_'N
I“):

meets-t

 

This is to certify that the
thesis entitled
A GENERAL APPROACH TO THE SYNTHESIS

OF NATURAL PRODUCTS CONTAINING
FIVE-MEMBERED HETEROCYCLES

presented by

Paul Matthew Herrinton

has been accepted towards fulfillment
of the requirements for

_M,_S_._degree in _Che.mi_s_tLy

wt {é

Major professor

Date 6/952?“

0-7639

3;; '1: g.- ,. a

 

 

MSU

LIBRARIES

W

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

A GENERAL APPROACH TO THE SYNTHESIS
OF NATURAL PRODUCTS CONTAINING
FIVE-MEMBERED HETEROCYCLES

BY

Paul Matthew Herrinton

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemistry

1982

(“13.00am

ABSTRACT

A GENERAL APPROACH TO THE SYNTHESIS
OF NATURAL PRODUCTS CONTAINING
FIVE-MEMBERED HETEROCYCLES

BY

Paul Matthew Herrinton

Most of the syntheses of natural products containing
five—membered heterocycles have been approached by careful
construction of a parent carbocycle to which the hetercyclic
ring is appended. This study demonstrates a general method
which acknowledges the heterocycle as an integral part of
the molecule.

Several 3—furyl epoxides were prepared in high yield by
either: 1) addition of the Grignard reagent derived from
(3-furyl)-chloromethane to halo-olefins and then oxidation with
meta-chloroperbenzoic acid, or 2) addition of the organo-
lithium reagent derived indirectly from (3-furyl)-chloro-
methane to iodoepoxides.

Cyclization of these 3-furyl epoxides was attempted
by treatment with Lewis acid. The best Lewis acids for
cyclization were found to be zinc iodide or triisopropoxy-
titanium chloride. Cyclizations which formed six- and seven-
membered rings proceeded smoothly but attempts to form

five-membered rings have been unsuccessful.

To my lovely Joan

ii

ACKNOWLEDGMENTS

The author wishes to thank Dr. Steven P. Tanis for
his patience, support and guidance throughout this project.

Financial support from Michigan State University in
the form of an assistantship from September, 1980 to June,
1982 is gratefully acknowledged.

The author also wishes to acknowledge the members of
the faculty and staff for their assistance and advice through-
out this project. The author wishes to thank his fellow
students for their advice and companionship.

I wish to thank my parents and family for their love
and support without which this work would not have been
possible.

Special thanks to my wife Joan for her invaluable
assistance in the preparation of this thesis.

iii

TABLE OF

List of Tables. . . . . .
List of Figures . . . . .
Introduction. . . . . . .

Results and Discussion. .

CONTENTS

Synthesis of epoxy—furans $3535.

Cyclizations . . . .
Conclusions . . . . . . .
Experimental. . . . . . .

Bibliography. . . . . . .

iv

vi

15
26
29
60

Cyclization Results.

LIST OF TABLES

l8

Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure
Figure

Figure

LIST OF

vi

FIGURES

ll
13
16
21
22
24

24

INTRODUCTION

The chemical literature contains numerous reports of
various biological activities which are associated with new
and interesting skeletal types. The variety of naturally
occurring biologically active compounds is as diverse as
the systems which produce them. However, upon closer
examination, some recurring themes do appear. Many biologically
active terpenoids contain five-membered heterocyclic rings.1
The rings may vary in oxidation state from furan l, or tetra-

hydrofuran 2 to butenolide 3 or butyrolactone g.

Q Q €101.10

Figure 1

Examples of such compounds include the clerodane
diterpenes Ajugarin I i2 and annonene g3, the drimane
sesquiterpene confertolin 14, the cytotoxic vernolepin gs's
the pseudoguaianolide confertin 26, and witchweed germination
promoter strigol $27.

As a result of the biological activities exhibited by
compounds gelg, which include insect antifeedant, anti-

complimental, cytotoxic and germination promotion activities,

as well as antimicrobial, antifungal and antitumor activities

these molecules have become attractive targets for total

chemical synthesis.

 

Figure 2

For the most part the syntheses of these terpenoids
have been approached by careful, stereocontrolled construc-
tion of a parent carbocycle, upon which a furan, butyrolac-
tone, of butenolide is appended. These schemes have not,
generally, acknowledged the basic five-membered ring hetero-
cyclic system as an integral part of the molecule which can
exert a measure of control upon bond forming reactions.

A methodology which would be generally applicable to the
synthesis of a terpenoid containing a five-membered hetero-

cyclic ring would therefore be very valuable.

Examination of compounds §r_g reveals three general

substitution patterns about the five-membered heterocyclic

rings (figure 3).
{\R'

R1? Cl? 0

A B C
Figure 3

Structure A represents a simple 3-substituted furan
equivalent such as compound g or g. Although many methods
for the synthesis of 3—substituted furans have been reported
most of them require a number of steps and/or proceed in
low yields.8 Recently Tanis reported a convenient high
yield synthesis of simple 3—substituted furans.9

Structure C represents a 3,4-disubstituted furan

equivalent such as compound 1. Oishi 25 al10

have demon-
strated that type C skeletons may be made by cyclization of

a judiciously functionalized furan (figure 4).

 

Figure A
Structure B represents a 2,3-disubstituted furan
equivalent. These type-B furans are found as integral parts
of numerous natural products. The activities associated
with molecules of this structural type range from the

germination stimulation activity for the parasite witchweed

exhibited by strigol lg7, to the cytotoxicity shown by the
pseudoguaianolide confertin 26, and the fish anti-feedant

activities possesed by the sponge, Dysidea fragilis, derived

ll

 

furanosesquiterpenes nakafuran-B 1; and nakafuran-9 li-
It was our goal in this study to devise a general methodology

for the synthesis of type B ring containing compounds.

QH

      

H
o
lw

 

Figure 5

RESULTS AND DISCUSS ION

If we consider the usual propensity exhibited by furans
for undergoing electrophillic substitution at C-2, the prepara-
tion of polycyclic materials related to structure g (figure 3)
can be simplified as in figure 6. The generation of an
electron deficient center (RflQ) in the 3-substituted furan
1; should provide compound lg after cyclization and
rearomatization. A simple synthesis of type B furans should
then be possible by synthesising a suitably functionalized
type A furan. If a latent electrophile is placed in the side
chain of a 3-substituted furan and then unmasked it should

lead to cyclization at the preferred 2 position.

R F j

_. (R .. .
R'/()\ R3 /0\ (,/0\

lg 16
Figure 6

 

 

Catonic cyclizations of this type have been the object
of intense study since 1950.12 However, there are few
examples in which the cyclization terminator is other than
a simple olefin or phenyl group. It has been demonstrated

by Boeckman 13

that epoxy vinyl ethers 11 could be cyclized
by treatment with Lewis acids (figure 7). We wished to
demonstrate that this type of cyclization can be routinely

performed with furan as the terminator and that five-, six-,

and seven—membered rings may be formed.

