Carbonyls as free-radical precursors

MISSING IMAGE

Material Information

Title:
Carbonyls as free-radical precursors cyclizations of unsaturated and epoxy carbonyls
Physical Description:
vii, 105 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kinter, Kevin S
Publication Date:

Subjects

Subjects / Keywords:
Free radicals (Chemistry)   ( lcsh )
Cyclic compounds   ( lcsh )
Ketones   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 98-103).
Statement of Responsibility:
by Kevin S. Kinter.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001890693
oclc - 29642201
notis - AJW5917
sobekcm - AA00004738_00001
System ID:
AA00004738:00001

Full Text













CARBONYLS AS FREE-RADICAL PRECURSORS:
CYCLIZATIONS OF UNSATURATED AND EPOXY CARBONYLS















BY


KEVIN S. KINTER


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



UNIVERSITY OF FLORIDA


1993































To Mom and Dad,

with all my love and appreciation
















ACKNOWLEDGEMENTS


The only words that I think aptly describe the last 10

years of my life are from Jerry Garcia, "What a long, strange

trip it's been." I can only hope that there is life after

education. It seems like it was just yesterday that I was

leaving home for the first time to go to a strange place

called college. I cannot even begin to comprehend all of the

things that have happened to me since that day. But there

are many people who I can thank for all their help, guidance,

understanding, and friendship.

First and foremost, I would like to thank my research

director, Dr. Eric Enholm, for all of his patience, guidance,

and assistance. His help in the preparation of this

manuscript was immeasurable. I would also like to thank my

undergraduate research director, Dr. Gary Crowther, for

introducing me to the wonderful world of chemistry. I would

like to thank all of the professors in the division of

organic chemistry who showed me that there were many facets

to being a Ph.D. chemist. I would also like to thank Dr.

Girija Prasad for training me and for his contributions to

the unsaturated ketone project.

The thing that I will remember best and miss most about

the University of Florida is the many friendships that I have









developed since I have been here. Thanks go to Keith Palmer

and Brent Kleintop for putting up with me this long and for

showing me the mechanistic side of organic chemistry (Keith)

and that analytic chemists are not just mechanics (Brent).

Thanks also go to fellow graduate students and fantasy

football friends Don Eades and Larry Villanueva for many

drunken hours of Gator sports prognosticating. Additional

thanks are extended to all of the past and present members of

the Enholm research group, especially Janet Burroff, Hikmet

Satici, and Jeff Schreier. I wish them all the best of luck

in their careers and personal lives.

I am also grateful to Karin Larsen for making this last

year very pleasant, despite the problems encountered at work.

I wish that we had met earlier, but I hope we will continue

to build on what we have begun. My family truly deserves the

most appreciation though. Thanks go to my oldest brother,

Mike, who showed that a Kinter could get his Ph.D., and to

Chris for being an organic comrade in the family. Thanks go

also to Jim and Tom for showing me that there is more to life

than chemistry. But most of all, I am truly indebted to my

parents. Without the emotional and financial support of my

parents I would never have been able to achieve this goal. I

feel very fortunate to have had these opportunities, which

would not have been possible without the support of all of

those persons mentioned above. For this I will always be

indebted to them.

















TABLE OF CONTENTS





ACKNOWLEDGEMENTS .......................................... iii

ABSTRACT ...................................... .............vi

CHAPTERS

1 INTRODUCTION..... ........ .............................1

2 CYCLIZATIONS OF a,P-UNSATURATED KETONES ..............28


3 CYCLIZATIONS OF c,P-EPOXY CARBONYLS .................. 48

4 SUMMARY..............................................71

5 EXPERIMENTAL.........................................75


General............. ....................... ........
Experimental Procedures and Results................

LIST OF REFERENCES .......................................

BIOGRAPHICAL SKETCH ......................................


S.75
..76

S.98

.104
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


CARBONYLS AS FREE-RADICAL PRECURSORS:
CYCLIZATIONS OF UNSATURATED AND EPOXY CARBONYLS

By

Kevin S. Kinter

May, 1993




Chairman: Eric J. Enholm
Major Department: Chemistry


This dissertation investigated the tributyltin hydride

induced cyclization reactions of unsaturated ketones and

epoxy carbonyls with olefins. The reaction of tributyltin

hydride with aldehydes and ketones produces an O-stannyl

ketyl, which has both anionic and radical character. The

goal of this study was to ascertain whether the radical

reactivity of the O-stannyl ketyl could be delocalized away

from the central-carbon atom by the participation of labile

functional groups.

The first area of study was the intramolecular coupling

of unsaturated ketones with alkenes to produce functionalized

cyclopentanes. Unactivated alkenes were found to be

unsuitable as radical acceptors and activation of the alkene

was essential to the cyclization. A dilution study revealed










that excellent anti stereoselectivities (>50:1) could be

achieved, and this was attributed to a reversible

cyclization. Product identification and enolate trapping

studies demonstrated that the anionic character of the tin

ketyl could also be utilized. This was the first reagent-

based approach to study the coupling of the P-carbons of

these systems.

The second area of study investigated the cyclizations

of epoxy carbonyls with olefins. The C-O bond of the epoxide

was fragmented by the adjacent tin ketyl radical to produce

an alkoxy radical. This project was originally designed to

produce tetrahydrofurans, however these reactions produced

functionalized cyclopentanols. Originally it was believed

that the reactions were proceeding through a hydrogen

abstraction mechanism, but after further experimentation a

rare 1,5-tin transfer mechanism was proposed.


vii















CHAPTER 1

INTRODUCTION



The term "free radical" applies to a species which

possesses one unpaired electron.1 Free radicals were once

thought to be indiscriminate, highly reactive intermediates,

but lately are being viewed in a kinder, gentler light.2

Radicals are formed by a homolytic cleavage of a covalent

bond. The central atom is usually sp2 hybridized and the

unpaired electron rests in the p-orbital.3 The reactivity

which radicals exhibit is dominated firstly by the nature of

the central atom and secondly by the substituents which are

attached to it.2

Free radicals have been known since 1900 when Gomberg

investigated the formation and reactions of triphenylmethyl

radical.4'5 Paneth not only discovered that less stable alkyl

radicals exist, but also measured their lifetimes in the gas

phase.6 Synthetic applications of radicals began in 1937

when Hey and Waters described the phenylation of aromatic

compounds by dibenzoyl peroxide.7 In the same year Kharasch

proposed that the anti-Markovnikov addition of hydrogen

bromide to an alkene was also a radical chain process.8

Until the 1970s, synthetic free radical chemistry developed









slowly and was mostly applied only to copolymerization

reactions.9, 10

In the 1970s though, new synthetic methods involving

free radicals began to be developed,11,12 and lately radical

cyclizations have become integral parts of many elegant

syntheses.13 The power and versatility of radical chemistry

is best demonstrated by Dennis Curran's synthetic work

towards the class of natural products known as triquinanes.14

Curran and Rakiewicz's landmark synthesis of hirsutene Z, as

shown in Scheme 1-1,15 utilizes a tandem radical cyclization

approach as the key step. The linear triquinane was produced

in a single step by a tandem sequence commencing with the



I H CH3
P TBTH,AIBN _
C6H6,A

H H
1 2

Scheme 1-1



generation of a 5-hexenyl radical from primary iodide i which

was captured by the olefin and finally terminated by addition

to the suitably disposed alkyne. Curran has also used tandem

radical cyclizations in the synthesis of other members in the

triquinane family (Figure 1-1), such as capnellene,16

coriolin,17 modhephene 18 silphiperfolene 4, 19 and

hypnophilin 5.17









CH3
SH OHI
H 0




SCH3 H O


3 4 5
Propellane Angular Linear

Figure 1-1

Other triquinanes synthesized by Curran



Before one can appreciate the complexity and grace of

the aforementioned syntheses one must understand the basic

principles of these reactions. Radical processes typically

are chain reactions, and the first element of a chain

reaction is initiation. Many reagents which will begin a

radical chain reaction are available to the synthetic

chemist.2 The reagent of choice, from both a safety and a

practical standpoint, is azobisisobutyronitrile (AIBN) E,

which is thermally decomposed to give cyanoisopropyl radicals

(Scheme 1-2) These radicals are not reactive enough to

abstract alkyl's hydrogens, but they are capable of

abstracting a hydrogen atom from the weak Sn-H bond of

tributyltin hydride (TBTH) 9 to produce the very useful

tributyltin radical 11.2

For more than 20 years tributyltin hydride has been

known to engage in free radical reactions.20 The chemistry

that TBTH engages in is too extensive to cover here, but good









reviews of its reactions are available.1,13,21,22,23 It is

commercially available and it can be prepared from

bistributyltin oxide and polymethylhydoxy siloxane.24 Lately,

there have also been some tin hydride reagents developed

which are immobilized on polystyrene beads.25 Although there

are many different initiators and reagents, this discussion

will focus mainly on the AIBN/TBTH system which was employed

in my work. The majority of current radical reactions are

based on the chemistry of this reagent combination.22




N=N 2 Y + N2 (1)

CN CN CN


6 7 8



S + Bu3SnH H + Bu3Sn- (2)

CN CN


7 9 10 11

Scheme 1-2



For a chain reaction to develop, the radicals formed in

the initiation steps must propagate. Propagation (Scheme 1-

3) occurs when a radical 11 interacts with a nonradical 12 to

produce a new radical species 13. Without propagation the

chain would never develop because if two radicals combine to

form a nonradical, the process is terminated. Careful

control of reaction variables, such as concentration and the









substrate to initiator ratio, helps to avoid such

nonproductive reactions. The organotin radical which is

produced is not synthetically useful until it reacts with

other functional groups to generate an organic-centered

radical.


Bu3Sn* + R-X R* + Bu3Sn-X (3)


11 12 13 14
X= Halogen, -SR', -SeR', -NO2

Bu3Sn
Bu3Sn* + (4)


11 15 16




Bu3Sn* + x Bu3Sn-X- (5)


11 17 18
X= C, O, S

Scheme 1-3



There are two broad classes of free radical reactions:

atom or group abstraction (eq. 3) and addition to multiple

bonds (eq. 4 and 5). Halide abstraction was discovered in

1957,26 and has achieved great importance in organic

chemistry since. Atom abstraction involves an SH2 reaction

by tin radical to generate an organic radical. Bromine and

iodine are most commonly used because chlorine reacts

sluggishly and fluorine is unreactive. The degree of









substitution of the carbon which bears the halide also has an

effect on the rate of this reaction, which is as follows:

10RX < 20RX < 3oRX.27 This trend parallels the relative

stability of the resultant alkyl radicals.

Thiols and thioethers can also be reduced cleanly with

TBTH, due to the strong tin-sulfur bond which is formed.21

Thioketals can also be reduced in the same manner, which

offers an interesting alternative for ketone reduction.28 The

most useful of the sulfur-related reductions involves the

addition, fragmentation reactions of thiocarbonyls which were

developed by Barton and McCombie.29 This deoxygenation

methodology was developed because of the inactivity of

alcohols towards TBTH reductions. This method can be applied

to a large variety of hydroxy compounds, including primary,

secondary, tertiary, and diols.21 It has been used in the

synthesis of many natural products such as compactin,30

anguidine,31 and gibberellin.32

Another way in which tin can generate a carbon-centered

radical is by its addition to multiple bonds. When 11 adds

to alkenes or alkynes the ensuing radical 16 or 18 can then

abstract a hydrogen from TBTH, giving the hydrostannated

products 19 and 20. In the process another molecule of

tributyltin radical, which satisfies the last criterion for a

radical chain reaction, has been generated. When a terminal

alkyne is reacted, the tributyltin will add to the terminus.

The reaction proceeds through overall anti-addition, which

gives the Z olefin as the kinetic product, but excess tin can









equilibrate the reaction mixture so that the

thermodynamically favored E-olefin is produced.21


Bu3Sn Bu3Sn
S+ Bu3SnH ""-- + Bu3Sn*
SH

16 9 19 11




Bu3Sn-X- + Bu3SnH Bu3Sn-X H + Bu3Sn*


18 9 20 11

Scheme 1-4



The chemistry which is most important to this thesis is

how tin reacts with carbonyls. Researchers have postulated

that the hydrostannation of a ketone or aldehyde carbonyl can

occur by two different mechanisms, depending on the reaction

conditions, shown in Scheme 1-5.21 When polar solvents and

Lewis acid catalyst are utilized the ionic pathway dominates

(eq. 6). In this mechanism, TBTH acts as a true "hydride

donor" giving intermediate 22 which reorganizes to give the

tin alkoxide. An alcohol can be obtained when 23 is treated

either with proton sources or with additional TBTH.21 Silica

gel33 and tributyltin triflate34 have proved to be valuable

catalysts.

An alternative mechanism is the way in which TBTH reacts

with carbonyls through a free radical pathway (eq. 7). The

formation of the O-stannyl ketyl 24 in this case arises from











0 Lewis [ 0 OSnBu3
A Acid
H-SnBu3 Bu3Sn L (6)
Polar ^(1 \ ^\\
Solvent H H

21 9 22 23



0 OSnBu3 OSnBu3
I AIBN TBTH
+ H-SnBu3 C6H6\ ------T + Bu3Sn- (7)
H

21 9 24 23 11

Scheme 1-5



the tributyltin radical reacting at the oxygen of the

carbonyl to form a carbon-centered radical. This radical

then abstracts a hydrogen atom to produce the tin alkoxide

species 23, and the tributyltin radical can now repeat the

process. It is this second type of reactivity, radical

addition to carbonyls, which will be the principal focus of

this dissertation.

Historically, rarely studied O-stannyl ketyls were first

described by Tanner et al.,35 and later by both Beckwith and

Roberts36 and Sugawara et al.37 in the synthesis of multiple

ring systems, discussed later. The tin ketyl can be

considered a pseudo-protected radical anion, where the O-Sn

bond has a large degree of ionic character due to

electronegativity differences (Scheme 1-6) The apparent

challenge to synthetic chemists is to develop methodology









which exploits both elements of reactivity which are

presented by this radical anion. Some of the work in Chapter

2 will address this challenge.


