C4HgOH 40~ 10 min H Br

~~,

----·

0

... (f;i'VH

~·''OH pTSA,

0

Br toluene

The configurational constitution of 28 and 29 was determined readily by cleavage of the ether linkage by treatment with p-toluenesulfonic acid in refluxing toluene for 10 minutes. This led to the formation of the known alcohols 5 and 7 (see Chapter I). The formation of 5 from 28 was accompanied by some isomerization to the cis alcohol, a phenomenon which is not unusual if we consider the strain at the double bond.

The two isomers could be separated by application of column chromatography. The result obtained by alcoholysing 10,10-dibromobicyclo(7.1.0ldecane (4) in t-butanol is in full agree-ment with the hypothesis that nucleophilicity and ring size determine the final cis/trans ratio. Upon treatment of 4 for 10 minutes at 40° with a 1M solution of silver perchlorate in t-butanol, a mixture of two t-butoxy ethers was obtained, viz. 80% of the trans isomer (31) and 20% of the cis isomer (30). The transient trans cation in the ten-membered ring is much less tained with strain and consequently the propensity to isomerization is strongly diminished.

H Br

~'

trans I eis N 80/20

The reaction of 9-exo-bromobicyclo[6.1.0lnonane (32) with silver perchlorate in t-butanol affords in a very rapid re-action mainly trans-3-t-butoxycyclonon-1-ene (33) which con-tained about 10% of the related cis compound (34).

®

The considerable differences between the cis/trans ratios in the products arising from 9,9-dibromobicyclo[6.1.0lnonane (2) and 9-exo-bromobicyclo[6.1 .Olnonane (32) at first glance seem rather unexpected. Probably this discrepancy can be explained in a quite simple way if we consider the stereochemistry of the cation in question (Fig. I).

Figure I

The difference in transition state between the monobromo derivative (32) and the dibromo derivative (2) is only the presence of a bulky bromine atom in the cation formed in the ring opening of 2. Apparently increased stability arising from the interaction of the pn-electrons of the bromine atom with the n-electrons of the allylic cation is of minor impar-tanee with respect to the steric influence of the bromine atom in the transition state. Therefore the enhanced propen-sity to isomerization in the ring opening of 2 might be due to a severe trans-annular interaction between the bramine atom and the methylene hydragen at

c

_{7 . Ring opening of 32}

with silver nitrate, alternatively, gave a 2:1 mixture of trans and cis nitrate 35 and 36 respectively. This observation fits well with the hypothesis that the configuration of the product will be dependent on the relative nucleophilicity of the anion. The influence of unfavourable steric interactions in the for-mation of anomalous products has been observed earlier in the dihalocarbene addition to norbornene11_-1_~.

. H Br

~''"

...

+

@

trans 1 cis cv 65/35

II-4 StPuctuPe assignments

The contiguration of products 26a-d and 27a-d arising from the ring expansion of 2 were determined with the aid of their 1H- and 13

c-

nmr spectra (see Table V). In the proton spectrum the trans products 27a-d exist as two rapidly equi-librating diastereoisomers (rotation of the trans double bond through the loop of the ring15

•16) . Their spectra display typical double doublets for the olefinic region. The methine part of the spectrum shows a characteristic double doublet and a lower field multiplet. The corresponding cis structures 26a-d can be detected readily by their typical olefinic tri-plet and by the signal of the methine proton; a multitri-plet which resonates always at lower field than the methine protons of the trans diastereoisomers. These observations are in good agreement with recently reported data for similar compounds17_, 18_• An additional and valuable methad of determining the contigu-ration of the product consisted in comparison of the 13c-nmr spectra. Of importance is the resonance of the allylic carbons. One of these, to which the alkoxy subsituent is attached may readily be found.

For the cis isomer this latter allylic carbon resonates ge-nerally at 5-10 ppm upfield relative to the allylic signal of the corresponding trans diastereoisomers19

,

In this way the mixtures 30/31 and 33/34, as well as the mixtures of trans- and cis-cyclonon-1-en~3-ylnitrate (36 and 35) were analyzed unambiguously. The structural assign-ments of the two trans compounds 28 and 29 were made in accordance with their 1H-nmr spectra20 _{• The coupling}

con-stants were in good agreement with the values measured in analogous compounds.