(1>-' (35”
(CH 2’
og2’n 033

Figure 7

Although furans are known to undergo electrophillic
aromatic substitution reactions more easily than benzene, the
availability and stability of suitable substrates has limited
the exploration of the reactivity and potential synthetic
utility of this functional group. We also wished to examine
the effect of the placement of the initiating function inside
the ring being formed (endocyclic) or outside the ring being
formed (exocyclic). According to the study of Baldwin gt
31,14 all of the exocyclic closures which generate five-,
six-, or seven-membered rings should be favorable while of
the endocyclic closures only the formation of a six-membered
ring is considered to be favorable.

When using epoxides as cyclization initiators, there is
the possibility that either epoxide carbon may serve as the
electrophilic center. In order to avoid this type of
regiochemical ambiguity in our studies, it was necessary to
bias the epoxide functions so that one mode of polarization
would be favored over the other. This is possible because
of the proposed cationic nature of the reaction intermediate.12f
When epoxides are treated with Lewis acids one of the carbon

oxygen bonds becomes polarized as in figure 8. This results

in positive charge being placed on the carbon. If one carbon

of the epoxide can stabilize that positive charge better

than the other, then the polarization will be mostly of that
carbon oxygen bond. Therefore, by making one side of the
epoxide resemble a more stable tertiary carbocation and the
other side resemble a less stable primary or secondary cation,

we can favor one mode of polarization over the other.

 

Figure 8

With this in mind, we chose the six epoxy furans in
figure 9 as substrates for cyclization.

The cyclization of compound 22 corresponds to a five-
membered endocyclic closure to give 22. 2g is representa-
tive of a five-membered exocyclic cloSure to provide gg,
cyclization of 22 should proceed 223 a six-membered endocyclic
closure to give 22, and cyclization of 22 should be considered
as a six-membered exocyclic closure to give 22. The cycliza-
tion of 22 represents a seven-membered endocyclic closure
to give 22, and cyclization of 21 corresponds to a seven—

membered exocyclic closure to give 39.

Synthesis of epoxyefurans 19-24

 

The most obvious and simplest path to the desired epoxy-
furans was assumed to be epoxidation of the corresponding
(3-furyl)-olefins. The necessary olefins may be prepared

by coupling the appropriate olefin with a judiciously

Designation

S-endo

5-exo

6—endo

6-exo

7-endo

7—exo

Epoxide Structure

 

/\
O 0
.2}.
(Q
g;
/\
0 o
_2__§_
/\
O
_2_51_
Figure 9

8

Desired Cyclization

Product
0H
/ \.
22
I, \‘ll" 0H
0
2g
I, \‘III} 0H
0
21
, x.
0 0H
22
, x.
o 0”
2g
, x.
0 0H
30

functionalized isoprenoid furyl synthon. Relative to the
standard bond forming reactions of such furans, having the
furan serve as electrophile and the alkyl group as nucleo-
phile is the "normal" bond forming polarity (path a, figure
10). We have examined the "reverse polarity" bond formation,
in which the furyl moiety serves as nucleophile and the alkyl

group as electrophile (path b, figure 10).
CH2®
F? €3ng
/o\ +R® LWFL/\ +96
0 0

Figure 10

This approach would involve the reaction of furyl
organolithum 22 or Grignard reagent 22 with an appropriate
electrophile. To the best of our knowledge 22 has been

9 has demon-

reported only once in the literature. Tanis
strated the utility of 22 in the synthesis of simple 3-
substituted furans. The requisite Grignard precursor 22
may be easily prepared from the readily available (3-furyl)—
methanol15 in 80-85% yield by the chlorination procedure
of Collington and Meyers.16
Treatment of (3-furyl)- chloromethane 22, in tetrahydro-
furan (THF), with magnesium, provides the Grignard reagent

22 quantitatively as determined by titration.17

Li MgC/ Cl

A variety of primary, secondary and allylic halides
have been reacted with 22 in the presence of Kochi's catalyst
LiZCuCl418 to give uniformly high yields of 3-substituted
furans.

The general route to the epoxy-furans would then be
the coupling of Grignard reagent 22 with an appropriate
haloalkene and then epoxidation of the product furyl olefin.
The coupling reactions all proceed smoothly, and in good
yield, as indicated in figure 11. The only exception is the
coupling of the vinyl bromide 22. There is no reaction
between 22 and 22 in the presence of LiZCuCl4, however,
Kochi notes that vinyl halides react only in the presence
of FeCl3, and indeed with ferric chloride as catalyst the
coupling proceeds in 82% yield.

Treatment of furyl olefins 22, g; and 3; under the
standard conditions with m-chloroperbenzoic acid results in
smooth conversion to the desired epoxy-furans in good to
excellent yields, as the only isolated products. Epoxida-
tion of furans 22, 22 and 22 proved more troublesome pro-
viding little or none of the desired product epoxides. It

became obvious that if the alkene is less than trisubstituted

10

>=/B" NR (LI'ZCuC/4)

If” 3"

 

 

 

 

39 (82%) H. (88%)
Mg 35 mm
L 009, H0 Ho
’2 U 4 (820
— 4” .2— (25%)
ON” “75/ /\
Li2CUC/4 Ho 0
1% 18/96) 2_1_ (850 /o)
\
cf“? 37236? m
\ / \
LiClJC/ o o
2 4 4 (8 o o
— 3/0) E (0 4;)
Mac/mil
/ \ . — /metza_. / \
0 LIZCUC/4 /\o
4__3 {730/0} _2_; (8/°/o)

Figure 11

ll

oxidation of the furan ring becomes a competitive reaction.
When furyl olefin 22 is reacted with m-chloroperbenzoic
acid only 25% of the desired epoxy-furan 22 is recovered.
The remainder of material isolated has been tentatively
identified as the furan oxidation product 22. Oxidation

of olefin 22 provided none of the desired product. Instead,
exclusive furan oxidation was observed. Other methods of
epoxidation which were tried included: treatment with
anhydrous t-butyl hydroperoxide in the presence of various
transition metal catalysts (Mo(CO)6, VO(acac)2, Ti(OiPr)4)19
and oxidation under basic conditions with benzonitrile,

hydrogen peroxide and sodium hydroxide.20

The only set of
reaction conditions which provided any product was t-butyl
hydrogen peroxide with Mo(CO)6 and that resulted in only a

25% yield of 22 and none of 22.