8- 5+
OSnBu3 O- SnBu3 O-SnBu3




24 25 26

Scheme 1-6



So far, this dissertation has discussed reductive free-

radical chain reactions and how they can produce carbon-

centered radicals, followed by hydrogen atom abstraction.

The next logical step is to examine their use in the

formation of carbon-carbon bonds, which is particularly

important because carbon-carbon bonds are the heart of

organic synthesis. Before one can use radicals in the

synthesis of complex molecules one must understand the

physical organic principles of free radicals. Some of the

factors which dictate how a radical will react include

orbitals, conformations, thermochemistry, and reaction rates.

The basic premise upon which thermochemistry is built is

that reactions which are exothermic (downhill in energy) are

favorable processes. Very often the behavior of radicals can

be rationalized from this standpoint. A good example is the

well known ability of oxygen-centered radicals to abstract

hydrogen. In this reaction a very strong O-H bond (-111









kcals)3 is formed at the expense of a weaker C-H bond (-99

kcals).3 Because a reaction is exothermic does not guarantee

that it will happen spontaneously, but it can be used as a

guide to see how feasible a transformation is.

The substituents on a radical also have a profound

effect on a radical's behavior. Giese noted that cyclohexyl

radical adds 8500 times faster to acrolein than to 1-hexene.38

In contrast, trifluoromethyl radical undergoes addition

reactions most efficiently with electron rich olefins such as

enamines and enol ethers.2 These results can be explained

using frontier molecular orbital theory.39 The singly

occupied molecular orbital (SOMO) of a radical interacts with

either the lowest unoccupied molecule orbital (LUMO) (Figure

1-2) or with the highest occupied molecule orbital (HOMO)

(Figure 1-3) of the alkene.2 In the trifluoromethyl example,

the inductive, electron-withdrawing effect of the fluorines

lowers the SOMO energy such that there is better orbital

overlap with the low-lying HOMO of the alkene.








--LUMO-
CH SOMO







Figure 1-2

Reaction of cyclohexyl radical with electron-poor alkene
III






C6H;-1 SOMO ... ,




Figure 1-2

Reaction of cyclohexyl radical with electron-poor alkene










S%

CF3. SOMO -

44, HOMO
S' EDG





Figure 1-3

Reaction of trifluoromethyl radical with electron-rich alkene


Another factor which influences the regioselectivity and

the stereochemistry of radical cyclizations is molecular

geometry. The regioselectivity of radical transformations

often complements or improves on the results which are
available through ionic reactions. 5-hexenyl radicals (29)

cyclize predominately to give five-membered rings (Scheme 1-

7),40 whereas cationic cyclizations yield six-membered rings.1

Conjugate addition of radicals to activated olefins occurs

solely at the B-carbon, but anionic conjugate addition often

shows a competition between 1,4 and 1,2 addition.1




+0

50 to 1

29 30 31


Scheme 1-7









The stereochemistry of 5-hexenyl radical cyclizations

has been thoroughly investigated by the Beckwith group.41 In

many additions to double bonds, such as cationic, attack

occurs at the center of a double bond, but radical reactions

are generally accepted to proceed via an early, unsymmetrical

transition state.42,43 Beckwith has formulated a model based

on the cyclohexane chair transition state (Figure 1-4) where

major products will be formed from the conformer with

substituents in the equatorial positions (Scheme 1-8).








2 2
1 13




Figure 1-4

Beckwith's chair-like transition state



Once the factors which influence radical reactivity have

been considered, the reactions in which they participate can

be examined. These reactions can be divided into two main

categories: intermolecular and intramolecular. The

intermolecular reactions are more difficult to accomplish

because the radicals must find the other reactive partner

before they are reduced by TBTH. This problem can be

circumvented by using either syringe pump or hydrideless tin

techniques to maintain a low hydrogen donor concentration.44












CH3


,CH3


3CH3




32


1 to 3


CH3




-o',
CH3
2 t


CH3


CH3


CH3





CH3
o 1


CH3





CH3


CH3


CH3


41


CH3


5 to


Scheme 1-8









The most basic intermolecular reactions that can occur

are recombination and disproportionation, which are both

radical-radical reactions. Recombination of radicals has

found some applications in synthetic chemistry. The Kolbe

electrolysis of carboxylate salts can lead to dimerization,45

or mixed couplings if one acid is used in excess.46

Disproportionation is the other pathway that two radicals can

take when they meet. Disproportionation and recombination

both occur when proponic acid is electrolyzed: butane is the

product of recombination, and ethane and ethylene are the

results of disproportionation.

Many intermolecular processes are simple substitution

reactions where the intermediate radical is quenched with

either a hydrogen source or a functional group.

Dehalogenation and deoxygenation reactions have become a

standard reaction in the armory of synthetic chemists2 and

have already been discussed in this dissertation. When the

thiocarbonyl methodology is applied to carboxylic acids, the

carboxylate can be replaced by many groups such as halogens,

chalcogens, and phosphorus groups.2


R R H
R= +___ TBTH
Y
Y Y

13 44 45 46


Scheme 1-9









The most important intermolecular methodology for the

synthesis of aliphatic carbon-carbon bonds via radical

reactions is the addition of alkyl radicals to alkenes1

(Scheme 1-9). For these reactions to be successful there is

a selectivity requirement which must be fulfilled. The

selectivity requirement pertains to the intermediate radicals

11, 13, and 45. These transient radicals all have specific

partners that they must react with for a successful synthesis

to occur. If A4 and 13 have the same tendency to add to

alkenes, then the reaction will result in polymerization.1

This problem can be avoided by careful selection of the

alkene appendage so that 13 and 45 have different

selectivities. To avoid polymerization the electronic

characteristics of the substituents on 13 and 4A5 must be

opposite in nature. An electron donating group on 13, (such


R

4 y Bu3SnH
44 9



R- 13


Bu3Sn-X
47
R H
Bu3Sn*
11 46
Y

R-X
12

Scheme 1-101








as a simple alkyl radical) and an electron withdrawing group

on the alkene is a popular combination (eq. 8).47 When these

appendages are used 13 becomes a nucleophilic radical which

prefers to add to electron deficient alkenes such as 49.

This gives A4 (Y=CN) which is now an electrophilic radical

because of the cyano group. Electrophilic radicals favor

adding to electron rich alkenes, therefore it will not add to

A4 (polymerization) and it will eventually be quenched by

TBTH. Equation 9 shows the other potential combination of

radical and olefin substituents.48



+' = TBTH (8
CN 95%


48 49 50




RO2C RO2 C
Rc2 + TBTH (9)
R0l + OC4H9 60% \ C4H9
RO2C RO2C OC4H9
51 52 53

Scheme 1-11



One of the main criticisms regarding radical chemistry

has been that due to the planarity of the intermediates their

application in chiral synthesis is impossible. Motherwell

rebuffs this argument by saying that "the planarity of

enolate anions, iminium ions and even free carbocations has

not hindered the effective operation of stereoelectronic








control elements leading to stereospecific reactions."2

Currently there are at least three groups working on the

control of acyclic stereochemistry, where the element of

control centers around the use of a chiral auxiliary on the

alkene. The Porter49 and Giese50 groups use the

dimethylpyrrolidine A4, and Curran and coworkers51 use sultam

55 as their chiral auxiliary. Scheme 1-12 shows how Porter,

Scott, and McPhail used the dimethylpyrrolidine auxiliary to

control the addition of cyclohexyl radical to alkene 56 which

produced 58 as the only diastereomeric product.52


0

N

/ 6


N
O=s
II


Figure 1-5

Chiral auxiliaries used by Porter, Giese, and Curran


EtO2C
CO2Et





56

56


CO2Et
'-CO2Et


5 hv
-780C


F-

Nli


Scheme 1-12








The intramolecular reactions that radicals can undergo

can be grouped into two different categories: rearrangements

and cyclizations. Rearrangements can take on many forms

ranging from simple hydrogen abstraction to

cyclopropylcarbinyl rearrangements. A popular intramolecular

reaction of radicals is a 1,5-hydrogen migration, which

proceeds through a six-membered transition state.53 An

interesting example of this reaction is by Rawal and

coworkers54 who recently reported the transformation shown in

Scheme 1-13. The olefin that is formed when the epoxide opens

eventually cyclizes with the carbon radical that is


S

0 N N /
TBTH,AIBN
= HO- H
0 C6H6,800C Me

59 64







O 0. OH HO H

H H Me



60 61 62 63


Scheme 1-13









formed when the alkoxy radical abstracts a hydrogen from the

8 carbon. The analogous 1,5-tin migration will be the

subject of Chapter 3 of this dissertation.

Carbon-carbon bonds can also be fragmented using

radicals. When a radical is generated a to a cyclopropane it

rapidly equilibrates to its homoallylic counterpart.2 This

ring-opening reaction has been used by Motherwell and

Harling55 to build bicyclospiro fused compounds (Scheme 1-14).

The radical which is formed by the fragmentation of the

thiocarboxylate opens the cyclopropane ring. This radical

then cyclizes onto the suitably disposed alkyne, and the

vinyl radical is eventually quenched by TBTH.56


S


0 N N TBTH,AIBN


C6H6,800C T
TMS


TMS 71% Yield
65 66

Scheme 1-14



As you can see from the examples above, radicals can

abstract hydrogens and fragment C-C bonds, but their main

function often lies in cyclization reactions. This thesis

will examine how carbonyls can be used to generate carbon-

centered radicals and the cyclizations that the resultant

radicals undergo. Prior to Enholm's studies very few papers









approached these reactions from a reagent-based standpoint.

The photolysis57 and electrochemistry58 of ketones differ from

the studies in this dissertation in that both are nonreagent

based methods which will be only briefly reviewed here.

When carbonyls are irradiated with light, an n-K*

transition occurs in which an electron from the nonbonding

orbital on the oxygen is promoted to X-antibonding orbital of

the ketone. The carbonyl in its excited state can either

abstract a proton from the solvent or an available hydrogen

donor, or it can fragment to give more stable molecules.

Fragmentation reactions can be divided into two categories,

type-I and type-II, based on the work of Ronald Norrish who

shared the Nobel prize in chemistry in 1967. In type-I

fragmentation reactions a carbonyl substituent is cleaved to

produce acyl and alkyl radicals. In most cases the acyl

radical is cleaved again to generate another alkyl radical

and carbon monoxide. The photolysis of 67 by Quinkert and

coworkers59 demonstrated this reaction, where subsequent

cleavage of the carbonyl appendages produced carbon monoxide

and a diradical species which coupled with itself to form 68.


Ph


O + CO

Ph
Ph
67 68 69


Scheme 1-15








The second type of Norrish cleavage happens when the

carbonyl compound being irradiated contains hydrogens on the
6-carbon. In this reaction the photoexcited carbonyl

abstracts a hydrogen from the 8-carbon through a six-membered

transition state to give a carbon-centered radical. The

diradical which is produced, 12, can either fragment to give

74 and 75 or couple to yield cyclobutanols.60 These

fragmentations can be distinguished by the bonds which are

broken. In type-I reactions the bond between the carbonyl and

the a-carbon is ruptured, but in type-II the Ca-Cp bond is

broken.


H OH
Shv OH Ring
Closure


71 72 73




0 I



74 75

Scheme 1-16



The electrochemistry of carbonyls has also shown that

they can effectively act as radical precursors.58 When

conducting solutions of ketones and aldehydes are

electrochemically reduced, there is a competition between

reduction to the alcohol and coupling of the intermediate









radical to give pinacol products. In recent years, however,

there have been successful attempts to trap the ketyl

radicals with a variety of appendages, such as allenes,

alkenes, alkynes (Scheme 1-17),61 and even benzene rings.58

The nature of the reactive species is sometimes clouded by

the conditions of the reaction, but it is generally believed

that either a ketyl (radical anion) or a protonated ketyl is

the reactive intermediate.


OH

S /// -e

H+

76 77

Scheme 1-17



Although sparse, there were two reports of the O-stannyl

ketyls being used in free radical cyclizations. In the

first, Beckwith and Roberts36 showed that tin ketyls can be

used to assemble bi- and tri-cyclic systems. The cyclization

(Scheme 1-18) proceeded with excellent yield, but they

remarked that the reaction was "sluggish" and required

additional TBTH and AIBN. Beckwith speculated that the

formation of the tin ketyl may be reversible or that either

the rate of the radical's initial formation or its subsequent

cyclization was slow. More likely, the problem with this

cyclization is that the nucleophilic tin ketyl would prefer

to react with a more electrophilic double bond.














90%
C6H6,A
C02Me C02Me

78 79

Scheme 1-18



More recent research indicates that cyclizations of tin

ketyl are effective and have potential applicability in

synthesis. In a second paper which preceded this work,

Sugawara and coworkers37 may have realized that an activated

alkene would solve the problems that Beckwith observed. They

used tin ketyls in the cyclization of aldehyde a8 onto the

activated olefin of a uridine ring system (Scheme 1-19).

Note that in this example, the skeleton is constrained to

prepare a six-membered ring, and the 6-exo closure of the 6-

heptenyl radical is known to be more than one order of

magnitude slower than the analogous 5-hexenyl cyclization.62

So an activated alkene is sometimes crucial to the success of

a radical cyclization.

Aside from these two synthetic reports, only the Enholm

group is actively reporting on studies directed towards the

use of tin ketyls in synthesis. This interest began in 1989

when Enholm and Prasad63 demonstrated that aldehydes and

ketones readily cyclized onto tethered olefinic appendages.

Although an unactivated olefin did cyclize, the yield was








low, and to achieve a synthetically useful yield the olefin

needed electron withdrawing groups, i.e., activation. They

also demonstrated the 6-exo cyclization of an activated

olefin, as shown in Scheme 1-20, where they produced cyclized

products in a 69% yield.


TBTH,AIBN


C6H, A


0 0

X


0 0


Scheme 1-19


CO2Me
82


TBTH,AIBN


C6H, A


Scheme 1-20


The next step for the Enholm group was to see if tandem

cyclizations were possible with tin ketyls. Enholm and

Burroff64 found that both spiro and fused bicyclic ring

systems could be synthesized with this methodology. In this


NH

.O


NH

1 O









study activated olefins were used in the first cyclization

and unactivated olefins were used in the second cyclization.