Compound 26a, 27a 26b' 27b 26c, 27c 26d, 27d 29 28 30' 31 33, 34 Tabl e V 1 H-nmr values (CC1₄ solutions) 3.91 and 3.47 (m and dd, methine H-3, trans, J~s and

10 Hz), 4.20 (rn, methine H-3,

cis)

3.62 and 4.02 (m and dd, methine H-3, trans, J•S.S and 1 0 Hz),

4.35 (m, methine H-3, cis)

4.11 (rn, methine H-3, trans),

4.40 (rn, methine H-3, cis), 6.20 (t, J=9 Hz, olefin H-1, cis)

3.72 and 4.07 (rn and dd, trans t

J methine H-3, J=5 and 10 Hz), 4.41 (m, cis, methine H), 6.02 (t, alefin ll-1, J=9 Hz, cis) 5.78 (t. 1, H-1, J=8 Hz), 5.19 (m, 2. H-5 and ll-6). 3.84 (m, 1, rnethine H-3) 5.99 (dd, 1, H-1, J=11 and 4. 5 3.96 (t' 1, H-31 J=B Hz) 6.24 (t. J=8 Hz, trans·, H-1), 5.82 (dd, H-1, cis, J=12 and 6 4.49 (m, H- 3 ~ cis) , 3.99 (m, H-3, trans) 3.59 (m, methine H, trans), 4.06 (m, methine H, cis) 5. 21 (m, olefin H) Hz), Hz), 13_{c-nmr, ppm downfield to} external TMS in c₂Br₂F₄ 56.9 85.9 , 79.2 (C-:i, cis), 87.2 (C-3, trans) 64.7 (CH 3-ç_H20), 77.4 (C-3, cis), 84.1 and 85.4 (C-3, trans)

69.2 ( (CH₃)₂f.), 74.2 (C-3,

cis), 81.3 and 82.7 (C-3, trans)

70.4 (C-3, cis), 74.7 ( (CH 3)3f_), 78.1 and 79.3 (C-3, trans) 75.1 (f_(CH 3)3), 78.7 (C-3, trans) 69.2 (C-3, cis), 75.0 (f.(CH₃)₃) 77.6 (C-3, trans) 68.9 (C-3, cis), 73.9 (f_(CH₃)₃), 75.9 (C-3, trans)

1) Only the most significant signals were tabulated, because in the methine region

the slgnals are often overlapped by alcoxy protons, whereas in the olefinic region the protons of cis and trans structures coincide.

II-5 Experimental seation

General. The starting materials 1-4 were prepared according to known procedures by the reaction o~ dibromocar-bene with the appropriate olefins 21 .

9-endo-Bromobicyclo[6.1.0]-nonane (19) was obtained from reduction of 2 with tri-n-butyl-tinhydride22·23. The isomerie 9-exo-bromobicyclo[6.1.0]nonane (32) was synthesized from 2 by reduction of the dibromide 2 with dimsyl anion in DMS0 24 . Silver tosylate was prepared from silver oxide and p-toluenesulfonic acid, according to the pro-cedure of Kornblum et al. 25 . Cyclononene, required for the preparation of 4 was obtained from 9,9-dibromobicyèlo[6.1.0l-nonane (2) by conversion to cyclonona-1,2-diene and subsequent reduction 26 . 1H-Nmr spectra were measured on a Varian T-60 spectrometer and 13c-nmr spectra were obtained on a Varian HA-100 apparatus at 25.12 Mc/s.

A typical experimental procedure for the preparation of the tosylates is exemplified with the synthesis of 14. The ring expansion reaction with silver nitrate is illustrated with the preparation of 23. The aleoholysis reactions were performed at 40°, starting with an initial 1M concentration of silver per-chlorate. A typical experiment is illustrated by the preparation of 28.