It was then necessary to explore alternate routes to
the epoxy furans 22, 22 and 22. Tanis has noted that Grignard
reagent 22 will react with an allylic halide containing a
remote epoxide function to give the epoxy furan 22 in 79%
yield. Products corresponding to attack at the epoxide
centers were not observed. It should, therefore, be possible

to couple epoxy bromides, iodides or tosylates to yield the

12

desired epoxy furans. We have discovered, however, that
reaction of 22 with 47a, 47b or 47c in the presence of
LiZCuCl4 resulted only in products which were the result

of attack at the epoxide function.

M96] 0
/ \ 0/ \ 0 \ I

g 22
§\/\ 0
X X
2.7. 4_ii
a) X=Br a) X=Br
b) X=I b) X=I
c) X=OTs c) X=OTs
Figure 12

Encouraged by reports21 that alkyl cuprates could be
reacted with epoxy-tosylates to give mixtures of displacement
and epoxide opening products, we explored the utility of
using the alkyl cuprate 22 in such a coupling reaction.

The synthesis of furyl cuprate 22 required the preparation
of the furyl organolithium 22. The reactive nature of the
halide 22 does not allow a direct metallation and so it

was deemed necessary to convert the halide to something less
reactive which can still afford 22. The tri(n—butyl)

tin compound 22 appeared to be an ideal furyl lithium 22

13

equivalent. The preparation of 22_was realized upon treat-
ment of chloride 22 with tri(n-butyl) tin lithium according

to the procedure of Still.22

This technique provided the
tri(n-butyl)-(3-furyl methyl) stannane 22 in 76% distilled
yield. Tin-lithium exchange was accomplished by reaction
with n-butyl lithium and the cuprate prepared by the pro—
cedure of Marino.23 Unfortunately, the addition of 22 to

bromo, iodo, or tosyl epoxides 47a-c and 48a-c provided

only epoxide opening products.

[far/v [fsm «any/)3
o o

49 50

Direct utilization of the organolithium reagent 22
was initially avoided because of the possibility of an allylic
type rearrangement. This sort of behavior has been previously

observed in the reaction of this Grignard reagent.24

However,
when the lithium reagent derived from 22, by tin-lithium
exchange, was reacted with iodoepoxides 212_and 222 in the
presence of hexamethylphosphoramide at -25°C the desired
epoxy furans 22_and 22 were obtained in 80 and 68% yields
respectively. None of the product which would result from
the rearranged anion were detected.

That left only epoxy furan 22 to be prepared in accept—

able yield. Reaction of the furyl organolithium reagent

l4

with the appropriate iodoepoxide should provide the deisred
compound in good yield. Unfortunately, the requisite iodo-
epoxide 22 is unknown and thus far has resisted all attempts
to affect its preparation. The synthesis of 22 remains under

study.

gyclizations

 

The chemical literature contains many examples of

epoxide initiated biomimetic type cyclizations.
majority of these examples have used simple olefins as the
terminator function. To the best of our knowledge there have
been few reports of cyclizations utilizing furan as the
terminator. We had two major concerns when considering the
proposed epoxy-furan cyclizations. The first was the
nucleophilicity of the furan terminator. Furans are known

to be more reactive toward electrophilic substitution than
most aromatics but less nucleophilic than simple olefins.

The crucial question is whether the electrophile generated

by the Lewis acid—epoxide complex will react with the furan
in a cyclization faster than it will yield other undesired
products. An important factor in the partitioning of the
reaction between the fruitful cyclization pathway and the
formation of other products may be the degree of epoxide

bond breaking by the Lewis acid. When the epoxide is com-
pletely opened the resultant carbocation may eliminate to
give an alkene or react in some other undesirable manner
before it can react with the weakly nucleophillic furan.

On the other hand, when the epoxide is weakly polarized,

15

the probability of elimination is lowered and cyclization
may become the major pathway. Therefore, Lewis acid strength
may play an important role in determining product distribution.
The second concern was stability of the cyclization
products. The product of the cyclization sequence is a
2,3-disubstituted furan which is decidedly more labile than
the starting material. Therefore, the cyclization conditions
must be carefully chosen so that the presence of strong

protic acid is avoided.

strong F, .
Lewisacid CW /\ 0H

m \’ . I
O
/’///’
wea
Lewis acid

 

Figure 13

The majority of epoxide initiated cyclizations reported
in the literature are carried out by brief exposure to boron
trifloride etherate in a non-polar solvent such as methylene
chloride. Under this set of conditions the reaction is
thought to occur in a concerted manner, that is, little
cationic character is ever generated on the electrophilic

12f Under these conditions olefin terminated cycliza-

carbon.
tions proceed smoothly to provide a 20-60% yield of desired
product. However, furans are much less nucleophilic than

alkenes and these reaction conditions may prove to be too

acidic to affect the desired cyclizations.

l6

The first set of conditions which were used were those
which appear to be "standard" for polyene cyclizations,
that is about l/3 of an equivalent of boron trifloride
etherate in methylene chloride at -25°C. The results of
the cyclization attempts under these conditions are summarized
in table 1. Only the six-membered endocyclic case 22
provided any of the desired cyclized product. The majority
of materials recovered from the attempted cyclizations of
22 (S-endo), 22 (S-exo), 22 (6-exo), 22 (7-endo), and 22
(7—exo) were mixtures of allylic alcohols. The total amount
of material recovered from the reaction was only about 50%
of the starting material. Evidently, in all cases except
the 6-endo, elimination to form allylic alcohols was faster
then cyclization. The lack of cyclization in every case
except 6-endo, coupled with the poor mass balance clearly
demonstrates that the standard cyclization conditions are
not generally applicable. Both the formation of the allylic
alcohols and cyclization generate protic acid which may
destroy labile furanoid products. This could account for
the poor mass balance of the reaction.

If the strength of the Lewis acid employed was responsible
for the destruction of the reaction products as well as the
general lack of cyclization then what was needed was a
weaker Lewis acid which would also be able to capture the
protic acid released by the reaction. Snider25 has reported
that alkyl aluminum halides act as Lewis acids which react

with Bronsted acids liberating alkanes and regenerate a

17

Mom

m mm e co mmucu mace How czonm mum mafismou QOmmmH umcu

u a Wmnmwmmm soHLB pmwamumo m U250m can m3 Hagan coaumuflaowo mcflumamuum
Eoum ms emumw>mum onmlmemm CMHDMwaomm mafiummwum CH wuHDOAMMHo wea

one
wow

wmh

Homrumwoewe

New
wqm
wmm
was
was

we

we

HEN

wm

N

we

v wmm

“mm: memaw

we

wea

wea

wNN

HUH<

muHSmwm COHDMNHHO>U .H wanme

.oxmum new;
.Ham3 xuos ou

whv fllcflocmmmaamm
hxouvhnlum
we we oxmln
we wea oecmlh
w e wem oxmle
wea whv oecmue
I we oxmum
we we oecmlm

macawum mum0.mmm

18

Lewis acid. The alkyl aluminum halides cover a wide range

of Lewis acitity. Replacing chlorines with alkyl groups

has been observed to decrease the Lewis acidity in a pre-
dictable fashion. Ethylaluminum dichloride is only slightly
less acidic than aluminum trichloride,22 while diethyl-
aluminum chloride is substantially less acidic and trimethyl-
aluminum is a very mild Lewis acid. The range of Lewis
acidity presented by alkylaluminum halides and their ability
to absorb protic acids seemed to make them ideal for
initiating epoxy-furan cyclizations.