The activated olefin which is involved in the first

cyclization must act as both an acceptor and a donor in this

reaction. Also notice that intermediate radical 86 is

electrophilic and therefore needs a nucleophilic olefin for

effective cyclization. All of these criteria are met and the

reaction proceeds smoothly to give bicyclic products in a 75%

yield.



SPh
Ph Bu3:Sn* Ph
Ph 0 75%
Bu3SnO HO



85 86 87

Scheme 1-21



This dissertation investigated the free radical behavior

of tin ketyls which are functionalized so that the radical

could be delocalized or separated from the carbonyl. This

was accomplished by the use of unsaturated and epoxy ketones.

TBTH was used to generate the tin ketyl, and activated and

unactivated olefins were used to trap the radical

intermediates.

In Chapter 2 the behavior of c,P-unsaturated ketones and

their cyclizations and additions to activated olefins was

examined. Highly functionalized monocyclic and bicyclic









cyclopentanes were obtained. An interesting concentration

effect was observed which can be attributed to the

reversibility of the cyclization. This concentration effect

enabled the reactions to achieve excellent

diastereoselectivities (>50:1). Both intramolecular and

intermolecular additions were examined and their

applicability towards five-membered ring natural products was

considered.

Chapter 3 continues to examine the reactivity of the tin

ketyl with labile a-substituents. The reactions of a,P-epoxy

ketones with TBTH and their subsequent cyclizations onto

unactivated olefins were examined. These reactions, which

could form tetrahydrofuran products by addition of oxygen-

centered radicals to olefins, were found to undergo an

unusual 1,5-tin transfer to yield carbon radicals. These

radicals were useful, though, and underwent cyclizations to

create highly substituted cyclopentanes.

Once thought to be too unruly for the delicate world of

synthetic organic chemistry, free radicals have exploded onto

the scene. These intermediates have been generated under

mild conditions, have been shown to be very selective, and

have tolerated a wide range of functionality. Their

viability as a powerful synthetic intermediate has been

proven by Dennis Curran and others who have synthesized

entire families of natural products with these species. TBTH

has shown that it is a valuable synthetic reagent for a

variety of transformations. This work attempts to





27


demonstrate that carbonyls can be used effectively as radical

precursors with a flexibility that has not been illustrated

by either electrochemistry or photochemistry.















CHAPTER 2


CYCLIZATIONS OF L,P-UNSATURATED KETONES



The foundation for this project was established in 1989

by .the work of Enholm and Prasad.63 They studied the

cyclizations of O-stannyl ketyls e8 onto both activated and

unactivated olefins (Scheme 2-1). The stereoselectivity of

the ring closure ranged from 1:1 (anti:syn, relative to the

hydroxyl) to 3:1. A natural extension of this work was to

attach appendages to the O-stannyl ketyl and see if the

radical could be separated from the oxyanion by resonance.

The work presented in this chapter was published in a

preliminary form as a JACS Communication in 1991.65




TBTH,AIBN

R O C6H6,A R1 R
Ii OSnBu3 OH
R2 R2 R2

88 89 90
RI = Hydrogen or Alkyl
R2 = Alkyl or EWG

Scheme 2-1



An unsaturated appendage seemed a natural selection due

to many factors, which included ease of preparation and









effective delocalization through resonance. The general

transformation, shown in Figure 2-1, would involve the

coupling of the P-carbons of dieneone 91. This is a

difficult coupling because the 0-carbons in this system both

carry a partial positive charge. Prior to our studies there

were no reagent-based methods to accomplish this

transformation,66 but there were electrochemical methods.58,67







EWG

91

Figure 2-1

General transformation of a,P-unsaturated ketones



These methods suffer from some of the basic drawbacks

that are associated with electrochemistry. Only conducting

solvent systems can be used, and this is usually limited to

CH3CN/H2O. The conditions need to be closely controlled to

avoid unwanted pinacol and aldol products. Aldol and

saponification (EWG = CO2R) products result because basicity

builds up at the cathode. Often the specialized equipment is

not readily available to a practicing synthetic organic

chemist, or it may be expensive. Lastly, this method is not

selective enough to be applicable to natural product

synthesis which may contain numerous labile functional

groups. The stereoselectivities obtained are usually low, 2:1









(anti:syn), although in a single report, Little and Baizer67a

observed that the addition of CeCl3 improved

stereoselectivities (Scheme 2-2).





e +
CH3CN '


O OMe 0 OMe CO2Me CO2Me CO2Me CO2Me
92 93 94

Without CeC13 2.6 to 1
With CeCl3 15 to 1

Scheme 2-2



Electrochemists have provided some interesting details

on the physical properties of the radical anion ESR spectra

of enones, which do not possess acidic hydrogens. These

studies show that 50% of the radical density is located on

the p-carbon 97 (Figure 2-2), while the remaining portion is

divided equally between the carbonyl carbon 96 and the

carbonyl oxygen 95.66a,68



S0- 0




95 96 97

Figure 2-2

Major resonance contributors for the enone radical anion









Our studies were initiated by addressing the question of

whether we could obtain an efficient cyclization with

substrates such as BA where the carbonyl is replaced by an

a,P-unsaturated ketone to prepare 91. The ketyl, which is

produced by the reaction of enone 98 with tributyltin

radical, has two major resonance contributors 99 and 100. One

might speculate that, if an analogy with the electrochemical

resonance structures can be drawn, then I00 should be the

major contributor. Also, electronegativity differences

between 0 and Sn should make radical 1Q0 electron-rich. As

mentioned before, an electron-rich or nucleophilic radical

will prefer to react with an electron-deficient olefin.1



0 OSnBu3 OSnBu3
BU3Sn*

RFt R2 R1 R R2 R2

98 99 100

Scheme 2-3



It was first necessary to determine whether activation

of the olefin was essential for a successful cyclization.

When 101 was subjected to radical cyclization conditions the

only product 102 which was observed was a result of simple

reduction of the conjugated olefin. These results were not

surprising considering that the radical reduction of 0c,p-

unsaturated ketones is a precedented reaction.69 We had hoped

that both activated and unactivated olefins could be used in









this study, but this result showed that an activated alkene

was a critical element for success.





J C5H1H TBTH,AIBN C5H11 92%

SC6H6,A 0


101 102

Scheme 2-4



The reason why this cyclization did not work is

supported by the frontier molecular orbital theory which was

presented in Chapter 1. Enholm and Prasad63 also found that

tin ketyls were reluctant to cyclize with unactivated

olefins. The tin alkoxylate in both of the intermediate

radicals 24 and i00 impart some negative character onto the

radical, which raises the energy of the SOMO. Higher energy

SOMO's have better orbital overlap with the LUMO's of

electron deficient olefins (Figure 1-2).2 It is apparent

from these cyclizations that the intermediate radical did not

effectively cyclize and was quenched by TBTH. Therefore, the

rate of cyclization with unactivated alkenes is slow relative

to the rate of hydrogen abstraction..

Once it was determined that an activated olefin was an

important prerequisite for successful cyclizations, our

attention was then focused on the synthesis of the required

starting materials. Glutaric dialdehyde 103 seemed ideally









suited for the two-step attachment of unsaturated appendages.

But, what seemed to be a trivial Wittig reaction to make

aldehyde J10 became somewhat of a challenge due to the

predominance of the di-addition product. When 1 equivalent

of Wittig reagent was added by addition funnel to a large

excess (4 eqs.) of dialdehyde the product ratio still favored

the bis-product 7:1 (di:mono) This was only a minor

inconvenience because the di-addition product 106 was one of

the intended starting materials. Eventually, the slow

addition of ylide (1 eq.) by addition funnel to 7 eqs. of 103

gave the desired mono-addition product 104 in 74% yield. The

second unsaturated appendage was added without problems to

yield the needed starting materials (Scheme 2-5).


0

SEWG

Ph3P Ph3P


O O CH2C12 O CH2C12



103 104 105 106 107
74% EWG= -CN, -COMe, -C02Me
47%, 52%, 49%

Scheme 2-5



In a typical cyclization, the unsaturated ketone was

dissolved in benzene (0.1 M), then 3 eqs. of TBTH, and 0.1

eq. of AIBN were added. The mixture was degassed with Argon












TBTH,AIBN

C6H6,800C

CN


+ H


"" CN


COMe


108

3.0


(94% Yield)


TBTH,AIBN

C6H6,800C


+ HI'


"-- COMe


COMe


0 HO


110

3.5


TBTH,AIBN
C6H6,800C


OMe


111


to
(93% Yield)


+
CO2Me


COMe


0 0
113


112


to
(85% Yield)


Scheme 2-6


105


NH2
109


106


107


2.1









and then heated to 850C (bath temp.). The reactions were

monitored by thin layer chromatography (TLC), and in most

reactions the starting material was consumed within two

hours. These reactions never seemed to suffer from the

sluggish behavior that Beckwith reported,36 and they were also

very clean reactions. A summary of the results of these

cyclizations and their products are shown in Scheme 2-6. The

ratios and yields reported are from the isolated products and

are in good agreement with the GC ratios.

The proposed mechanism for these reactions is presented

in Scheme 2-7. The initiation steps which produced

tributyltin 11 radical were covered in Chapter 1, Scheme 1-2,

and will not discussed further here. Tributyltin radical 11

added to the ketone carbonyl of the starting material to

generate the O-stannyl ketyl 115. This ketyl radical, in

conjugation with the allylic double bond, has a resonance

form with the radical at the p-carbon 116. This resonance

form, a 5-hexenyl radical, subsequently cyclized with the

activated olefin. The resultant radical species 117

underwent hydrogen atom transfer with a second molecule of

TBTH, generating another tributyltin radical which continued

the chain process. Tin enolate 118 was then quenched by a

proton source to yield the final cyclized product 11..

The minor bicyclic syn products were formed by a second

cyclization. The EWG used to activate the olefin could also

undergo a second intramolecular reaction with the tin

enolate, as in 121. When the appendages were anti to one









another they were not close enough to react and the anti-ring

fusion in 118 is too strained,70 but when they were syn a

second aldol-like cyclization occurred (Scheme 2-8). These

products demonstrate that the dual reactivity of tin ketyls

can be utilized. They undergo an efficient radical reaction

followed by a smooth two electron condensation reaction.


Bu3Sn*


EWG


114


H+


0

119


EWG


EWG EWG
OSnBu3 OSnBu3

118 117


Scheme 2-7


Syn
---- H


O NS
SnBu3


120


O N
SnBu3


SnBu3


121


122


Scheme 2-8


H









Structural identification of the bicycle syn products

was not always a trivial pursuit. It was believed that the

bicycle syn product 109 from the cyclization of nitrile 105

would be an imine which could be hydrolzed to 113.

Unfortunately, this compound resisted attempted hydrolysis

with 1 M HC1. After comparison with standard library spectra

and conversations with Dr. John Greenhill it was decided that

the actual structure was the keto-eneamine shown in Scheme 2-

6.

The syn product from the cyclization of ester 107 was a

1,3-diketone. The literature H1 NMR spectra of 2-

acetylcyclopentahone showed contributions from all three

tautomeric forms (Scheme 2-9). 123 was the most stable form

and has a methyl group at 1.94 ppm. The methyl group of 125.

was assigned a chemical shift of 2.21 ppm, and the enolizable

proton was barely visible at 3.35 ppm. 113 had a very strong

singlet at 2.06 ppm, and a very small singlet at 2.22 ppm,

where the ratio is about 10:1 respectively from 1H NMR

integration. This analogy was not strong enough to

accurately identify the tautomer, therefore, its structure

was unknown.






OH O 0 O O OH

123 124 125


Scheme 2-9


















I"-'11111i HRF





HB


Proton(s) Percent Effect Observed
Irradiated HA HB HC HRF H6H CH3(A) CH3(K)

HA 30.1 6.0 2.7 2.1 *
HB 26.1 6.0
He 2.8 7.8 2.5 *
HRF 1.5 7.2 0.76
H6H
CH3(A) 2.8 2.2
CH3(K) 2.4 3.01

= Proton difficult to observe due to proximity with irradited proton.
1 = Molecule can adopt a conformation which could give these results.

Figure 2-3


Difference NOE data for 111









The syn product 111 from the cyclization of ketone J10

was surprisingly isolated as a single diastereomer. The

structure was confirmed by extensive NOE difference studies

summarized in Figure 2-3. Additionally, the hydroxyl

proton's chemical shift was independent of the concentration

of the sample. Therefore, it was believed to be

intramolecularly hydrogen bonded to the ketone. It is

interesting to note that in this product, four stereogenic

centers were created, however only isomer 111 was observed.

Particularly disturbing to us was the anti-

stereochemistry of the appendages which stood in marked

contrast to nearly all other 5-hexenyl radical

cyclizations.1,2,13 Normal radical cyclizations which produce

substituted cyclopentanes conform well to the classic

Beckwith model.40 Beckwith's model (Fig 1-4) suggested that a

1-substituted 5-hexenyl radical gave mostly syn products; our

results conflict with his model. An explanation of these

deviations are based on several differences between the two

radical chair-like systems. Beckwith's system (Scheme 1-8)

involved the cyclization of a radical with an alkene trap,

and essentially this reaction is kinetically controlled and

irreversible. The product ratio was determined by which ever

isomer formed the fastest, and was said to be under kinetic

control. However, our cyclization involved a resonance

stabilized allylic radical reacting with an activated olefin.

The resonance stabilization of this radical made the reaction

reversible and allowed the less stable syn product to









equilibrate and form the more stable anti isomer; this is

thermodynamic control. It could also be argued that the high

degree of polarity that 105-107 possesses is not adequately

represented by the bare skeleton of Beckwith's model.

If our system was indeed under thermodynamic control,

then we believed that varying the concentration and the

amount of TBTH would lead to improved stereoselectivities.

When the reaction was examined at greater dilution, we were

delighted to obtain much higher levels of stereoselectivity.