2-Bromo-3-tosyloxycyclonona-trans,ais-1 ,5-diene (14)

To a solution of 2.80 g (0.01 mol) of 3 in 10 ml of acetoni-trile was added a solution of 3.10 g (0.011 mol) of silver tosylate in 15 ml acetonitrile. The mixture was stirred with gentie reflux for 2 hr. After cooling and addition of an equal

vo~ume of ether the precipitate was filtered and the filtrate evaporated to dryness. The resulting gummy product was chroma-tographed through a short silica gel column and afforded 3.3 g (89%) of white crystalline tosylate 14. Recrystallization from diisopropyl ether gave analytica! material; mp 89-91°. Anal. Calcd for c 16H19Brü3S: C, 51.75; H, 5.12. Found: C, 51.64; H, 5.12.

cis-2-Bromocyclonon-1-en-3-ylnitrate (23)

A solution of 2.82 g (0.01 mol) of 9,9-dibromobicyclo[6.1.0l-nonane (2) and 3.33 g (0.02 mol) of silver nitrate in 20 ml of acetonitrile was refluxed with stirring for 4 hr (progress of the reaction was monitored by tlc). After cooling the re-action mixture was poured onto 100 ml of saturated sodium chloride solution and 75 ml of ether. Stirring was continued for 10 minutes and the mixture was filtered through Celite. The organic layer was separated from some tlc immobile mate-rial by chromatography through a short silica gel column, using pentane as eluent. Upon bulb to bulb distillation 1.86 g (78%) of 23 was obtained as a colourless oil. Nmr data are presented in the Table; ir (neat) cm- 1 2930, 2860, 1630, 1460, 1440, 1360, 1320, 1290, 1270, 1160, 1020, 1005, 970, 962, 946, 920, 890, 845, 750. Anal. Calcd for

c

₉H₁₄N0₃Br: C, 40.91; N, 5.30. Found: C, 40.99; H, 5.36; N, 5.19.

Product 23 was obtained also by nitration of cis-2-bromocyclo-non-1-en-3-ol as follows. To 3 ml of acetic anhydride was added 265 mg (1.1 mmol) of Cu(N0₃)₂.3H₂0.After the appearance of the typical precipitate of cupric acetate the mixture was stirred for an additional 15 minutes and then cooled to 0°. A solution of 210 mg (1 mmol) of cis-2-bromocyclonon-1-en-3-ol (11) in 2 ml of methylene chloride was added in 1 minute. The mixture was stirred for an additional 5 minutes and then poured on

20 ml of water. After neutralization with solid sodium bicar-bonate the product was extracted with ether. The oil which remained after evaporation of the organic phase was chromato-graphed over a short silica gel column (pentane as eluent) and afforded 240 mg (91%) of 23.

trans-2-Bromo-3-t-butoxycyclooct-1-ene (28)

Toa solution of 4.14 g (0.02 mol) of silver perchlorate in 20 ml of t-butanol c~ 15 g) was added at 40° with vigorous stirring 2.68 g (0.01 mol) of 1. Precipitation of silver bromide began immediately. After 20 minutes the starting bromide had disappeared (as evidenced by the thin layer chro-matogram; Merck silicagel plates, benzene as eluent).

Then 20 ml of saturated sodium chloride solution were added. The mixture was stirred for about 5 minutes and filtered through Celite. After dilution with 100 ml of water the pro-duct was extracted twice with ether. Upon washing, drying and evaporation of the solvent 2.16 g (83%) of pure 28 re-mained as a colourless oil. The nmr data are presented in Table V.

A solution of 1.3 g (0.005 mol) of 28 in 10 ml of toluene, containing about 100 mg of p-toluenesulfonic acid was re-fluxed for 10 minutes. The mixture was washed twice with 10% sodium bicarbonate solution and once with water. After drying and evaporation of the organic phase 0.75 g (74%) of a 3:2 mixture of trans and cis alcohol was obtained. These two al-cohols were separated by column chromatography (silica gel chloroform-2% methanol as eluent); Rf (cis alcohol) 0.35; Rf (trans alcohol) 0.29. Nmr (CDC1 3) for the trans alcohol 5:

ó 4.18 (dd, 1, methine H-3, J 10 and 5 Hz), 6.11 (dd, 1, ole-fin H-1, J 10.5 and 4.5 Hz). Nmr (CDC1₃) for the cis isomer: ó 4.71 (dd, 1, methine H-3, J 10 and 5 Hz), 6.21 (t, 1, ole-fin H-1, J 8.5 Hz).

trans,ais-2-Bromocyclonona-1,6-diene (22)

To a suspension of 0.8 g (0.021 mol) of lithium aluminohydride in 10 ml of dry THF was added with stirring a solution of 2.62 g (0.01 mol) of 7 in 5 ml of THF. No reaction took place27

• Then the mixture was refluxed for 0.5 hr. After work up in the usual manner the resulting oil was chromatographed through a short silica gel column, using pentarre as eluent. Upon bulb to bulb distillation 0.7 g (35%) of 22 was obtained; bp 67-70°

(1 mm). Nmr (CDC1₃) ó 5.67 (m, 1, H-1 olefin), 5.37 (m, 2, H-6 and H-7 olefin).

This compound proved to be identical with an authentic sample prepared by reduction of the corresponding tosylate.

Rf!ferences and

notes

1. A. STREITWIESER, Jr., "Solvolytic Displacement Reactions", McGraw-Hill, Inc., New York 1962

2. N. KORNBLUM, W.J. JONES and G.J. ANDERSON, J. Amer. Chem. Soc.,~. 4113 (1959)

3. W.D. EMMONS and H.F. FERRIS, J. Amer. Chem. Soc., 7 2257 (1953)

4. H.M.R. HOFFMANN, J. Chem. Soc., 6748 (1965)

5. The 9-exo-bromobicyclol6.1.0lnonane reacts within 5 minutes completely with silver tosylate under the common reaction conditions. Though a crude tosylate could be isolated, the purification failed.

6. Y. POCKER and P.N. KEVILL, J. Amer. Chem. Soc.,~. 4760 (1965)

7. N. KORNBLUM and D.E. HARDIES, J. Amer. Chem. Soc.,~. 1707 (1966)

8. Acetylnitrate was prepared by modifying a known method; T. SATO, T. AKIMAand K. UNO, J. Chem Soc. (C), 891 (1973) 9. The nitrate 21 could be prepared actually by reaction of

the parent alcohol with acetyl nitrate, but this reaction suffers from giving a lot of side products, probably arising from the addition of acetyl nitrate to the double bondor from trans-annular reactions.

10. C.B. REESE and A. SHAW, Chem. Comm., 787 (1972)

11. W.R. MOORE, W.R. MOSER and J.E. LaPRADE, J. Org. Chem., ' 2200 (1963)

12. R.C. DeSELMS and C.M. COMBS, J. Org. Chem., ~. 2206 (1963) 13. C.W. JEFFORD and R.T. MEDARY, Tetr., ~. 4123 (1967)

14. C.W. JEFFORD, S. MAHAJAN, J. WASLYN and B. WAEGEL, J. Amer. Chem. Soc.,~. 345 (1965)

15. A.C. COPE, K. BANHOLZER, H. KELLER, B.A. PAWSON, J.J. WHANG and H.J.S. WINKLER, J. Amer. Chem. Soc.,

!I•

3644

(1965)

16. G. BINSCH and J.D. ROBERTS, J. Amer. Chem. Soc.,

!I•

5157 (1965)

17. C.B. REESE and A. SHAW, J. Chem. Soc., (Perkin I), 2422 (1975)

18. C.B. REESE and A. SHAW, Chem. Comm., 1365, 1367 (1970) 19. J.W. de HAAN and L.J.M. van de VEN, Org. Magn. Resonance,

~. 147 (1972) and references cited herein

20. The trans-2-bromo-3-t-butoxycyclooct-1-ene (28) isomerizes spontaneously on standing for several days

21. L. SKATTEB~L, Acta Chem. Scand.,

12,

1683 (1963)

22. M.S. BAIRD and C.B. REESE, J. Chem Soc. (C), 1808 (1969) 23. D. SEYFERTH, H. YAMAZAKI, D.L. ALLESTON, J. Org. Chem.,

~. 703 (1963)