Treatment of the various epoxy-furans with two equiva-
lents of ethylaluminum dichloride at -25°C in methylene
chloride provided very disappointing results (table 1).

As with boron trifloride etherate, only the six-endo case
22 provided any cyclized material. The majority of the
products in all cases were allyclic alcohols resulting from
elimination. However, it was noted that the mass balance
of the reaction had improved markedly to about 70%.

We appeared to be on the right course using a Lewis
acid which is also a base. However, the Lewis acidity was
too great and elimination was still faster than nucleophilic
attack by furan. Diethylaluminum chloride is much less
acidic than ethylaluminum dichloride and so this was the
next logical choice as a Lewis acid. Exposure of the epoxy-
furans to two equivalents of diethylaluminum chloride at
-25°C in methylene chloride led to some cyclization in the

six-endo (22), six-exo (22), and seven-endo (22) examples.

19

The yields of the desired products are very low (10-22%)
with the major products being allylic alcohols. However,
the material balance of the reaction is very good with 80%
of the starting mass being recoverd.

The results from these cyclization attempts seemed to
indicate that the Lewis acids being used were still much
too strong and we began to search for examples of weaker
Lewis acids as catalysts for cyclization. Boeckman13 has
reported success in cyclizing epoxy vinyl ethers by exposure
to basic alumina (A1203, activity I) at room temperature in
hexane. Although vinyl ethers are recognized among the most

nucleophilic of olefinic bonds26

and furans are among

the least, there are enough similarities to indicate that
basic alumina might be a suitable reagent for these cycliza-
tions. Stirring epoxy-furans 22 and 22:22_with alumina in
hexane at room temperature for 24 hours resulted in very
high yields (BO-90%) of elimination products. Only 22
showed any cyclization product and that was obtained in
only 23% yield.

Recovering high yields of allylic alcohols and little
or no cyclized material was an indication that a milder
Lewis acid was necessary. Basic alumina is a mild Lewis
acid and that reagent is too potent. Magnesium bromide is
a weak Lewis acid which has a high affinity for oxygen and
a low affinity for nitrogen. This high affinity for oxygen
allows magnesium bromide to be used as a Lewis acid in the

presence of tertiary amines such as triethylamine. This

system (magnesium bromide, triethylamine) seemed appropriate

20

for the epoxy-furan cyclization. Exposure of epoxy-furan 22
(6-endo) to three equivalents of magnesium bromide-tetra-
hydrofuran complex and one equivalent of triethylamine in
methylene chloride at room temperature for 24 hours provided
43% of the cyclized product 22 and 53% unreacted starting
material. Magnesium bromide, under these conditions appears
to be too mild a Lewis acid. Although we have not demonstrated
the need for the added triethylamine, since the derived
magnesium alkoxide is likely to be an efficient acid scavenger,
this system has shown some promise.

Titanium alkoxides have been shown to be effective
Lewis acids for the catalysis of aldol condensation.27
The Lewis acidity of titanium alkoxides may be varied by
replacing alkoxides with halogens. Titanium tetrachloride
is a very strong Lewis acid which has been observed to react
with epoxides to yield/B‘Chlorotitanates (figure 14).28
Replacing halides by alkoxide groups, such as isopropoxy,
lessens the Lewis acidity. Monoalkoxytitanium trichlorides
are only slightly less acidic than titanium tetrachloride.
Dialkoxytitanium dichlorides and trialkoxytitanium chlorides

are intermediate to titanium tetrachloride and the very mild

tetraalkoxytitanium compounds (figure 15).

Ti (:14

 

I?

C!
/O
RCHJCHZ THOCHZA‘HRQ

Figure 14

21

TiCl4> Ti (OR) Cl3> Ti (OR) zClz > Ti (OR) 3C1 > Ti (OR) 4

 

Decreasing Lewis Acidity -—————-—’

Figure 15

Exposure of epoxy-furan 22 (6-endo) to three equivalents
of the very mild Lewis acid titanium tetraisopropoxide in
methylene chloride at room temperature for 24 hours resulted
in no reaction. The next most acidic compound in the series,
triisopropoxytitanium chloride, was prepared by mixing three
equivalents of titanium tetraisopropoxide and one equivalent
of titanium tetrachloride in methylene chloride.28 The
triisopropoxy titanium chloride was never isolated but simply
used as a stock solution in methylene chloride which was
stored at -20°C. Stirring epoxy-furan 22 with three equiva-
lents of triisopropoxy titanium chloride in methylene chloride
at room temperature for three hours provided 71% of the
desired cyclization product. Similar reactions with the
Iremainder of the epoxy-furans provided 64% of the six-exo
product 22, 88% of the seven-endo product 22 and 24% of the
seven-exo product 22. Both five-membered precursors 22
and 22 failed to provide even trace amounts of the desired
products yielding only mixtures of allylic alcohols.

The final Lewis acid employed in this preliminary study
was zinc iodide. It has been shown that zinc iodide is

useful as a Lewis acid in the addition of allylic acetates

22

to silyl enol ethers.29

Stirring epoxides 22722 with three
equivalents of freshly prepared zinc iodide etherate and
one equivalent of sodium acetate in methylene chloride for
18 hours at room temperature provided good yields of the
cyclized materials 22, 22, 22 and 24% of 22. The two five-
membered ring cases 22_and 22 yielded none of the desired
materials, but instead a 75-80% of the material was recovered
as mixtures of allylic alcohols. Again in this series of
reactions the utility of sodium acetate has yet to be demon-
strated.

Tanis30 has noted that treatment of epoxy dendrolasin
22 with boron trifloride etherate provides 47% of 3-fl3-
hydroxy pallesencin A 22. According to our study, this
reaction should proceed in higher yield with either zinc
iodide-sodium acetate or triisopropoxy titanium chloride as
the Lewis acid. Exposure of 2,6-dimethyl—9-(3-furyl)-2,6
nonadiene-oxide-Z 22 to 3 quivalents of triisoprOpoxy titanium
chloride for one hour at room temperature provided 57% of
B-p -hydroxy pallesencin A and treatment of 22 with three
equivalents of zinc iodide and one equivalent of sodium
acetate at room temperature for three hours provided 54% of 22,
The yields are actually higher than indicated as the samples
of 22 were contaminated with at least 25% epoxy geranyl
chloride as determined by lHNMR. The preparation of
larger samples of 22 should allow more efficient purifi-

cation providing material of much higher purity for future

study.