Thus, 107 was cyclized at three different dilutions as shown

in Scheme 2-10. Increasing levels of anti-stereoselectivity

for the ring appendages were obtained as the reaction was

diluted. A ratio of greater then 50:1 for the anti:syn

products could be achieved at 0.01 M in benzene.




TBTH,AIBN + H,- ..H
C6H6,800C "-CO2Me

0 O OMe COMe
OH 0

107 112 113



Reaction Conditions 112:113 (% Yield)

1. 1.00 M in benzene, 1.1 eq. TBTH 9:1 (75%)

2. 0.10 M in benzene, 1.1 eq. TBTH 25:1 (82%)

3. 0.01 M in benzene, 1.1 eq. TBTH >50:1 (82%)


Scheme 2-10









This dramatic increase in anti-stereoselectivity can

likely be attributed to the reversibility of the cyclization

and the decreased availability of TBTH. Once the cyclized

radical 117 reacted with a second molecule of TBTH it was no

longer able to open up to 116. Making the reaction more

dilute and decreasing the equivalents of TBTH allowed the

intermediate radical 117 more time to equilibrate before it

was quenched. By the time the radical was quenched the

equilibrium would have shifted from the kinetic syn product

to the thermodynamic anti product.

It could be that Beckwith's model was applicable to this

system, and that with even more concentrated solutions and an

excess of TBTH the syn isomer could be the major product.

However, it is difficult for a intermolecular quench to

compete with an intramolecular equilibrium and it is doubtful

that these conditions could be found. The more likely

scenario was that Beckwith's model was not entirely

applicable to this very polar substrate, and that ratios

observed at high concentration may be the kinetic product.

Regardless, the dilution study showed that this reaction was

reversible and that excellent anti-stereoselectivities could

be achieved.

Now that the one-electron chemistry of the tin ketyl had

been explored, we attempted to establish the ability to

conduct two-electron chemistry as well. Although this has

already been demonstrated by the isolation of the minor

bicyclic products, two additional experiments were performed









which clearly establish the presence of the stannyl enolate.

Since these reactions were not run in a strictly tandem

sequence, we use the term "serial reactions" to refer to

these studies. In each experiment, the one-electron

reactivity of an allylic O-stannyl ketyl was induced to

cyclize under the dilute anti-selective conditions (run # 3)

as shown in Scheme 2-10. The resultant tin enolate was then

immediately quenched in the same pot with Br2 or D20 to

produce a ca. 2:1 mixture of 126 and 122 as shown in Scheme

2-11. From 1H NMR integration of the methylene group a to

the ketone deuterium incorporation was calculated to be

greater then 85%.


1.TBTH,AIBN
C6H6,A


2. Quench with CO2Me

0 0 OMe Br2 of D20 COMe

107 126 Z=Br (86%)

127 Z=D (87%)

Scheme 2-11



These results and those for the previously discussed

bicyclic products clearly demonstrated the presence and

utility of the tin enolate. Also the monocyclic products and

dilution studies showed that the radical cyclization of

allylic O-stannyl ketyls could be both efficient and highly

stereoselective. Collectively these studies show that the









one-electron reactivity in the allylic O-stannyl ketyl can be

separated from the two-electron chemistry by sequential

transformations, and by the correct choice of experiment.

Thus, both types of reactivity can be achieved.

Now that the effectiveness of a,P-unsaturated ketones

was demonstrated on an intramolecular level, the next task

was to see if it was equally effective on an intermolecular

basis. The electrochemistry of a,P-unsaturated ketones has

shown that intermolecular coupling is a promising route to

highly functionalized cyclopentanes, and they have termed

these reactions electrohydrodimerization EHD.58'67 This name

was derived from the overall transformation where the

products represented dimerization of the starting material

with the addition of two hydrogens.

Often many problems are encountered when one goes from

intramolecular reactions, where the course of the reaction

can be controlled by the prudent selection of starting

materials, to intermolecular reactions which have a great

deal of freedom to react along many different pathways.

Coupling of unsaturated ketones could occur from both the

carbonyl carbon (head) and the 3-carbon (tail). This meant

that three different coupling products were possible: head to

head (pinacol), head to tail, or tail to tail. The

hydrodimerization of enone 128 (Scheme 2-12) exemplified the

frustrating product mixture which could be produced.71

Electrochemists believe that these reactions go through








either the combination of two radical anions or the addition

of a radical anion to the starting enone.67c


-- O


e CH3CN OH
Bu4NBF4 (
HO

-2 2 -2
128 129 130 131

52% 31% 16%

Scheme 2-12



A preliminary attempt to hydrodimerize methyl vinyl

ketone under radical conditions produced a low yield of a

compound whose 1H NMR spectra resembled an

acetylcyclopentanol. This hydrodimerized product further

reacted through its tin enolate to give the cyclopentanol.

The hydrodimerization of cyclohexenone resulted in the

recovery of 37% of the starting material's mass in form of

hydrodimerized products. The failure of these reactions to

achieve usable yields was most likely due to either the

volatility of the starting materials and products or the

lifetime of the radical was not long enough to allow for

coupling.

Both of these potential problems were circumvented in

the hydrodimerization of trans-chalcone 132. The reaction

conditions and product ratios are shown in Scheme 2-13.








Notice the delicate balance that existed between the

concentration and the equivalents of TBTH. Reaction 1 showed

that, if too much TBTH was used, simple reduction of the

olefin would be the major product 133. In an effort to

increase the amount of hydrodimer 134 that was formed, the

reaction was run at even higher concentration with the

minimum amount of tin required. We were delighted to find

that the previous ratio was reversed and that we were able to

isolate hydrodimer 134 in a 72% yield. It was isolated as an

inseparable mix of isomers whose spectra and melting point

agreed with published reports.72


0 O O HO

Ph TBTH,AIBNr Ph + Ph
I C6H6,800C
Ph Ph Ph Ph


132 133 134



Reaction Conditions 133:134 (% Yield)

1. 0.5 M in benzene, 1.50 eq. TBTH 4:1 (93%)

2. 1.0 M in benzene, 0.55 eq. TBTH 1:3 (94%)

Scheme 2-13



A proposed mechanism for the hydrodimerization of trans-

chalcone is shown in Scheme 2-14. 134 was formed by the

initial 03- coupling of the allylic O-stannyl ketyl 135 with





46


a molecule of trans chalcone, and resultant radical 136 was

quenched with TBTH to give tin enolate 131. This enolate

then condenses intramolecularly on the ketone to give tin

alkoxide 138 that upon hydrolysis yielded the final

hydrodimer 134.


0 OSnBu3 OSnBu3 0

Ph B3 Ph 1 Ph Ph

Ph Ph Ph Ph

132 135 136
TBTH





TBT*

O HO Ph 0 OSnBu3 OSnBu3 O
Ph

Ph H Ph Ph Ph


Ph Ph Ph Ph Ph Ph


134 138 137

Scheme 2-14



Although hydrodimers were achieved by this method it was

believed that the scope of this reaction was not sufficiently

broad to make it synthetically useful. Stabilization of the

allylic O-stannyl ketyl radical with phenyl groups seemed to

be imperative to the success of this reaction. The tentative

balance between reaction concentration and TBTH concentration








also limited this reaction. This problem could be avoided by

the use of either hydrideless tin sources or syringe pump

techniques. Both of these methods keep the concentration of

hydride donor at very low levels, but if these reactions were

run at high reaction concentration and low hydride density

then the most likely products would be polymers.

In conclusion, the intramolecular coupling of a,P-

unsaturated ketones to activated olefins can be applied to

the synthesis of substituted cyclopentanes. Excellent anti

stereoselectivities can be achieved with the proper selection

of reaction conditions. Additionally, the dual reactivity of

the allylic O-stannyl ketyl radical can be separated and

utilized in sequential one- and two-electron reactions.

Although, the intermolecular coupling of unsaturated ketones

was demonstrated, additional work is needed for this

methodology to be synthetically useful. This additional work

should be focused on hydrideless tin and syringe pump

techniques, also the prospect of cross-coupling reactions

should be investigated. Collectively, this work has

illustrated that Ct,-unsaturated ketones are a viable radical

precursor, and the degree of functionality and stereocontrol

which can be achieved allows this methodology to be applied

in the synthesis of natural products.















CHAPTER 3


CYCLIZATIONS OF (X,P-EPOXY CARBONYLS



In continuation of our studies directed towards the use

of carbonyls as free radical precursors, we examined the free

radical ring-opening reactions of epoxy ketones and

aldehydes. The radical ring opening of cyclopropanes is

known to be an extremely rapid process,73 likewise, the

analogous opening of epoxides is thought to be equally

rapid.2 The main difference between these two system is the

radicals which can be produced. The cyclopropyl case can

yield only carbon-centered radicals, while the epoxide

example can afford either oxygen-centered or carbon-centered

radicals, depending on which bond of the epoxide fragments.

The direction of fragmentation is largely controlled by

the nature of the R1 substituent (Scheme 3-1). Jorgensen74

has suggested that the bond dissociation energy of the C-O

bond is lower than the C-C bond except when stabilizing

groups are present to lower the energy of the C-C bond. When

this is a radical-stabilizing group such as a vinyl or aryl

group, then fragmentation of the C-C bond leads to a low

energy and delocalized carbon-centered radical. When no

stabilizing groups are present, cleavage of the weaker C-O

bond will be the predominant reaction. Synthetic methods








have been developed which utilize both of these fragmentation

pathways.



C-C Bond C-0 Bond.
Cleavage Cleavage
R1 Rl Ri

141 142 143

Scheme 3-1



The route which has received the least attention has

been the fragmentation of the carbon-carbon bond to give

stabilized radical 141. The reason why there has been less

research focused on this methodology is because the

requirement of a radical stabilizing appendage tends to limit

its applicability. Additionally, some researchers have found

that the presence of either vinyl75 or aryl76 substituents do

not guarantee the formation of carbon-centered radicals.

But, it is generally accepted that the presence of these

groups will lead to cleavage of the C-C bond.77

Feldman and Fisher75 used this fragmentation to generate

functionalized THFs. Radical 147 was generated by addition

of phenylthio radical to epoxyalkene A44. The epoxide ring

was opened by cleavage of the C-C bond to give stabilized

radical 148, which reacted with the activated olefin to give

adduct 149. This radical then cyclized onto the olefin which

was created by the epoxide opening, and the final product was

produced by elimination of thio phenyl radical.











A Ar + /EWG Ph2S2,AIBN
0 C6H6, hV
EWG
144 145 146


PhS. -PhS*

SPh
Ar 0
145 0

hAr EWG
EWG
147 148 149 150

Scheme 3-2



The syn-2,5 products were the only diastereomers which

were observed, and this conformed to the stereochemistry

which was predicted by Beckwith's model (Figure 1-4). Very

little control at the C-4 position was demonstrated, but this

can be blamed on the epimerizability of this center and not

necessarily on the cyclization. This reaction represents a

formal [3 + 2] addition where essentially both electrons of a

carbon-carbon bond are used to make the new bonds. This is a

very interesting approach to the synthesis of THFs and

demonstrates that all retrosynthetic disconnections should be

explored.

The majority of the classical work in the late 1960s and

early 1970s dealt with the opening of epoxides, mostly

concerning the pathway which produced oxygen-centered radical

14a. In these papers, there are a wide variety of methods









for the production of the a radical, and many different

synthetic and mechanistic goals. The common thread which

runs through these papers is that a radical is generated a to

an epoxide which causes the epoxide ring to open, and the

resultant alkoxy radical helps accomplish their

transformation. Some of this work, which pertains to this

dissertation, will be reviewed here.

If radical 143 was simply quenched by hydrogen atom

transfer with TBTH then allylic alcohols would be the

product, and many papers have utilized the fragmentation in

this manner.78 This transformation could be thought of as a

convenient alternative to the Wharton reaction,79 where epoxy

ketones were reacted with hydrazine to give allylic alcohols.

It was discovered in 1971 by H. C. Brown and coworkers78a when

alkyl radicals from organoboranes added to vinyl epoxides to

form allylic alcohols.

More recently in 1981, Barton and coworkers78b found that

product formation was highly dependent on the reaction

conditions. The fragmentation of epoxide 151 under "normal

addition" conditions did not give the expected allylic

alcohol as the major product, but instead, gave ketone 153

which resulted from an alkyl migration caused by p-scission

of the C-C bond. "Normal addition"80 refers to TBTH and AIBN

being added dropwise to a refluxing solution of starting

material, which effectively maintains a low concentration of

hydride donor. This low concentration of hydride donor

allows the intermediate radicals to do other chemistry.
















J TBTH,AIBN
/--^S^^ S<^'^------------^
0 Toluene,A





151


Reaction Conditions


1. 2 eqs. TBTH, Normal Addn.

2. 9 eqs. TBTH, Inverse Addn.


OH+
OH H


152


153


152:153 (% yield)

1 : 2.0 (71%)

4.3 : 1 (72%)


Scheme 3-3



Barton and coworkers felt that, if they could maintain a

very high concentration of hydride donor, they would be able

to isolate the allylic alcohol in useful yields. The

"inverse addition" method that they used maintained a very

high concentration of TBTH by dropwise addition of the

starting material to a refluxing solution of TBTH (9 eqs.)

and AIBN. The high concentration of hydride donor quenched

the intermediate alkoxy radical before it could do other

chemistry. Other examples in this paper showed that "inverse

addition" helped avoid unwanted side reaction products which

were formed when the "normal addition" method was used.

However, these unwanted products showed that alkoxy radicals

can do a great deal more then simple hydrogen abstraction,

and groups began to explore their potential.









Johns and Murphy81 were intrigued by some of the

products which were described in Barton's paper and set out

to see if the alkoxy radical could be used in the synthesis

of tetrahydrofurans. Starting material 154 is easily

prepared from commercially available geranial. To keep the

concentration of TBTH at a low level they used the "normal

addition" process and isolated THF products in pretty good

yields (Scheme 3-4). Although no ratios were given, they

mentioned that in the R=n-butyl case, the major diastereomer

had an anti-relationship between the vinyl and the isopropyl

group, which was in agreement with the major products formed

in Beckwith's studies (Scheme 1-8). Interestingly, bicyclic

product 156 was formed by a tandem radical cyclization where

the radical formed after the first cyclization reacted with

the vinyl group in a second cyclization.