24. C.L. OSBORN, T.C. SHIELDS, B.A. SHOULDERS, C.G. CARDENAS and P.D. GARDNER, Chem. Ind., 766 (1965)

25. N. KORNBLUM, W.J. JONES and G.J. ANDERSON, J. Amer. Chem. Soc., .§_1, 4113 (1959)

26. P.D. GARDNER and M. NARAYANA, J. Org. Chem., ~. 3518 ( 1961)

27. Normally this reaction should afford the corresponding alcohol. See: L.M. SOFFER, C.W. PAROTTA and J. DI DOMENICO, J. Amer. Chem. Soc.,

.z.!,

5301 (1952)

GRAPTER

!//

stereoselective synthesis

rif

geminal

endo-iodo-exo-bromo-cyc lopropanes

III-1 Introduetion

In the preceding chapter it has been discussed how the configuration of the product is determined by both ring size and nucleophilicity of the solvent (solvolytic ring expansion) or the anion (non-solvolytic ring expansion). The following general rules are now available.

1. If the expanded ring would become too small to accommodate a trans double bond, then the endo halogen atom is lost. This always leads to cis products via a disrotatory ring opening. Some representative examples are presented below.

H

Ck

r ; r H

®

AgOAc-HOAc re flux - Br- ( endo) AgOAc-HOAc re flux -Cl- (endo)

ö-

OAc Br

l

{ref. 1)

-(ref. 2 I

2. In ring expansions leading to enlarged rings which can bear a trans double bond, but in which the geometry of the en-larged system does not allow the existence of a free trans cation, the exo halogen atom is lost. The nucleophile en-ters concertedly with the ring opening and the product formed does have a trans geometry. These reactions display two-fold stereospecificity. The r opening proceeds in a stereo-electronic controlled way3_{•~ and the attack of the}

nucleo-phile at the incipient allylic centre implies an inversion of configuration5

• So only one diastereoisomer is formed. A representative example is given below.

H ~

~

~a~.

H~H

~OH Br

---5°/o aq.acetone

3. Another important class of bicyclic systems actually

exhi-46

bits opening via free cationic intermediates. Now two pathways exist for product formation. If the cation is trapped immediately the products will have a trans geometry. From molecular models it is obvious that the steric builcl-up of such a cation will not allow the formation of both diastereoisomers one would expect. As such, the final stereo-selectivity of these reactions might be similar to the situ-ation presented for the fully conceited reaction.

If the cation does not react immediately with the solvent or an anion then isomerization to a cis cation may occur. Generally this occurs preferentially in the presence of weak nucleophiles such as No;, t-C₄H₉0H, TosO • The products will show now a cis double bond arrangement.

AgOTos

The formation of either cis or trans products is encountered only in two extreme cases. A reasonably stabilized trans cation reacting with a hard nucleophile leads to a trans configuration, whereas a trans cation which exhibits strong propensity towards trans+cis isomerization reacts with a weak nucleophile under formation of a cis product.

In practice these extreme prerequisites will seldom be en-countered and in most cases hitherto known mixtures of cis and trans products are formed. Representative examples are

e.g. the ring opening of 10,10-dibromobicyclo[7.1.0)decane

(4) with silver nitrate and the t-butanolysis of 9,9-dibromo-bicyclo[6.1.0lnonane (2). H Br

cj::~

d~r

AgCI04

_(R%"

₊ !-C4HgOH - crt-c _{- 4 9} H

0

@

Br

8

H Br

c:x~r

AgN03

-~H

w:ON02

+ C~CN ~ ., _Br

0

@

Br "'oNo2

®

Only in the case of very weak nucleophilic anions like the tosylate anion, the ring expansion of 2 and 4 results in cis products exclusively.