23

 

 

Figure 16

During the course of this investigation it was necessary
to have large amounts of epoxy-furan 22 on hand. The prepa-
ration of 5—iodo-Z—methyl-Z-pentene 22 requires a long and
tedious synthesis while 4-chloro-2-methyl-2-butene 22 is
readily available by reduction of 3,3-dimethyl acrylic acid
and chlorination of the resulting alcohol. Coupling of the
readily available 22 with a one carbon homologue of 22 or
_2 would provide the furyl olefin 22. This synthon could be

used in the preparation of nakafuran-9 22 (figure 17).

O

 

Figure 17

The necessary couplings could be accomplished using
either the organo-lithium 22 or Grignard reagent 22 which
both may be derived from the corresponding tri-(n-butyl)

tin compound 22. The desired one carbon chain extension

24

has been achieved in 92% yield by the addition of tri—(n-butyl)-
iodomethyl stannane22 to Grignard reagent 22 in the presence

of LiZCuCl4. The Grignard reagent 22 was then prepared by
tin-lithium exchange on 22, with n-butyl lithium, and addition
of the organo-lithium to a solution of magnesium bromide

in THF. Reaction of the resultant Grignard reagent with

4-chloro-2-methyl-2-butene in the presence of LiZCuCl4 pro-

vided the desired furyl olefin 22 in 82% yield.

Wu WMQC/ / \ 3,,"_Bu,3

51 52

Id: CD

25

CONCLUS IONS

The general method devised for synthesis of type B
furans (figure 3) appears to be very promising. The synthesis
of the required 3-furyl epoxides proceeds smoothly and in
high yield. The only major problem encountered was in the
oxidation of the 3-fury1 olefins. These reactions do not
provide the desired epoxide, if the alkene is not at least
trisubstituted. This problem is overcome simply by coupling
the organo-lithium reagent 22 with the proper iodo-epoxide.
Only 2-methyl-4-(3-furyl)-l-epoxy-butene 22 cannot yet
be prepared in at least 65% overall yield. The successful
synthesis of 22 requires a convenient, high yield route to
the requisite iodoepoxide.

The results of our initial cyclization attempts are
surprising in several respects. The first and most dis-
appointing of these results is the failure of both the five-
endo and five-exo precurors to undergo cyclizations. Exami-
nation of molecular models clearly demonstrates that the type
of orbital overlap necessary to effect cyclization is nearly
impossible in the five-endo case and therefore, we did not
expect cyclization to occur in this case. However, in

3 carbons in the ring being

the five-exo closure the two sp
formed allow for enough flexibility to make overlap possible
and we anticipated a successful closure. There are examples
of cyclizations with similar geometry in the literature.

Snider31a has reported a case (figure 16) in which the ring

26

being formed contains three sp2

sp3 carbons as is the case in the cyclization of 22. This

carbons and two adjacent

example and examination of molecular models seem to indicate
that this cyclization should work but a milder Lewis acid
will be required because the poor orbital overlap slows

the nucleophilic attach of furan making elimination

d,3lb

possible. Sutherlan and Yamamoto31C have reported

cyclizations with identical steric constraints.

EfAlC/g .
s

We had anticipated difficulty in affecting the cycliza-
tion of seven-exo precursor 22, As mentioned earlier, the
electrophilic carbon and nucleophilic carbon must be
brought close enough together to have orbital overlap before
cyclization may occur. In the case of 22 there are four
sp3 carbons between the electrophile and nucleophile. The
probability that the two required centers are adjacent is
very low, requiring a longlived reactive Lewis acid complex
if cyclization is to occur. Given the predisposition of
these systems to provide allylic alcohols to the exclusion
of cyclization, we were pleased to obtain a 24% yield of
cyclized product.

In this study we have learned a great deal about the

range of Lewis acidity which is useful in cyclizing 3-furyl

27

epoxides. There are many Lewis acids in this range which may
prove useful. We are continuing to experiment with magnesium
salts and species such as B(OCH3)nCl3_n among others. Thus
far the only solvent used in the attempted cyclizations has
been methylene chloride, other solvents such as ether, THF or
acetonitrile may prove useful.

As mentioned earlier, many functionalites have been used
as cyclization initiators. We are planning to expand this
study by the examination of cyclizations initiated by at r/3

unsaturated carbonyls and allylic alcohols, etc.

28

29

EXPERIMENTAL

General: Tetrahydrofuran (THF) was dried by distillation,
under nitrogen from sodium benzophenone ketyl; methylene
chloride was dried by distillation under nitrogen from calcium
hydride; 2&2 dimethylformamide (DMF) was dried by distilla-
tion at reduced pressure from phosphorous pentoxide; hexa-
methylphosphoramide (HMPA) was dried by distillation at
reduced pressure from calcium hydride; pyridine was dried
by distillation, under nitrogen, from calcium hydride;
diisopropyl amine was dried by distillation, under nitrogen,
from calcium hydride. Pet. ether refers to the 30—60°C
boiling point fraction of petroleum benzin. Diethyl ether
was purchased from Mallinkrodt Inc., St. Louis, Mo., and
used as received. n-Butyl lithium in hexane was purchased
from Aldrich Chemical Co., Milwaukee, Wis. and titrated by
the method of Watson and Eastham.17 Ethylaluminum dichloride
and diethylaluminum chloride were purchased as hexane
solutions from Alfa Products, Danvers, Ma., and used as re-
ceived. Magnesium metal was activated by successive
washings with lN aqueous hydrochloric acid, water, acetone,
ether and dried in a dessicator over phosphorous pentoxide
at reduced pressure.

Unless otherwise stated, all reactions were carried
out under an atmosphere of argon with the rigid exclusion
of moisture from all reagents and glassware.

Melting points were determined on a Thomas—Hoover

capillary melting point apparatus and are uncorrected.
Infrared spectra were recorded on a Pye-Unicam SP-1000
infrared spectrophotometer with polystyrene as standard.
Proton magnetic resonance spectra were recorded on a Varian
T-60 at 60MHz of a Bruker WM-250 spectrometer at 250MHz as
indicated, as solutions in deuterochloroform unless other-
wise indicated. Carbon magnetic resonance spectra were
recorded on a Bruker WM-250 spectrometer at 68.9MHz. Chemical
shifts are reported in parts per million on theS scale
relative to a tetramethylsilane internal standard. In NMR
descriptions br=broad, s=singlet, d=doublet, t=triplet,
q=quartet, m=multiplet and J=coupling constants in Hertz.
Electron impact (EI/MS) and chemical ionization (CI/MS)
mass spectra were recorded on a Finnigan 4000 with an INCOS
4021 data system.