R S


0 N N R R
STBTH,AIBN
S+
C6H6,A
O -





154 155 156

R = Me 53% 14%

R = n-Bu 63% 22%

Scheme 3-4








Alkoxy radicals are well known for their ability to

abstract hydrogen atoms through a six-membered transition

state from neighboring atoms which bear hydrogens, which was

demonstrated in the classic Barton reaction.82 There was no

mention of any products derived from this pathway in the

reduction of 154. The only hydrogens which could have been

abstracted through a six-membered transition state were next

to the olefin, and these were not accessible due to the trans

geometry of the olefin. However, some studies have used this

prominent reaction of alkoxy radicals to generate polycyclic

compounds.

Rawal and coworkers54 used the hydrogen-abstracting

ability of alkoxy radicals to facilitate the cyclization

shown in Scheme 3-5. Addition of tributyltin radical to the

thiocarbonyl of 151 caused it to decompose to radical 160.

The epoxide ring was then cleaved by this radical to produce

alkoxy radical 161, which then abstracted a hydrogen atom

from the 8-carbon. The resulting stabilized radical 162 was

then in position to cyclize onto the olefin produced from the

fragmentation of the epoxide.

Bicyclic products were formed in a 69 % yield, and the

ratio between the diastereomers 158:159 was 2.7:1. Product

159, which has the ester pointing into the cup of this cis-

fused bicyclic, was epimerized to the more stable 158 by

treatment with catalytic amounts of t-BuOK. This method also

produced successful cyclizations on other cyclic and acyclic








systems. Good results were also achieved when alkyl

appendages, with no radical-stabilizing groups, were used.


S
SH CMe H CO2Me

N 0 TBTH,AIBN
SC2Me C6H6,A

OH OH


157 158 159





I I C02M e
CO2Me




CO2Me OH
160 161 162

Scheme 3-5



Kim and coworkers83 used the addition of tributyltin

radical to vinyl epoxides in much the same way. The addition

of tin radical to the vinyl group of 163 provided the radical

which opened up the epoxide, which then abstracts a hydrogen

from the s-carbon. This radical then cyclized back onto the

olefin which was created when the epoxide was fragmented, and

the final product is formed by ejection of tributyltin

radical. Kim's group also had success with alkyl appendages

that contained both electron withdrawing and electron

releasing substituents.











SR TBTH,AIBN
0 C6H6,A

OH

163 164

Scheme 3-6



In the same paper, Kim and coworkers reported similar

cyclizations which occurred through a rare 1,5-tin transfer.

This mechanism was first proposed by Davies and Tse84 who did

ESR spectroscopy on the reactions of glycidyl tributyltin

ethers 165 with t-butoxy radicals. These studies found that

the predominant radical in solution was 168. The authors

reported that this was the first time that a 1,5-

organometallic transfer had been observed. This mechanism

seemed suspicious at first, but we now believe that our

reactions are occurring by the same mechanism, which will be

discussed later.


M M

H t-BuO* O ? 0

R R
00 R

165 166 167 168

M = SnBu3
R = H or alkyl


Scheme 3-7








Other mechanistic studies have looked at the radical-

induced opening of epoxides from another angle. These papers

were more concerned with the stereoelectronic factors which

governed the opening of epoxides. Bowman and coworkers85 used

the norborane skeleton as a template for their reactions.

They envisioned that the exo-oxiranes 169 would lead to C-O

bond cleavage products 1I1, while the endo-oxiranes 172 would

produce C-C bond fragmentation products 174. This prediction

was based on the stereoelectronic control observed for the

radical ring opening of analogous cyclopropyl and cyclobutyl

systems.86 It was proposed that the transition state required

maximum overlap of the radical's SOMO with the a-bond which

was undergoing scission.




_R R_ R
R R,
O R O*

169 170 171




R R

0 0 RI
R1 RI
172 173 174

R = H or OSnBu3
RI = H or Aryl

Figure 3-1

Stereoelectronic model for Exo- and Endo-epoxides








Bowman and coworkers envisioned this system as the

perfect test for these stereoelectronic factors. Many of the

current papers dealing with epoxide openings invoke this as a

potential mechanism, but it had never been tested. However,

Bowman found that no stereoelectronic effect could be

observed. When aryl derivatives of 169 and 172 (R=H and

R1=Aryl) were generated from the corresponding bromides, no

products from the cleavage of the C-O bond could be isolated.

Additionally, when non-aryl derivatives of 169 and 172

(R=OSnBu3 and R1=H) were created from the corresponding

ketones, no products from the cleavage of the C-C bond could

be found. They-concluded that no stereoelectronic effects

were operational, and that the determining factor was the

strengths of C-C and C-O bonds.

Many of the twists and turns which were observed in the

works reviewed above, also found their way into the study

which is presented in this chapter. Many of these diversions

were fully anticipated because of the well-known ability of

alkoxy radicals to abstract hydrogen atoms. However, what

was not anticipated was that the rare 1,5-tin transfer

mechanism would be the proposed mechanism. This project

started out as an extension of my oral research proposal,

which presented the idea of synthesizing THFs, such as 175,

through the fragmentations of epoxy ketones. However, the

fragmentation of epoxy ketone 176 which was used in this

project ended up yielding cyclopentanols 177.









0
COR
OH


TBT NUJ1BN TBTH,AIBN
0 H CsH6,A R





175 176 177

R = H, Me, n-Bu

Scheme 3-8



The molecule which is shown in Scheme 3-8 was very

similar to the system which Johns and Murphy81 used to

synthesize tetrahydrofurans in Scheme 3-4. The only

difference between the studies was that they used

thiocarbonyls to generate the radical and we used ketones.

One would expect that minor differences, such as the

generation of the initial radical, would have little effect

on the outcome of the reaction. However, so often when you

least expect something unusual, that is exactly what you get.

Anyway, we embarked on the study of the free-radical

fragmentation of epoxy ketones, with the hope of developing

methodology related to their cyclizations.

The synthesis of the starting materials for this study

were prepared from geranial, which either could be purchased

or was easily prepared by oxidation of geraniol, and is shown

in Scheme 3-9. The geraniol which was commercially available

was a mixture of cis and trans isomers in about a 1:2 ratio,





60


respectively. No attempt was made either to separate them or

to examine their individual reactivity.




R


PDC, CH2C12
4 A sieves

90%


RLi,THF
00C


R = Me 179 95%

= nBu 180 96%


PDC, CH2C12
4 A sieves


H202, K2CO3
MeOH, 0C


R = Me 182 71%

= nBu 183 39%


R = H 185 27%

= Me 186 80%

= nBu 187 98%


Scheme 3-9


177


178








The allylic oxidation of geraniol 177 proceeded smoothly

to produce geranial 2178 in 90% yield. The 1,2 addition of

organolithiums to geranial was carried out at 00 C in THF and

gave allylic alcohols 179 and 180 in very good yields. No

products from the 1,4 addition of the organolithium were

observed. The second series of allylic oxidations with PDC

did not proceed as smoothly, but gave unsaturated ketones 182

and 183 in reasonable yields. The alcohol-directed

epoxidation of the allylic alcohols was not attempted because

in a similar case that we had tried it was found that the

other olefin was sometimes epoxidized. So the epoxidation of

the unsaturated ketones was attempted by using a 30% aq.

hydrogen peroxide solution and K2CO3 as the base. This

protocol prepared the desired starting materials with overall

yields of 24% for 185 (R=H, 2 steps), 49% for 186 (R=Me, 4

steps), 33% for 187 (R=n-Bu, 4 steps).

In a typical cyclization, the epoxy ketone was dissolved

in benzene (0.5 M), then 2.5 eqs. of TBTH, and 0.1 eq. of

AIBN were added. This mixture was degassed with argon for 20

minutes, then the reaction was heated to 850 C (bath temp.).

The reactions were monitored by TLC, and in most cases the

starting material was consumed within a few hours. These

reactions were run at higher concentrations because at first

it seemed that the epoxide's opening may have been sluggish,

but epoxide openings were also' achieved at low

concentrations, such as 0.025 M for 185.








The conditions that the reactions were run at and the

ratio of products are shown in Scheme 3-10. The precursor

was used several times because it was easily accessible from

commercially available materials. However, the drawback to

this precursor was that the products which were formed were

generally a mixture of isomers due to four appendages on the

products. This mixture was not easily separated, and a

thorough identification of stereochemistry was not considered

possible. In general, it was believed that there were only

two isomers formed. This was less then might be expected

with three different chiral centers in the product, which

could give rise to four different diastereomers.


0

R 0 R
0
TBTH,AIBN
C6H6, A
HO






Substrate Conditions Major:Minor (% Yield)

186 (R=Me) 0.5 M, 2.5 eq TBTH 189:190
3:2 (83% Yield)

187 (R=nBu) 0.5 M, 2.5 eq TBTH 191:192
1:1 (81% Yield)


Scheme 3-10





63


The proposed mechanism at this point is shown in Scheme

3-11 for the methyl example. We believed that the products

shown in Scheme 3-10 were a result of the proficient ability

of alkoxy radicals to abstract hydrogens, such as in 194.

This would generate allylic-stabilized radical 195 which also

can be represented as resonance form 196. The 5-hexenyl

cyclization appeared to proceed smoothly yielding 192, which,

after additional reactions with TBTH and a proton source,

would yield the products shown in Scheme 3-10.


IBu3


Bu3Sn"


186


193


194


OSnBu3




OH


Pdts. 42


197


196


195


Scheme 3-11











For this mechanism to be operative there needed to be

hydrogens on the group which was attached to the ketone, as

in 192 in Scheme 3-11. If there were no hydrogens, then

radical 195 could never be generated. Compound 185 possessed

no abstractable hydrogens, and so its cyclization, or

inability to cyclize, would be a key test for this proposed

mechanism. When this reaction (Scheme 3-12) was carried out

we were surprised to find that 185 cyclized efficiently to

afford cyclopentanols.


0

R O R
0
TBTH,AIBN
C6H6,A
HO






Substrate Conditions Major:Minor (% Yield)

185 (R=H) 0.1 M, 2.0 eq TBTH 198:199
3:1 (81% Yield)

Scheme 3-12



At this point we realized the original mechanism that

was proposed (Scheme 3-11) was not possible. After careful

examination of the literature, the references to 1,5-tin

transfer (Scheme 3-7) were found,76b,83,84 which at first









seemed to be a bit obscure, but in time we realized that this

was indeed the operative mechanism. The proposed mechanism,

now revised as it applies to the cyclization of 185, is shown

in Scheme 3-13.


Bu3Sn*
--_----^


185


202


Bu3


H
16----


206


H
id- --


205


203


OSnBu3


204


Scheme 3-13




The deciding factor between these mechanisms was the

geometry of the olefin which was produced when the epoxide

was fragmented (Figure 3-2). If the alkoxy radical and the

tin enolate are cis to one another, as in 209, then 1,5-

migration can occur; but, if they are trans, then the









migration cannot occur. If a ketone bears a group with a

hydrogens then the alkoxy radical in 210 could abstract them

and the reaction could proceed by the mechanism in Scheme 3-

11.


0 0


R V< R


Bu3SnO H H OSnBu3

H H

207 208





Bu3SnO O0 H O0

H R Bu3SnO R

H H
209 210

Figure 3-2

Transition states model of cis and trans tin enolate



So the dilemma we were faced with was that there were

two possible mechanisms which the reactions discussed in this

chapter could follow. The initial reactions in Scheme 3-10

could go by either mechanism, while the reactions in Scheme

3-12 most likely followed the 1,5-tin transfer mechanism.

The tin transfer mechanism must proceed through transition

state 207 which is more sterically congested. We believe

that a weak chelation effect between the tin enolate and the








resulting alkoxy radical would help overcome any steric

problems. This six-membered ring chelate would eventually

develop into the transition state for the 1,5-tin transfer.

A new system was devised which would illuminate the reaction

pathway of the initial reactions in Scheme 3-10. The

synthesis of this compound is shown in Scheme 3-14.


0
1.DIBAL, CH2C12
O -780C
2.VinylMgBr,THF
0C
51% Yield
211


Dess-Martin
Periodinane,,
CH2C12

54%


213


OH mCPBA, CH2C2l

79%


212




CO2Me


0 Ph3P

0 CH2C12

68%

214


O CO2Me





215
215


Scheme 3-14









The low temperature reduction of E-caprolactone with

DIBAL produced a product which seemed to polymerize rapidly,

and was not characterized. Instead, after a crude

purification, it was treated with vinyl magnesium bromide

which yielded diol 212 in 51% yield for the two step

reaction. This diol was then epoxidized using mCPBA to give

a very polar product which was extremely hard to

chromatograph. PDC oxidation of diol 213 did not produce a

large amount of the desired product. This might have been

because the highly polar substrate was difficult to separate

from the PDC. The highly touted Dess-Martin Periodinane87 was

synthesized, and 2 eqs. of it was used for the oxidation

which proceeded smoothly to produce 214 in a 68% yield. The

final product -was synthesized by a stabilized Wittig reaction

on the aldehyde to give 215 in 10% overall yield in five

steps from caprolactone.

The reason the cyclization of 215 was important was

because it could help provide additional information about

whether the hydrogen abstraction mechanism is valid for these

fragmentations. If this cyclization succeeded in producing

cyclopentanes, then the alkoxy radical must be abstracting a

hydrogen; if it does not succeed, then the 1,5-tin transfer

mechanism must be valid. The starting material was dissolved

in benzene (0.01 M), then 1.5 eqs of TBTH and 0.1 eq. of AIBN

were added. The reaction was degassed with argon and the

reaction was refluxed. The only product which was isolated








was the opened epoxide 216. No cyclized products could be

isolated.