III-2 Iodination of endo-Zithio-exo-bromocyaZopropanes

In a similar way the synthesis of a vinyl iodide has been described by ring opening of a geminal diiodocyclopro-pane precursor6 _{(41, 42).}

@

The examples of diiodocyclopropanes described in the litera-ture are scanty, because these compounds are in most cases too unstable to be isolated or disproportionate rapidly at room temperature. Some efforts directed towards the synthe-sis of functionalized nine-membered systems bearing a trans vinyl iodide moiety, were rewarded with very moderate success due to the extreme instability of the prerequisite geminal diiodocyclopropanes7_•8_{• As it is to be expected that mixed}

dihalocyclopropanes are more stáble precursors than the corresponding diiodo derivatives different chemical routes for their preparatien will be discussed. Synthesis of these hitherto unknown substrates by stereoselective addition of iodobromocarbene to cyclic olefins seemed to be less promi-sing, with regard to the results of a variety of mixed car-bene additions to cyclic olefins9_- 15_{• Köbrich et aZ.} 16 _found that upon treatment of 7,7-dichloronorcarane with n-butyl-lithium at -95° a reasonably stable mixture of endo- and exo isomers of 7-lithio-7-chloronorcarane was obtained.

Recently Hiyama et aZ. 11

, showed that the corresponding

7,7-dibromonorcarane could be lithiated stereoselectively at -95° leading to 7-endo-lithio-7-exo-bromonorcarane. Subsequent addition of methyl iodide afforded the corresponding 7-endo-methyl-7-exo-bromonorcarane (43). 1) Buli, THF, -95° 2lCH3I, -95°

-H

CK,

_H

@)

Based on the above mentioned considerations it was to be ex-pected that iodination of such a lithiate would afford the corresponding endo-iodo-exo-bromo derivative. This method had been used for the precursor 9,9-dibromobicyclo[6.1.0lnonane

(2). Treatment of 2 with one equivalent of n-butyllithium at -95° results indeed in the specific formation of 9-endo-lithio-9-exo-bromobicyclo[6.1.0lnonane. This was evidenced by a

quenching experiment with methyl iodide (see Chapter IV). Upon treatment of this lithiate with iodine in THF at -95° a single isomer was isolated which proved to be structure 44.

@

This compound displayed a typical 13c-nmr resonance at 4.1 ppm downfield to TMS ( fBri, see Table VI). With the aid of simple silver ion promoted ring expansions the stereochemistry of 44 was determined as 9-endo-iodo-9-exo-bromobicyclo[6.1.0lnonane

On treatment of 44 with a molar excess of silver perchlorate in 10% aqueous acetone a mixture of the diastereoisomers of trans-2-iodo-3-hydroxycyclonon-1-ene (45) and cis-2-iodo-3-hydroxycyclonon-1-ene (46) was obtained (approximate ratio 45/46 ~ 75/25). Close examination of the crude cis product revealed the presence of about 20% of cis-2-bromo-3-hydroxy-cyclonon-1-ene (11). The presence of this product may readily be explained by assuming some competition of a mo4e of ring opening in which the endo iodine atom is expelled. This side reaction is not unexpected since the endo site now holds a much better leaving group18

• The ring expansion of 44 with silver tosylate gave, as expected, cis-2-iodo-3-tosyloxycyclo-non-1-ene (47) only. lts identity could be established by synthesis from the corresponding cis-2-iodo-3-hydroxycyclo-non-1-ene (46).

Scheme I

H Br

I

d~

Ag+

w

H20

O::X"

----

.

' _OH

@

@I

OT os

'

OH

C)-1

TosO-

w-1

HzO

...

Cj-1

@

@

50

In a similar way as outlined for the preparation of 44 the compounds 1, 2, 37 and 48 were converted into their corres-ponding geminal endo-iodo-exo-bromo derivatives 50, 44, 49 and 51 respectively (see Table VI).

Same final remarks should be made with regard to the iodina-tion of the lithiates. Attempts to synthesize 6-endo-iodo-6-exo-bromobicyclo[3.1.0lhexane (53) from the corresponding dibromo derivative (52) met with failure. Presumably this lithiate disproportionates at -95° via an intermediate allene19 •20•

d"Bc

Buli, THF

a·'"

_...

[0]

-95°

...

@)

+

I

'

'

~·'"'

dimers

®

Another interesting observation is the fact that 9,9-dibromo-bicyclo[6.1.0]non-4-ene (3) could not be converted into the corresponding 9-endo-iodo-9-exo-bromobicyclo[6.1.0lnon-4-ene. The lithiate derived from 3 proved to decompose instantaneous-ly into the corresponding allene (54) even at -95° as evidenced by an attempt to quench this lithiate with water.