Flash chromatography was performed according to the
procedure of Still 22 22 32 using the Whatman silica gel
mentioned and eluted with the solvents mentioned. Analytical
thin layer chromatography was run on either Macherey-Nagel
Polygram SIL G/UV254 precoated plastic sheets or Brinkman
Instruments SIL G/UV precoated glass plates. Spots were
visualized by either dipping into a solution of Vanillin
(1.59) in absolute ethanol (100ml) and concentrated sulfuric
acid (0.5ml) and heating with a heat gun or spraying with a
5% solution of molybdophosphoric acid in absolute ethanol

and heating to 120°C.

30

(3-furyl)-chloromethane (33)l6 - To a mechanically

 

stirred solution of LiCl (2.129, 0.05 mole) in anhydrous
DMF (40ml) was added a mixture of (3-fury1)-methanol (4.99,
0.05 mole) and 2,4,6-trimethylpyridine (6.669, 0.055 mole).
The resulting solution was cooled to 0°C in an ice-water bath
and methanesulfonyl chloride (6.39, 0.055 mole, distilled
from calcium hydride) was added over a period of 20 minutes.
The mixture became bright yellow and a thick suspension.
After stirring at 0°C for 2 hours the mixture was cast into
ice-water (150ml) and ether-pentane (1:1,150ml). The organic
phase was separated and washed with saturated aqueous cupric
nitrate (3x150m1), dried (Na2804) and concentrated 22 22222
to give a light yellow liquid. Distillation provided 4.89
76%, of 22 as a colorless liquid B.P.(25mm)=40°c. (lit.
B.P.%§7mm)=42-43°C)
EI/MS (70eV):118(M++2,ll.l), ll6(M+,34.5), 81(base)
lHNMR (60MHz)¢§:7.32 (t,J=2Hz,2H), 6.28 (d,J=2Hz,lH), 4.56
(s,2H)

3-methyl-but-2-en-l-ol - To a suspension of lithium

 

aluminum hydride (6.659, 0.175 mole) in ether (250ml) cooled
to 0°C was added a solution of 3,3-dimethyl acrylic acid

(5.09, 0.05 mole) in ether (100ml) over a period of 30 minutes.
The suspension was then heated under reflux for 14 hours.

The mixture was cooled to 0°C in an ice-water bath and 20%
aqueous sodium hydroxide (50ml) was carefully added dropwise.
The resulting suspension was filtered through celite 545,

the filter cake was rinsed with ether and the combined

filtrates were concentrated 22 vacuo to yield a colorless

31

liquid. Distillation provided 3.579, 83%, of 3—methyl-but-

2-en-l-ol as a colorless liquid. B.P. =72°C

(45mm)
lHNMR (60MHz)é‘:5.34(m,lH), 4.10(d,J=6Hz,2H), 3.8-3.6
(br s, 1H), l.88(s,3H), l.78(s,3H)

4-chloro-2-methyl-but-2-ene (36)16 - To a mechanically

 

stirred solution of LiCl (2.129, 0.05 mole) in DMF (40ml)
was added a mixture of 3-methyl-but-2-en-1-ol (4.39, 0.05
mole) and 2,4,6-trimethylpyridene (6.669, 0.055 mole).
The resulting solution was cooled to 0°C and methanesulfonyl
cloride (6.39, 0.055 mole, distilled from calcium hydride)
was added over a period of 20 minutes. After stirring at
0°C for 2 hours the mixture was cast into ice-water (150ml)
and ether-pentane (1:1,150ml). The organic phase was washed
with saturated aqueous cupric nitrate (3x150m1), dried
(NaZSO4), and concentrated 22_y2222 to give a yellow liquid.
Distillation gave 2.249, 43% of 22 as a colorless liquid
B.P.(93mm)=55°C
lHNMR (60MHz)é':5.34(m,lH), 4.00(d,J=9Hz,2H), 1.70

(br s,6H)

3—methyl-but-3-en-l-ol4petoluenesulfonate - To a solu-

 

tion of 3—methyl-but-3-en-l-ol (2.6g, 30 mmole) in pyridine
(20ml), cooled to 0°C in an ice-water bath, was added freshly
crushed p-toluenesulfonyl chloride (7.639, 40 mmole) in one
portion. The mixture was stirred at 0°C for 1 hour and then
placed in a freezer (-20°C) overnight. The resulting suspen-
sion was cast into a mixture of ice-water and concentrated

hydrochloric acid (509-50ml) and extracted with ether (150ml).

32

The organic phase was separated and washed with 1N aqueous
hydrochloric acid (100ml), saturated aqueous sodium bicar-
bonate (100ml), brine (100ml), dried (NaZSO4), and concentrated
22_22222 to yield 6.09, 83%, of a viscous yellow liquid which
was used without further purification.

3-methyl-3-epogy-buten-l-ol p-toluenesulfonate (47c) -

 

To a solution of 3-methyl-but-3-en-l-ol p-toluenesulfonate
(7.629, 30 mmole) in methylene chloride (50ml), cooled to
0°C in an ice-water bath, was added a solution of m-chloro-
perbenzoic acid (8.089, 30 mmole, 85%) in methylene chloride
(50ml) over a period of 30 minutes. The mixture was allowed
to stir for 3 hours at 0°C and the resulting suspension was
then suction filtered and the filtrate was taken up in ether
(150ml) and washed with 10% aqueous sodium bisulfite (2x100m1),
saturated aqueous sodium bicarbonate (100ml), water (100ml),
brine (100ml), dried (NaZSO4), and concentrated 22 ygggg to
yield a viscous liquid. The crude product was purified by
chromatography on a column of silical gel (60-230 mesh, 509,
40mm o.d., ether-pet. ether l:l, 30ml fractions) using the
flash technique. Fractions 8-13 provided 5.529, 68%, of
222 as a colorless liquid.
EI/MS (70eV):256(MI,2.l), 155(11), lOl(ll.6), 91(38.5),
84(24.4), 68(23.7), 43(base)
lHNMR (60MHz)J :7.76(d,J=8Hz,2H), 7.31(d,J=8Hz,2H)
4.l4(t,J=6.5Hz,2H), 2.61(s,lH), 1.96

(t,J=6.5Hz,2H), 1.31(S,3H)