0 CO2Me 0 C02Me
TBTH,AIBN _

O C6H6, A
HO
70%
215 216

Scheme 3-15



This suggested that the mechanism in Scheme 3-10, or the

hydrogen abstraction mechanism, was a less-likely

possibility. We felt that hydrogen abstraction by the alkoxy

radical would generate a 5-hexenyl radical which would likely

cyclize onto the activated olefin. It should be noted that

hydrogen abstraction may have actually occurred but the

ketone stabilized radical was electron poor and could not add

to the LUMO of the alkene The probable mechanism which can

now be proposed is the 1,5-tin transfer which was originally

proposed by Davies.84 When this tin transfer occurred in the

reaction of 215 a radical was generated a to the ketone, but

too far away to cyclize with the activated olefin, and it was

simply quenched with TBTH to give the final product.

In conclusion, the work which is presented in this

chapter has shown that the fragmentation of epoxy ketones

proceeds by a relatively novel 1,5-tin transfer. The

products of these reactions are highly functionalized

cyclopentanols. Although the stereochemistry of this








cyclization could be improved, the number of isomers isolated

was far less then might be expected from such a densely

functionalized molecule. In this regard, radical

cyclizations were conducted on f18 and 187 with SmI2 and the

same products as mentioned above were isolated. However,

initial TLC showed that a much higher degree of

stereoselection (around 10:1) was possible. But the ketone

center turned out to be very epimerizable, and these

selectivities were never realized. Future work in this

direction could potentially yield a very a fruitful study.















CHAPTER 4

SUMMARY



The studies described in this dissertation are an

attempt to expand the realm of free-radical chemistry. Free-

radical methodology has been dominated by halogen and group

abstraction as the source of organic radicals. Although,

these methods have helped establish free-radical reactions as

a valuable tool in the synthesis of complex molecules,

hopefully, this thesis will show that there are quality

alternatives to these methods. The mild reaction conditions,

and the ability to control reactivity, stereoselectivity, and

regioselectivity in these reactions are certainly different

qualities from the original "unruly" reputation of free

radicals 40 years ago.

A problem associated with "classical" radical chemistry

is illustrated by the work of Dennis Curran shown in Scheme

1-1. In his synthesis of hirsutene, Curran took a molecule

which had three exploitable functional groups (LL and

produced a molecule which had only functional group (2). The

use of free-radical cyclizations on this molecule had

virtually stripped it of usable functionality. Although, the

loss of functionality was intended in the case of hirsutene,

it still demonstrates that sometimes radical cyclizations








result in compounds which can not be further functionalized.

A goal of this thesis was to show that radical cyclizations

could be conducted successfully with carbonyls as free-

radical precursors.

In the first study, TBTH was used to couple the P-

carbons of a,3-unsaturated ketones and activated olefins to

produce highly functionalized cyclopentanes. This study

revealed that unsaturated ketones were very effective at

delocalizing the radical away from the tin ketyl which was

formed when tributyltin radical added to the ketone.

Unactivated alkenes did not undergo cyclization reactions,

but instead underwent simple reduction of the unsaturated

ketone. Activated alkenes that were tethered to this O-

stannyl allylic radical participated in 5-hexenyl radical

cyclizations to form cyclopentanes. Excellent anti-

stereoselectivities were achieved (>50:1) when these

cyclizations were conducted at low concentrations. The

effect of concentration on the product ratios seemed to

indicate that these cyclizations were an artifact of a

reversible cyclization. These observations are in direct

contrast to how an ca,P-unsaturated ketone is normally viewed

in free radical reactions. These studies show that for the

first time that unsaturated ketones can serve as radical

precursors in a reagent-based study.

This study also demonstrated that the anionic character

of the tin ketyl can be trapped by electrophilic reagents.

All of the activating groups on the alkenes reacted with the








tin enolate in a second cyclization when both appendages were

syn to the newly formed cyclopentane. When the appendages

were anti to one another, ring strain would not allow them to

react, but electrophilic reagents such as bromine and

deuterium oxide were able to trap this anti intermediate.

This is important because the sequencing of one- and two-

electron reactions is rapidly emerging as an important

synthetic tool.87

Chapter 2 also demonstrated that the intermolecular

coupling of ia,P-unsaturated ketones was possible. Although

this methodology appeared very limited to substrates bearing

phenyl or radical-stabilizing groups, the implementation of

either syringe pump or hydrideless tin techniques to these

reactions may expand the applicability of this reaction.

The third chapter examined the reactions of epoxy

ketones. The original goal of synthesizing tetrahydrofurans

by this methodology was never realized. The well-known

ability of alkoxy radicals to abstract hydrogen from adjacent

carbons was not the problem with this project. It was a

little-known 1,5-tin transfer mechanism which kept the alkoxy

radical from cyclizing onto adjacent olefins. However, the

carbon-centered radical, which was produced by this tin

transfer, underwent cyclization reactions with tethered

olefins. The product of these cyclizations were highly

substituted cyclopentanols.

Collectively, these studies show that carbonyls can be

effectively utilized as free-radical precursors. Hopefully








this work will begin to establish the use of ketones and

aldehydes in that regard. The cyclizations of unsaturated

ketones and epoxy carbonyls have shown that these systems can

offer excellent stereoselectivities along with a greater

degree of manipulatable functionality. These cyclizations

are well suited towards the synthesis of complex, and highly

functionalized cyclopentanes, similar to those found in

natural products. They will hopefully become a new and

effective weapon in the arsenal of synthetic chemists.















CHAPTER 5

EXPERIMENTAL SECTION


General


Melting points were determined on a Thomas-Hoover

capillary melting point apparatus and are uncorrected.

Infrared spectra were recorded on a Perkin-Elmer 283B FTIR

spectrophotometer and are reported in wave numbers (cm-1). 1H

Nuclear magnetic resonance (NMR) spectra were recorded on a

Varian VXR-300 (300 MHz) spectrometer, and General Electric

QE-300 (300 MHz). 13C NMR spectra were recorded at 75 MHz on

the above mentioned spectrophotometers. Chemical shifts are

reported in ppm downfield relative to tetramethylsilane

((CH3)4Si) as an internal standard in CDC13. Mass spectra and

exact mass measurements were performed on Finnigan MAT95Q,

Finnigan 4515, or Finnigan ITD mass spectrometers. Elemental

analysis was performed by Atlantic Microlab, Inc., Norcross,

GA 30091.

All reactions were run under an inert atmosphere of

argon using flame or heat dried apparatus. All reactions

were monitored by thin layer chromatography (TLC) and judged

complete when starting material was no longer visible in the

reaction mixture. All yields reported refer to isolated

material judged to be homogeneous by thin layer








chromatography and NMR spectroscopy. Temperatures above and

below ambient temperature refer to bath temperatures unless

otherwise stated. Solvents and anhydrous liquid reagents

were dried according to established procedures by

distillation under nitrogen from an appropriate drying agent:

ether, benzene, and THF from benzophenone ketyl; CH2C12 from

CaH2. Other solvents were used "as received" from the

manufacturer.

Analytical TLC was performed using Kieselgel 60 F-254

precoated silica gel plates (0.25 mm) using phosphomolybdic

acid in ethanol as an indicator. Column chromatography was

performed using .Kieselgel silica gel 60 (230-400 mesh) by

standard flash89 and suction chromatographic techniques.

All GC experiments were performed on a Varian 3500

capillary gas chromatogarph using a J & W fused silica

capillary column (DB5-30W; film thickness 0.25 g).




Experimental Procedures and Results



(3E.8Z)-Tetradecadien-2-one (101)


To a previously dried 25 ml round bottom flask (RBF) was

added methyl ketone Wittig reagent (1.70 g, 5.31 mmol) along

with a magnetic stirrer. 5-Undecenal90 (0.40 g, 2.38mmol) was

weighed into a separate flask and transferred with 2.6 ml of

CHC13. Reaction was quenched with water and extracted with

Et20. The Et20 layer was washed with sat. brine soln., dried








over Na2SO4, and concentrated in vacuo. The concentrate was

purified by flash chromatography on a silica gel column to

yield a clear oil (0.40 g, 81.0%): Rf 0.75 (70% Et20/hexane);

300 MHz 1H NMR (CDC13) 6 6.81 (1H, m), 6.08 (1H, d, J=16 Hz),

5.38 (2H, m), 2.26 (3H, s), 2.24 (2H, m), 2.03 (4H, m), 1.54

(2H, m), 1.30 (6H, m), 0.89 (3H, t, J=7 Hz); 75 MHz 13C NMR

(CDC13) 8 198.58, 148.23, 131.45, 131.01, 128.61, 32.58,

32.01, 31.56, 29.41, 28.18, 27.29, 26.29, 22.60, 14.08; IR

(neat) 3006, 1700, 1628, 1459, 1253, 977 cm-1; MS (CI), m/e

(relative intensity) 209(m++1, 9), 150(20), 137(21), 97(35),

95(28), 84(25), 81(34), 69(28), 67(29), 43(100); HR MS (CI)

209.1910 (calc. for C14H250: 209.1905).


7-Oxo-5-octenal (104)91


Glutaric dialdehyde (100 ml of a 50% aq. solution, 552.3

mmol) was placed into a 300 ml RBF along with a magnetic

stirrer. The methyl ketone Wittig reagent (25.5 g, 80.1

mmol) was dissolved in CH2C12 (100 ml) and placed into a large

addition funnel. CH2C12 (50 ml) was added to the dialdehyde

in the reaction flask and the ylide solution was slowly added

to the reaction mixture. The addition of the ylide took

about 1.5 hours, and the funnel was rinsed with 10 ml of

CH2C12. The reaction was allowed to proceed overnight and was

extracted with H20 (2 x 100 ml). The organic layer was dried

with Na2SO4 and concentrated in vacuo. Column chromatography

of the residue produced a colorless oil (8.2784 g, 59.0 mmol,

74% yield): Rf 0.40 (70% Et20/hexane); 300 MHz 1H NMR (CDCl3)








6 9.79 (1H, s), 6.77 (1H, dt, J=16.0, 6.9 Hz), 6.09 (1H, d,

J=16.0 Hz), 2.52 (2H, t, J=6.9 Hz), 2.28 (5H, m), 1.84 (2H,

m); 75 MHz 13C NMR (CDC13) 8 201.51, 198.31, 146.65, 131.87,

42.98, 31.56, 26.95, 20.38.


General Procedure for Preparation of Activated Olefins


To a previously dried flask was added the appropriate

stabilized ylide (7 mmol) and a magnetic stir bar. CHC13

(3.5 ml) was then added and the reaction was stirred. Once

the ylide was completely dissolved, 7-oxo-5-octenal 104 (3.5

mmol) was added. Reaction was followed by TLC and starting

material was usually consumed in 1-2 days. The reaction

mixture was concentrated in vacuo, and this residue was flash

chromatographed to yield the activated olefin.


9-Oxo-2E.7E-decadienenitrile (105)


Yield (47%); Rf 0.55 (90% Et20/hexane); 300 MHz 1H NMR

(CDC13) 6 6.75 (2H, m), 6.09 (1H, dt, J=16.0, 1.4 Hz), 5.38

(1H, dt, J=16.0, 1.4), 2.27 (7H, m), 1.68 (2H, m); 75 MHz 13C

NMR (CDC13) 8 198.24, 154.81, 146.36, 131.93, 117.28, 100.55,

32.60, 31.51, 27.06, 26.08; IR (neat) 2933, 2863, 2222 (s),

1731, 1696, 1673, 1631, 1363, 1256, 976 cm-1; MS (EI), m/e

(relative intensity) 164(M++1, self CI,18), 148(35), 55(54),

43(100), 41(33), 39(37); HR MS (EI) 163.09857 (calc. for

C1oH13NO: 163.09971).








3E.8E-Undecadiene-2,10-dione (106)


Yield (52%); Rf 0.55 (50% EtOAc/hexane); 300 MHz IH NMR

(CDC13) S 6.78 (2H, dt, J=16.0, 7.5 Hz), 6.04 (2H, d, J=16.0

Hz), 2.37 (10H, m), 1.67 (2H, m); 75 MHz 13C NMR (CDC13)

198.27, 146.77, 131.78, 31.73, 27.02, 26.51; IR (neat) 2932,

1697, 1673, 1626, 1428, 1363, 1255, 1185, 977, 607 cm-1; MS

(EI), m/e (relative intensity) 180(M+, 1), 137(43), 81(23),

43(100); HR MS (EI) 180.1151 (calc. for CIIH1602: 180.11503).


Methyl-9-oxo-2E.7E-decadienoate (107)


Yield (49%); Rf 0.40 (70% Et20/hexane); 300 MHz 1H NMR

(CDC13) 8 6.95 (1H, dt, J=16.0, 6.9 Hz), 6.78 (1H, dt, J=16.0,

6.9 Hz), 6.08 (1H, dt, J=16.0, 1.5 Hz), 5.84 (1H, dt, J=16.0,

1.5 Hz), 3.73 (3H,s), 2.24 (7H, m), 1.67 (2H, m); 75 MHz 13C

NMR (CDC13) 8 198.29, 166.78, 148.13, 146.90, 131.63, 121.53,

51.34, 31.57, 31.38, 26.87, 26.28; IR (neat) 1723, 1697,

1674, 1627, 1436, 1362, 1272, 1256, 1202, 977 cm-1; MS (EI),

m/e (relative intensity) 197(M++1, self CI, 82), 196 (M+, 2),

165(80), 164(31), 137(75), 136(55), 121(42), 93(61), 81(89),

79(28), 68(32), 55(51), 53(52), 43(100), 41(41), 39(56); HR

MS (EI) 196.10913 (calc. for C11H1603: 196.10995).


General Procedure for Radical Cyclization of Olefins


A round bottom flask and a condenser were flame dried

and allowed to cool under a argon atmosphere. After the

flask had cooled, the activated olefin (0.5 mmol) was weighed








into the flask. Then AIBN (0.05 mmol), followed by TBTH (1.5

mmol) and freshly distilled benzene (5 ml) were syringed into

the flask. This solution was carefully degassed for 15 min

with argon. Then the reaction was warmed to 850C until the

reaction was complete by TLC. When no starting material was

observed by TLC, the reaction mixture was concentrated in

vacuo, and the mixture was separated by flash chromatography

to give the cis and trans cyclized products.