H Br

c}(·'

0

1)8uLi,-95°

---~'

Table VI

GE>m1nal ~-bro mo- E>ndo- iodocyclopropanE>s

Substra!E>s Produc ts 1 YiE>ld,% Bp(mm].Mp{°C I spE>ctro ( ppm val u es H Br

d'

èt''

I 77 69-72 (0.01) 10.0 (CBri), 29.4, 33.9, 35.0, 36.2

CD

@ H Br

c!l

C : X ' B r H''I 82 77-79 (0.02) 4.3 (CBri), 27.9, 29.3, 31.4, 34.7

®

@

ä

H ar 'Br

0

~1fi

_{90 102-103}2 2.3 (CBri), 27.2, 28.3, 32.8, 33.1, 34.2, 34.5, 81.7, 82.2, 0 -J---o''~~.

-1-

< H 108.4

"'

0 @ H

~'

(tl.'

69 71-75 (0.05) 11.3 (CBri), 21.7, 25.5, 29.4 , 1 'ar @H @)H

1) Satisfactory elemental analyses were obtained (~ 0.3\ for C and H) for all products. 2) Crystallized from 95% ethanol.

3) Spectra obtained from neat products; except 51 which was measured as a 30t salution

This phenomenon is discussed in detail in the next Chapter. The silver ion assisted salvolysis of 49 is worth recording since 7,7-dihalobicyclo[4.1.0lheptanes are prone to undergo ring expansion by expelling the endo halogen atom18 _{• Upon}

heating 49 with silver acetate in acetic acid for several hours only ais-2-bromo-3-acetoxycyclohept-1-ene (38) was isolated in 74\ yield.

This synthesis of endo-iodo-exo-bromocyclopropanes is a hither-te unprecedenhither-ted approach which permits the shither-tereoselective introduetion of the elements of iodobromocarbene into an ole-finic system via an indirect way. Finally it is noteworthy to comment that in contrast to the diiodocyclopropanes, the corresponding endo-iodo-exo-bromocyclopropanes show indeed a high stability. Even on standing at room temperature for seve-ral weeks no disproportienation could be detected.

III-3 SiZver fluoride as an effeative oataZyst in ring expansions of geminaZ dihaZooyaZopropanes

The ring expansions which have been presented in the preceding section can be characterized as follows.

If the ring size and the relative nucleophilicity of Nu are known (either solvent or anion) it must be possible to make rather accurate forecasts which products (and in which ratios) are formed (see Scheme II).

--X, Nu coneerled

r-t~

,

\/"'"~'\

y

Nu

---·

'

Nu Nu

--Y, Nu isomerisation _

~+

..

_,

H

H

y

One interesting set of experiments with silver fluoride is finally described. Treatment of 9,9-dibromobicyclo[6.1.0]non-4-ene (3) with two equivalents21 _{of silver fluoride gives via} a disrotatory ring opening with concomitant attack of the fluoride anion at the incipient allylic centre

trans,eis-2-bromo-3-fluorocyclonona-1 ,6-diene (55). Details have been pre-sented in Table VII.

AgF, CH3CN reflux, 3hr

-~H

~~F

tf?l Br ~

The 1H-nmr spectrum displayed a typical double doublet coup ling pattern for the allylic hydragen (J 10 and 6 Hz) and a geminal F-H coupling of 46 Hz. The olefinic proton adjacent to the bramine atom resonates as a lower field triplet (J 8 Hz). This is in good agreement with the observations made upon reaction of 3 and a number of nucleophiles under silver ion assisted ring opening. The dibromides 2 and 4 gave mix-tures of cis and trans products upon reaction with silver fluoride in acetonitrile. H Br

c:z,~

AgF,CH3CN

~

F

----=---

V-rBr

eq;

0

@)

Br cis/trans C\l 60/40 cis/ trans C\l 20/80

The cis and trans products could be separated by column chroma-tography in both cases. The less polar compounds proved to be the trans isomers. Structure assignment of 56 and 57 can be made readily.