33

4-iodo-2-methyl-1-epoxy-butene (47b) - To a solution
of 3—methyl-3-epoxy-buten-l-ol p-toluenesulfonate, 222, (2.029,
7.89 mmole) in acetone (25ml, dried over CaClz) was added
sodium iodide (1.509, 10 mmole) in one portion and the
solution heated under reflux for 4 hours. The resulting
suspension was cooled to room temperature and suction filtered.
The filtrate was diluted with ether (150ml) and washed with
water (100ml), 10% aqueous sodium bisulfite (100ml), water
(100ml), brine (100ml), dried (NaZSO4) and concentrated 22
22222 to yield a colorless liquid. Distillation of the crude
product provided 1.479, 88%, of 212_as a clear, colorless
liquid. B'P°(25mm)=58°C
EI/MS (70eV):212(M+,4.3), 194(1.13), 110(14.2), 85(25.4),

55(66.l), 43(base)

lHNMR (60MHz)J:3.ll(t,J=8Hz,2H), 2.60(AB,J =4Hz,2H), 2.12

AB
(m,2H), 1.28(S,3H)
IR(neat): 3000 2920, 1430, 1375, 1215, 1150, 1050, 895,
790, 720cm‘l
4-methyl-pent-4-en-1401 p-toluenesulfonate - To solution
of 4-methyl-pent-4-en—l-ol33 (3.09,30 mmole) in pyridine (16ml)
cooled to 0°C in an ice-water bath was added freshly crushed
p-toluenesulfonyl chloride (7.63g,40 mmole) in one portion.
The mixture was allowed to stir at 0°C for 1 hour and then
placed in the freezer (-20°C) overnight. The mixture was
cast into ice—concentrated hydrochloric acid (SOg-SOml) and
extracted with ether (150ml). The organic phase was washed

with 1N aqueous hydrochloric acid (100ml), brine (100ml),

dried (NaZSO4), and concentrated 22 vacuo to yield 7.629,

34

100% of a viscous yellow liquid. This product was used
without further purification.

4-methyl-4—epoxy-penten-l-ol p-toluenesulfonate (48c) -
To a solution of 4-methy1-pent-4-en-l-ol p-toluenesulfonate
(7.629, 30 mmole) in methylene chloride (40ml) cooled to
0°C in an ice-water bath was added a solution of m-chloroper-
benzoic acid (6.089, 30 mmole, 85%) in methylene chloride
(50ml) and the resulting suspension was stirred at 0°C for
1 hour and then overnight at room temperature. The mixture
was suction filtered and the filtrate was diluted with ether
(200ml) and washed with 10% aqueous sodium bisulfite (150ml),
saturated aqueous sodium bicarbonate (150ml), water (150ml),
brine (150ml), dried (NaZSO4) and concentrated 22.22222 to
yield a cloudy colorless liquid. The crude product was
purified by chromatography on a column of silica gel (60-230,
509, 40mm o.d., ether-pet. ether 1:1, 25ml fractions) using
the flash technique. Fractions 10-14 yielded 5.529, 68% of
222 as a colorless liquid.
lHNMR (60MHz)J :7.73(d,J=7.5Hz,2H), 7.30(d,J=7.5Hz,2H),

4.03(t,J=6Hz,2H), 2.44(s,3H), 1.63(m,4H),
l.29(s,3H)

5-iodo-4-methyl-l-epoxyjpentene 48b - To a solution of

 

4-methy1-4-epoxy-penten-1-ol p-toluenesulfonate, 222, (5.69, 20
mmole) in acetone (50ml, dried over CaClZ) was added sodium
iodide (3.39, 22 mmole) in one portion and the solution was
heated under reflux for 6 hours. The resulting suspension

was cooled to room temperature and suction filtered. The
filtrate was cast into water (200ml) and ether (200ml). The

organic phase was separated and washed with, 10% aqueous
35

sodium bisulfite (100ml), saturated aqueous sodium bicar-
bonate (100ml), water (100ml), brine (100ml), dried (NaZSO4),
and concentrated 22 22222 to yield a water white liquid.
Distillation of the crude product provided 3.799, 84.5%, of
222 as a colorless liquid. B°P‘(20mm)=62°c
lHNMR (60MHz) :3.20(m,2H), 2.58(s,2H), 2.10-1.53(m,4H),
l.29(s,3H)
EI/MS (70eV) 227(M++l,22), 226(M+,8), 199(26), 141(14), 100(82),
43(base)
IR(neat): 3000, 2930, 1460, 1800, 1385, 1225, 1180, 915,
840, 750cm‘1

4-methyl-pent-3-en-l-ol - To a suspension of lithium

 

aluminum hydride (0.959, 25 mmole) in ether (50ml) cooled

to 0°C was added a solution of ethyl-4-methyl-3-pentenoate34
(3.239, 23 mmole) in ether (50ml) over a period of 30 minutes.
The suspension was then heated under reflux for 3 hours. The
mixture was cooled to 0°C and 20% aqueous sodium hydroxide
(30ml) was carefully added. The resulting suspension was
filtered through celite and concentrated 22 22222 to yield a
colorless liquid. Distillation of the crude product provided
1.479, 67%, of 4-methyl-pent-3—en-l-ol as a colorless liquid.
=105°C)

B.P. =110°c. (lit. B.P.35

(112mm) (110mm)
lHNMR (60MHz)6 :5.12(m,lH), 3.60(t,J=7Hz,2H), 2.98(brs,lH),
2.17(M,2H), l.78(s,3H), 1.73(s,3H)
4-methyljpent-3-en-l-ol p-toluenesulfonate - To a solu-
tion of 4-methyl-pen-3-en-l-ol (3.159, 37.5 mmole) in pyridine
(18ml) cooled to 0°C in an ice-water bath was added freshly

crushed p-toluenesulfonyl chloride (7.629, 40 mmole) in one

portion. The mixture was stirred at 0°C for 1 hour and then
36

stored in the freezer (—20°C) overnight. The mixture was then
cast into ice-concentrated hydrochloric acid (509-50ml) and
extracted with ether (150ml). The organic phase was washed
with 1N aqueous hydrochloric acid (100ml), saturated aqueous
sodium bicarbonate, (100ml), water (100ml), brine (100ml),

dried (Na $04) and concentrated 22 vacuo to yield a viscous

2
yellow liquid which was used immediately in the next reaction.