8Z-Tetradecen-2-one (102)


Yield (90%); Rf 0.77 (70% Et20/hexane); 300 MHz 1H NMR

(CDCl3) 8 5.38(2H, m), 2.42 (2H, t, J=8 Hz), 2.14 (3H, s),

1.97 (4H, m), 1.58 (2H, m), 1.30 (10H, m), 0.89 (3H, t, J=7

Hz); 75 MHz 13C NMR (CDC13) 8 209.0, 130.6, 129.9, 43.7, 32.3,

31.3, 29.4, 29.3, 28.8, 27.29, 26.9, 23.6, 22.5, 14.0; IR

(neat) 2926, 2855, 1719, 1462, 1358, 1286, 1161, 1075, 968,

723 cm-1; MS (CI), m/e (relative intensity) 211(m++l, 92.7),

193(32), 125(22), 85(41), 43(100); Anal. C14H260: 79.82% C,

12.40% H (calc. 79.94% C, 12.46% H).


trans-(2-(2-Oxopropyl)cyclopentyl)ethanenitrile (108)


Yield (71%); Rf 0.60 (90% Et20/hexane); 300 MHz 1H NMR

(CDCI3) 8 2.7-2.3 (4H, m), 2.14 (3H, s), 1.98 (3H, m), 1.77

(1H, m), 1.55 (2H, m), 1.44 (1H, m), 1.13 (1H, m); 75 MHz 13C

NMR (CDC13) 6 207.89, 119.15, 48.75 ,41.87, 40.28, 32.66,

32.06, 30.23, 23.55, 21.96; IR (neat) 2954, 2872, 2244 (s),

1713, 1452, 1424, 1359, 1294, 1227, 1174 cm-1; MS (EI), m/e








(relative intensity) 166(M++1, self CI, 20), 81(24), 58(36),

43(100), 34(48); HR MS (EI) 165.11593 (calc. for CloH15NO:

165.11536).


cis-2-Acetyl-3-iminobicyclo[3.3.01octane (109)


Yield (24%); Rf 0.40 (90% Et20/hexane); 300 MHz 1H NMR

(CDC13) 8 3.37 (1H, m), 2.70 (2H, m), 2.23 (1H, m), 2.10 (3H,

s), 1.88 (1H, m), 1.74 (1H, m), 1.53 (4H, m), 1.36 (1H, m),

1.24 (1H, br s); 75 MHz 13C NMR (CDC13) 8 196.19, 162.76,

110.90, 48.31, 41.29, 38.22, 34.87, 34.83, 27.95, 25.81; IR

(KBr) 3355, 2929, 2856, 1627, 1498, 1340, 1316, 1284, 1270,

926 cm-1; MS (El), m/e (relative intensity) 165(M+, 29),

136(100), 43(30), 34(52); HR MS (EI) 165.11526 (calc. for

C10H15NO: 165.11536).


trans-l-(2-(2-Oxopropyl)cyclopentyl)-2-propanone (110)


Yield (73%); Rf 0.70 (50% EtOAc/hexane); 300 MHz 1H NMR

(CDC13) 8 2.60 (2H, dd, J=18, 5 Hz), 2.36 (2H, dd, J=18, 9

Hz), 2.14 (6H, s), 1.89 (4H, m), 1.58 (2H, m), 1.25 (2H, m);

75 MHz 13C NMR (CDC13) 8 208.68, 48.97, 40.93, 32.35, 30.26,

23.50; IR (neat) 3602, 2949, 2871, 1713, 1407, 1359, 1274,

1230, 1176, 1154 cm-1; MS (EI), m/e (relative intensity)

183(M++1, self CI,100), 125(25), 124(45), 81(32), 43(95); HR

MS (EI, self CI) 183.13866 (calc. for C11H1902: 183.13851).








cis-2(R)-Acetyl-3(S)-hydroxy-3-methylbicyclo 3.3.01octane (111)


Rf 0.65 (50% EtOAc/hexane); 300 MHz 1H NMR (CDC13) 8

4.09 (1H, br s), 2.86-2.67 (1H, d, J=9.3 Hz), 2.23 (3H, s),

2.00 (1H, dd, J=7.8, 13.2 Hz), 1.80-1.55 (6H, m), 1.31 (3H,

s), 1.13 (1H, dd, 9.3, 13.2 Hz); 75 MHz 13C NMR (CDC13) 6

214.64, 82.81, 66.34, 48.28, 47.68, 42.04, 33.26, 32.84,

32.15, 26.21, 25.63; IR (neat) 3474 (br), 2943, 2862, 1689,

1453, 1372, 1237, 1176, 1138, 934 cm-l; MS (EI), m/e (relative

intensity) 182(m+, 0.1), 167(1), 124(71), 81(10), 66(20),

43(100); HR MS (CI) 183.1381 (calc. for C11H1902: 183.1385).


trans-Methyl(2-(2-oxopropyl)cyclopentyl)ethanoate (112)


Yield (58%); Rf 0.45 (70% Et20/hexane); 300 MHz 1H NMR

(CDCl3) 8 3.67 (3H, s), 2.65-2.17 (4H, m), 2.14 (3H, s), 1.89

(4H, m), 1.59 (2H, m), 1.23 (2H, m); 75 MHz 13C NMR (CDC13) 8

208.66, 173.66, 51.46, 48.91, 42.01, 40.76, 39.04, 32.33,

32.10, 30.27, 23.44; IR (neat) 2952, 2872, 1738, 1716, 1436,

1359, 1258, 1194, 1176, 1155 cm-1; MS (CI), m/e (relative

intensity) 199(M++1, 63), 167(92), 141(27), 81(60), 67(37),

43(100); HR MS (EI) 167.10746 (calc. for C11H1603 -OCH3:

167.10721).


cis-2-Acetyl-3-oxobicyclof3.3.01octane (113)

Yield (27%); Rf 0.65 (70% Et20/hexane); 300 MHz 1H NMR

(CDC13) 8 13.8 (1H, br s), 3.26 (1H, m), 2.70 (2H, m), 2.06

(3H, s), 1.92 (3H, m), 1.58 (2H, m), 1.45 (2H, m); 75 MHz 13C









NMR (CDC13) 6 202, 180.39, 115.27, 43.43, 42.93, 37.16, 34.80,

34.39, 26.04, 21.44; IR (neat) 2948, 2864, 1710, 1652, 1448,

1389, 1286, 1235, 937, 893 cm-1; MS (EI), m/e (relative

intensity) 166(M+, 76), 137(100), 124(30), 95(48), 43(98),

41(31), 39(32); HR MS (EI) 166.09905 (calc. for C10H1402

166.09938).


General Procedure for Enolate Trapping Studies


A round bottom flask and a condenser were flame dried

and allowed to cool under a argon atmosphere. After the

flask had cooled, activated olefin 107 (0..250 mmol) was

weighed into the flask. Then AIBN (0.025 mmol), followed by

TBTH (0.265 mmol) and freshly distilled benzene (25 ml) were

syringed into the flask. This solution was carefully

degassed for 15 min with argon. The reaction was warmed to

850C for one hour, then either D20 (XS) or Br2 in CC14

(.750mmol) was added. When no starting material was observed

on TLC, the reaction mixture was concentrated in vacuo, and

the mixture was separated by flash chromatography to give

mostly trans cyclized products.


Methyl(2-(l-bromo-2-oxopropyl)cyclopentyl)ethanoate (126)


Yield (86%); Rf 0.68 (90% Et20/hexane); 300 MHz 1H NMR

(CDC13) 6 4.45 (1H, d, J=7.5 Hz), 3.68 (3H, s), 2.47 (1H, dd,

J=7.5, 15 Hz), 2.38 (3H, s), 2.27 (1H, dd, J=9, 15 Hz), 2.3-

2.1 (2H, m), 1.90 (2H, m), 1.63 (2H, m), 1.45-1.2 (2H, m); 75

MHz 13C NMR (CDCl3) 6 201.96, 173.09, 60.27, 51.48, 46.41,








40.02, 38.98, 32.65, 30.16, 27.48, 23.81; IR (neat) 2953,

2871, 1736, 1436, 1358, 1257, 1198, 1166, 1086, 1014 cm-1; MS

(EI), m/e (relative intensity) 278(m+, 81Br,-), 276(m+, 79Br,-

), 165(11), 141(16), 123(15), 95(17), 81(18), 43(100); HR MS

(CI) 279.0417 (calc. for C11H1881BrO3: 279.0419).


Methyl (2- (1-deutero-2-oxopropyl) cyclopentyl) ethanoate (127)


Yield (87%), 85% Deuterium incorporation; Rf 0.60 (90%

Et20/hexane); 300 MHz 1H NMR (CDC13) 8 3.67 (3H, s), 2.62

(0.6H, dd, J=6, 18 Hz), 2.46 (1H, dd, J=6, 15 Hz), 2.35

(0.5H, dd, J=9, 18 Hz), 2.21 (1H, dd, J=9, 15 Hz), 2.14 (3H,

s), 1.89 (4H, m), 1.59 (2H, m), 1.23 (2H, m); MS (CI), m/e

(relative intensity) 200(m+, 25), 199(24), 168(100), 167(99),

149(3), 140(12), 139(9), 121(4).


Hydrodimerization of Chalcone


Chalcone 132. (0.2057 g, 0.9877 mmol) was weighed into a

previously dried 5 ml RBF. Freshly distilled Benzene (1 ml),

a magnetic stir bar, AIBN (7.0 mg, 0.04 mmol), and TBTH (0.15

ml, 0.5577 mmol) were added and the reaction was degassed for

15 min. After TLC showed that chalcone was consumed, the

reaction was concentrated in vacuo, and the residue was

separated by column chromatography. The simple reduced

product, 1,3-diphenyl-l-propanone,92 133 was obtained as a

colorless solid which melted at 69-700C (lit. 70-720C). The

hydrodimerized and cyclized product,72 134 was obtained as a








mixture of isomers which was a colorless solid which melted

at 93-1010C (lit. 91-990C).


4,8-Dimethyl-3.7-nonadiene-2-ol (179)93


Geranial (mixture of citral and neral) (2.0327 g, 13.353

mmol) was weighed into a 50 ml RBF, then THF (25 ml) was

added. The temperature was lowered to 0C and 1.5M MeLi

(13.5 ml, 20.25 mmol) was syringed into the reaction mixture.

After an hour, the reaction was quenched with a sat. NH4C1

soln., and then extracted with Et20 (3 x 25 ml). The ether

layer was dried over Na2S04 and concentrated to give a

colorless oil (2.1418 g, 12.728 mmol, 95.3% yield) which was

a mixture of cis and trans double bond (spectra reported for

major peaks of mixture): 300 MHz 1H NMR (CDC13) 8 5.22 (1H, d,

J=9 Hz), 5.09 (1H, t, J=7 Hz), 4.57 (1H, m), 2.08 (2H, m),

2.00 (2H, t, J=7 Hz), 1.88 (1H, br s), 1.69 (6H, s), 1.60

(3H, s), 1.22 (3H, d, J=8 Hz); 75 MHz 13C NMR (CDC13) 8

137.23, 130.23, 129.17, 123.90, 64.64, 39.39, 26.35, 25.58,

23.55, 17.60, 16.33.


7.11-Dimethvl-6,10-dodecadiene-5-ol (180)94


Geranial (mixture of citral and neral) (1.9955 g, 13.108

mmol) was weighed into a 50 ml RBF, then THF (25 ml) was

added. The temperature was lowered to 0C and 2.5M BuLi (7.9

ml, 19.75mmol) was syringed into the reaction mixture. After

an hour, the reaction was quenched with a sat. NH4Cl soln.,

and then extracted with Et20 (3 x 25 ml) The ether layer








was dried over Na2S04 and concentrated to give a colorless oil

(2.65418 g, 12.617 mmol, 96.3% yield) which was a mixture of

cis and trans double bond (spectra reported for major peaks

of mixture): 300 MHz 1H NMR (CDCl3) 8 5.22-5.04 (2H, m), 4.33

(1H, m), 2.08 (5H, m), 1.69 (6H, s), 1.61 (3H, s), 1.31 (6H,

m). 0.90 (3H, t, J=7 Hz); 75 MHz 13C NMR (CDCl3) 8 138.12,

131.53, 128.16, 123.92, 68.56, 39.53, 37.41, 27.58, 26.33,

25.60, 22.68, 17.62, 16.51, 14.02.


4,8-Dimethyl-3.7-nonadiene-2-one (182)93


Allylic alcohol 179 (2.1418 g, 12.728 mmol) was weighed

into a 100 ml RBF with PDC (9.6 g, 25 mmol), CH2C12 (25 ml),

and crushed 4A molecular sieves.95 The next day the reaction

was diluted with Et20 (75 ml) and this soln. was allowed to

stir for four hours. Suction chromatography was used to

separate the bulk of the PDC, and sieves. The eluent wss

concentrated and purified by column chromatography to give a

colorless oil (1.5105 g, 9.085 mmol, 71.4% yield) which was a

mixture of cis and trans double bond (spectra reported for

major peaks of mixture): 300 MHz 1H NMR (CDC13) 8 6.08 (1H,

s), 5.09 (1H, m), 2.16 (10H, m), 1.71 (3H, s), 1.62 (3H, s);

75 MHz 13C NMR (CDC13) 8 198.59, 158.15, 132.35, 123.54,

122.96, 41.11, 31.65, 26.08, 25.57, 19.18, 17.59.


7,11-Dimethyl-6.10-dodecadiene-5-one (183)96


Allylic alcohol 180 (2.6541 g, 12.617 mmol) was weighed

into a 100 ml RBF with PDC (9.5 g, 25 mmol), CH2C12 (25 ml),








and crushed 4A molecular sieves95. The next day the reaction

was diluted with Et20 (75 ml) and this soln. was allowed to

stir for four hours. Suction chromatography was used to

separate the bulk of the PDC, and sieves. The eluent was

concentrated and purified by column chromatography to give a

colorless oil (1.0289 g, 4.939 mmol, 39.1% yield) which is a

mixture of cis and trans double bond (spectra reported for

major peaks of mixture): 300 MHz 1H NMR (CDCl3) 8 6.05 (1H,

s), 5.08 (1H, t, J=7.2 Hz), 2.60 (2H, t, J=8.1 Hz), 2.42 (2H,

m), 2.14 (5H, m), 1.69 (3H, s), 1.62 (3H, s), 1.56 (2H, m),

1.33 (2H, m), 0.91 (3H, t, J=7,2 Hz); 75 MHz 13C NMR (CDCl3)

201.43, 157.77, 132.34, 123.88, 123.20, 44.11, 41.15, 33.74,

26.10, 25.60, 22.37, 19.23, 17.62, 13.86; IR (neat) 2960,

2930, 2873, 1688, 1619, 1450, 1377, 1131, 1076, 1040 cm-1; MS

(EI), m/e (relative intensity) 208(m+,7), 166(22), 165(21),

151(70), 123(46), 98(41), 85(31), 83(100), 82(17), 69(38),

57(18), 55(17); HR MS (EI) 208.1856 (calc. for C14H240:

208.1827).


3,7-Dimethyl-2,3-epoxy-7-octenal (185)


Unsaturated aldehyde 178 (0.9895 g, 6.500 mmol), MeOH

(12 ml), 30% H202 in water (1.4 ml, 13.7 mmol), and a sat. aq.

soln. of K2C03 (2.9 ml) were added to a 25 ml RBF. The next

day the reaction was quenched with sat. NaHCO3. This soln.

was extracted with Et20 (3 x 25 ml), then the ether layer was

dried over Na2SO4 and concentrated in vacuo. Column

chromatography of the residue produced a colorless oil (0.291








g, 1.73 mmol, 26.6% yield): Rf 0.62 (50% Et20/Hexane); 300

MHz 1H NMR (CDCl3) 8 9.43 (1H, d, J=6 Hz), 5.07 (1H, t, J=7

Hz), 3.14 (1H, d, J=6 Hz), 2.12 (2H, m), 1.9-1.5 (11H, m); 75

MHz 13C NMR (CDCl3) 8 198.65, 132.54, 122.58, 64.39, 63.41,

38.18, 33.28, 25.51, 21.98, 17.06; IR (neat) 2970, 2928,

2858, 1723, 1674, 1452, 1409, 1383, 1110, 800 cm-1; MS (CI),

m/e (relative intensity) 169(m++1,33), 151(72), 137(57),

135(25), 123(100), 111(17), 109(66), 95(16), 82(15), 81(17);

HR MS (CI) 169.1228 (calc. for C10H1702: 169.1226).


4,8-Dimethyl-3,4-epoxy-7-nonene-2-one (186)


Unsaturated ketone 182 (1.5105 g, 9.0851 mmol), MeOH (16

ml), 30% H202 in water (2.0 ml, 19.6 mmol), and a sat. aq.

soln. of K2C03 (3.3 ml) were added to a 25 ml RBF. The next

day the reaction was quenched with sat. NaHCO3. The reaction

mixture was extracted with Et20 (3 x 25 ml), then the ether

layer was dried over Na2S04 and concentrated in vacuo. Column

chromatography of the residue produced a colorless oil

(1.3225 g, 7.256 mmol, 79.9% yield): Rf 0.54 (50%

Et20/hexane); 300 MHz 1H NMR (CDC13) 8 5.11 (1H, t, J=7 Hz),

3.43 (1H, s), 2.21 (3H, s), 2.13 (2H, m), 1.85-1.50 (8H, m),

1.25 (3H, s); 75 MHz 13C NMR (CDCl3) 8 204.05, 132.36, 122.84,

64.74, 63.11, 38.06, 32.19, 27.80, 25.58, 23.60, 16.00; IR

(neat) 2970, 2922, 1723, 1452, 1404, 1382, 1356, 1241, 1187,

1079 cm-1; MS (CI), m/e (relative intensity) 183(m++1,11),

181(100), 165(24), 163(17), 143(47), 139(40), 137(10),








125(36), 123(11), 121(28); HR MS (CI) 183.1361 (calc. for

CllH1902: 183.1385).


7.11-Dimethyl-6.7-epoxy-10-dodecene-5-one (187)


Unsaturated ketone 183 (0.9697 g, 4.654 mmol), MeOH (8

ml), 30% H202 in water (1.0 ml, 9.8 mmol), and a sat. aq.

soln. of K2C03 (1.8 ml) were added to a 25 ml RBF. The next

day the reaction was quenched with sat. NaHC03. The reaction

mixture was extracted with Et20 (3 x 25 ml), then the ether

layer was dried over Na2SO4 and concentrated in vacuo. Column

chromatography of the residue produced a colorless oil

(1.0176 g, 4.535 mmol, 97.5% yield): Rf 0.70 (50%

Et20/hexane); 300 MHz 1H NMR (CDC13) 8 5.11 (1H, t, J=7 Hz),

3.43 (1H, s), 2.50 (2H, m), 2.02 (2H, m), 1.87-1.47 (10H, m),

1.35 (2H, m), 1.22 (3H, s), 0.91 (3H, t, J+7 Hz); 75 MHz 13C

NMR (CDC13) 8 206.38, 132.46, 122.93, 64.53, 63.25, 40.54,

38.16, 32.17, 25.65 ,25.28, 23.67, 22.30, 16.05, 13.74; IR

(neat) 2961, 2931, 2873, 1722, 1454, 1407, 1382, 1249, 1133,

1068 cm-1; MS (CI), m/e (relative intensity) 224 (m+, 1.8),

223(20), 205(2), 157(2), 143(6), 141(5), 139(8), 137(3),

125(8), 121(7); HR MS (CI) 223.1657 (calc. for C14H2302:

223.1698).


General Procedure for Cyclizations of Epoxy Ketones


The appropriate C,P-epoxy ketone (1.0 mmol) was weighed

into a 10 ml pear-shaped flask (PSF). Then TBTH (2.0 mmol),

AIBN (0.1 mmol), and benzene (2.0 ml) were added to the same








flask. This mixture was degassed for 20 min. After it was

degassed, the temperature was raised to 850C. After the

starting material had been consumed, the reaction was

concentrated and purified by column chromatography to yield

the cyclized products.


(2-Hydroxy-5-isopropyl-2-methylcyclopentyl)ethanone


This compound was isolated as a mixture of two products

in a 3:2 ratio (1H NMR integration). The isomers were not

fully separated by column chromatography. The combined yield

for both compounds was 82.7%. Despite the chromatography

problems some fractions did contain pure compound and the

spectra for each are reported below.

Major Product (189): Rf 0.31 (50% Et20/hexane); 300 MHz

1H NMR (CDC13) 8 2.88 (1H, s), 2.65 (1H, d, J=10 Hz), 2.38

(1H, m), 2.27 (3H, s), 1.94 (1H, m), 1.83-1.48 (4H, m), 1.34

(3H, s), 0.84 (6H, dd, J=7, 12 Hz); 75 MHz 13C NMR (CDC13) 8

213.67, 81.22, 64.61, 49.60, 41.45, 33.02, 32.07, 27.78,

26.08, 21.51, 19.38; IR (neat) 3446, 2959, 2872, 1699, 1466,

1422, 1369, 1245, 1170, 1129 cm-1; MS (CI), m/e (relative

intensity) 185(m++l,100), 184(1), 168(5), 167(50), 149(3),

141(2), 123(9), 109(1); HR MS (CI) 185.1526 (calc. for

C11H2102: 185.1541).

Minor Product (190): Rf 0.25 (50% Et20/hexane); 300 MHz

1H NMR (CDC13) 8 2.80 (1H, d, J=9 Hz), 2.26 (3H, s), 2.24 (1H,

m), 1.75 (4H, m), 1.50 (2H, m), 1.18 (3H, s), 0.84 (6H, dd,

J=7.5, 17 Hz); 75 MHz 13C NMR (CDC13) 6 210.72, 81.59, 66.76,








47.00, 42.49, 32.97, 32.26, 26.03, 24.61, 21.29, 20.10; IR

(neat) 3462, 2960, 2872, 1702, 1466, 1368, 1242, 1167, 1127,

932 cm-1; MS (CI), m/e (relative intensity) 185(m++1,61),

169(2), 168(11), 167(100), 165(1), 149(8), 141(2), 124(1),

123(16), 109(3); HR MS (CI) 185.1545 (calc. for C11H2102:

185.1541).


1- (2-Hydroxy-5-isopropyl-2-methylcyclopentyl) --pentanone


This compound was isolated as a mixture of two products

in a 1:1 ratio (by isolated weights). The combined yield for

both compounds was 81.5%. The spectra for each are reported

below.
Major Product (111): Rf 0.48 (50% Et20/hexane); 300 MHz

1H NMR (CDC13) 8 3.10 (1H, s), 2.63 (1H, d, J=ll Hz), 2.54

(2H, t, J=7 Hz), 2.49 (1H, m), 2.00-1.50 (7H, m), 1.35 (2H,

m), 1.31 (3H, s), 0.92 (3H, t, J=7 Hz), 0.85 (6H, dd, J=7, 13

Hz); 75 MHz 13C NMR (CDC13) 6 215.94, 81.26, 63.83, 49.81,

45.86, 41.46, 31.74, 27.78, 25.80, 25.23, 22.28, 21.60,

19.14, 13.86; IR (neat) 3481, 2959, 2872, 1692, 1466, 1370,

1129, 1073, 942 cm-1; MS (CI), m/e (relative intensity)

227(m++1,100), 219(6), 210(9), 209(65), 191(8), 183(4),

149(7), 127(11), 124(4), 123(55); HR MS (CI) 227.2015 (calc.

for C14H2702: 227.2011).

Minor Product (1.92): Rf 0.40 (50% Et20/hexane); 300 MHz

1H NMR (CDC13) 6 2.78 (1H, d, J=9 Hz), 2.60-2.41 (2H, m), 2.23

(1H, m), 1.92-1.42 (8H, m), 1.32 (2H, m), 1.15 (3H, m), 0.93

(3H, t, J=8 Hz), 0.83 (6H, dd, J=7, 18 Hz); 75 MHz 13C NMR








(CDC13) 8 212.62, 81.75, 66.07, 47.18, 44.83, 42.50, 32.82,

26.01, 25.41, 24.80, 22.30, 21.33, 20.02, 13.89; IR (neat)

3474, 2958, 2872, 1694, 1466, 1377, 1307, 1244, 1109, 1068

cm-1; MS (CI), m/e (relative intensity) 227(m++1,11), 226(2),

214(1), 211(1), 210(6), 209(53), 208(1); HR MS (CI) 227.1999

(calc. for C14H2702: 227.2011).


2-Hydroxy-5-isopropyl-2-methylcyclopentanecarboxaldehyde


This compound was isolated as a mixture of two products

in a 3:1 ratio (1H NMR integration and Capillary GC). The

isomers were not fully separated by column chromatography,

and the combined yield for both compounds was 81%. Despite

the chromatography problems some fractions did contain pure

compound and the spectra for each are reported below.

Major Product (198): Rf 0.31 (50% Et20/hexane); 300 MHz

1H NMR (CDCl3) 8 9.80 (1H, s), 2.46 (1H, m), 2.30 (1H, d, J=9

Hz), 2.13 (2H, m), 1.8-1.4 (4H, m), 1.42 (3H, s), 0.88 (6H,

dd, J=7, 13 Hz); 75 MHz 13C NMR (CDC13) 8 206.01, 82.78,

64.42, 46.78, 41.67, 32.26, 27.85, 26.83, 21.18, 19.99; IR

(neat) 3437, 2960, 2872, 2730, 1715, 1466, 1378, 1268, 1136,

1074 cm-1; MS (CI), m/e (relative intensity) 171(m++1,7),

169(4), 155(8), 154(5), 153(69), 137(12), 135(23), 127(12),

114(4), 100(11); HR MS (CI) 173.1354 (calc. for C10H1902:

171.1385).

Minor Product (199): Rf 0.25 (50% Et20/hexane); 300 MHz

1H NMR (CDC13) 8 9.70 (1H, d, J=3 Hz), 2.54 (1H, dd, J=3, 7

Hz), 2.18 (1H, m), 1.9-1.5 (5H, m), 1.32 (3H, s), 1.25 (1H,









s), 0.88 (6H, dd, J=7, 12 Hz); 75 MHz 13C NMR (CDC13) 8

203.93, 66.62, 45.97, 42.10, 32.92, 29.70, 27.08, 25.04,

21.31, 20.33; IR (neat) 3430, 2957, 2872, 2729, 1716, 1463,

1376, 1075, 926 cm-1; MS (CI), m/e (relative intensity)

171(m++1,40), 155(17), 153(100), 135(22); HR MS (CI) 171.1351

(calc. for C10H1902: 171.1385).


7-Octene-1.6-diol (212)


E-Caprolactone (5.21 g, 45.6 mmol) and CH2Cl2 (92 ml)

were added to a 250 ml RBF equipped with a magnetic stir bar.

This solution was chilled to -780C with a dry ice/acetone

slush bath. DIBAL (50 ml of a 1M soln., 50.0 mmol) was

cautiously added over a 20 min. into the flask by placing the

syringe needle tip onto the upper neck of flask so that DIBAL

did not drip directly, but rather ran down the flask and was

cooled before it gets to solution. After 1 hr. the reaction

was quenched with MeOH (3 ml) at -780C. The reaction mixture

was then poured into a freshly prepared aq. sat. Rochelle's

salt soln. (300 ml) with rapid stirring, and this was allowed

to stir overnight. The reaction mixture was extracted with

EtOAc (4 x 150 ml), then the organic layer was dried over

Na2SO4 and concentrated to yield crude 6-hydroxyhexanal. The

crude product polymerized readily so it was used without

further purification.

The crude 6-hydroxyhexanal was placed in a 500 ml RBF

along with freshly distilled THF (150 ml) and a magnetic

stirrer. This mixture was chilled to 0C with an ice bath.