5—iodo-2-met2yl-Zipentene 38 - To a solution of crude

 

4—methyl-pent-3-en-l-ol p-toluenesulfonate (4.409, 17.3
mmole) in acetone (50ml) was added sodium iodide (3.09, 20
mmole) and the solution heated under reflux for 14 hours.
The mixture was cooled to 0°C and suction filtered. The
filtrate was cast into water (200ml) and ether (150ml).
The organic phase was separated and washed with 10% aqueous
sodium bisulfite (150ml), water (150ml), brine (150ml) dried
(NaZSO4), and concentrated 22 22222 to yield a colorless
liquid. Distillation of the crude product provided 3.349,
92% of 22 as a colorless liquid. B.P.(45mm)=63°c (lit.
B.p.i§0mm)=72°c .
lHNMR (60MHz)J :5.04(m,lH), 2.50(t,J=8Hz,2H), l.99(m,2H),
1.66(s,3H), l.60(s,3H)

tri-(n-butyl)-stanyl methyl furan (50) - To a solution
of diiSOpr0pyl amine (4.449, 44 mmole) in anhydros THF (50ml)
cooled to 0°C in an ice-water bath was added n-butyl lithium
(25.8ml, 44 mmole) over a period of 10 minutes and the mix-
ture was allowed to stir for an additional 10 minutes after
the addition was complete. To the resulting solution was

added tri-(n-butyl) tin hydride (11.69, 40 mmole) over a

period of 10 minutes and the mixture allowed to stir for an

37

additional 15 minutes and then cooled to -25°C in a dry ice-
carbon tetrachloride bath. To the resulting yellow solution
was added (3-fury1)-chloromethane (4.669, 40 mmole), over a
period of 10 minutes. The cooling bath was removed and

the reaction allowed to stir at room temperature for 1 hour.
The mixture was then cast into ether (300ml) and saturated
aqueous NH4C1 (200ml). The organic phase was separated

and washed with water (200ml), brine (200ml), dried (NaZSO4)
and concentrated 22_y2222 to yield a yellow liquid. Dis-
tillation provided 11.239, 76%, of a colorless liquid.
=125°c (lit. 3.9.37

(005mm) (0.55mm)
EI/MS (70eV) 372(1.3), 355(6), 315(10), 291(28), 235(32),

B-P- =116-1l9°C)
201(18), 179(base)
lHNMR (60MHz):5:7.23(t,J=2,1H), 7.18(m,lH), 6.21(s,1H),
2.0-0.7(m,29H)

2—methyl-4-(3furyl)-but-2-ene (39) - To activated

 

magnesium metal turnings (0.2439, 10 mmole) covered by THF
(15ml) was added (3-furyl)-chloromethane (1.169, 10 mmole)
in one portion. The mixture was stirred at room temperature
until all the magnesium had reacted (about 1 hour). The
resulting golden solution was cooled to 0°C and l-bromo-2-

methyl-propene38

, 22, (1.359, 10 mmole) was added followed
immediately by FeCl3 (16mg, 0.01 mmole). The reaction mixture
immediately turned deep red. After the mixture had stirred

at 0°C for 1 hour it was cast into saturated aqueous NH4C1

(100ml) and ether (150ml). The organic phase was separated

38

and washed with water (100ml), brine (100ml), dried (NaZSO4),
and concentrated 22 22222 to yield a golden liquid. The crude
product was purified by chromatography in a column of silica
gel (230-400 mesh, 1009, 50mm o.d., ether-pet. ether 1:99,
30ml fractions), using the flash technique. Fractions 6‘9
provided 1.129, 82% of 22 as a colorless sweet-smelling
liquid.
EI/MS (70eV) l36(M+, base), 121(42), 93(41), 91(37), 77(36)
lHNMR (250MHz)6 :7.22 (t,J=2Hz,lH0), 7.04(m,lH), 6.l3(br s,
1H), 5.24(t,J=10Hz,lH), 3.10(d,J=10Hz,2H),
2.62(s,3H), 2.50(s,3H)
IR(neat): 2900, 1500, 1450, 1375, 1155, 1070, 1010, 870,
780cm-1

2-met2yl—4-(3-furyl)-but-l-ene (40) - To activated

 

magnesium metal turnings (0.2439, 10 mmole) covered by THF
(15ml) was added (3-fury1)-chloromethane (1.169, 10 mmole)

in one portion. The mixture was stirred at room temperature
until all the magnesium had reacted (about 1 hour). The result-
ing golden solution was cooled to 0°C in an ice-water bath

and 3-chloro-2-methyl-propene, 22, (0.909, 10 mmole) was added
followed immediately by LiZCuCl4 (0.12ml, 0.1M in THF). The
reaction mixture immediately warmed and turned black. After
the solution had stirred at 0°C for 1/2 hour it was cast into
saturated aqueous NH4C1 (100ml) and ether (100ml). The
organic phase was separated and washed with water (100ml),

brine (100ml), dried (NaZSO4) and concentrated 22 vacuo to

yield a colorless liquid. The crude product was purified

39

by chromatography on a column of silica gel (230-400 mesh,
1009, 50mm o.d., ether-pet. ether 1:99, 30ml fractions),
using the flash technique. Fractions 6-11 provided 1.109,
81% of 22 as a colorless liquid.
lHNMR (250MHz)5 :7.28(t,J=l.8Hz,lH), 7.13(m,1H), 6.19(br s,
1H), 4.62(br s,2H), 2.27(m,4H), l.76(s, 3H)
EI/MS (70eV) 136(M+,15), 121(ll.7), 94(46.7), 81(base)
IR(neat): 2950, 2870, 1500, 1150, 1080, 1025, 900, 890,
780cm"l

2-methyl-5-(3-furyl)-pent-2-ene (41) - To activated
magnesium turnings (0.2439, 0.01 mole) covered by THF (15ml)
was added (3-furyl)-chloromethane (1.169, 0.01 mole) in
one portion. The mixture was stirred at room temperature
until all the magnesium had reacted (about 1 hour). The
resulting golden solution was cooled to 0°C in an ice-water
bath and 4-chloro-2-methyl-2-butene, 22, (l.04g,0.01 mole) was
added, followed immediately by LiZCuCl4 (0.12m1, 0.1M in
THF). The reaction mixture immediately warmed and turned
black. After the solution had stirred at 0°C for 1/2 hour
it was cast into saturated aqueous NH4C1 (100ml) and ether
(100ml). The organic phase was separated and washed with
water (100ml), brine (100ml), dried (NaZSO4) and concentrated
22 ygggg to yield a golden liquid. The crude product was
purified by chromatography on a column of silica gel (230-
400 mesh, 1009, 50mm o.d., ether-pet. ether 1:99, 30ml
fractions) using the flash technique. Fractions 8-14

yielded 1.239, 82%, of 22 as a colorless sweet smelling liquid.

40

EI/MS (70eV) 150(M+,52), 135(12), 81(67), 69(base)
lHNMR (250MHz)5 :7.32(t,J=2Hz,lH), 7.12(m,lH), 6.23(s,1H),
5.12(t,J=10Hz), 2.6-2.0(m,4H), 1.73(s,3H),
1.65(s,3H)
IR(neat): 2890, 1500, 1440, 1380, 1160, 1080, 1025, 870,
780cm-1

2-metgyl-6-(3-furyl)-hex-2-ene (43) - To "activated"
magnesium turnings (0.2439, 0.01 mole) covered by anhydrous
tetrahydrofuran (15ml) was added (3-fury1)-chloromethane
(1.169, 0.01 mole) in one portion. The mixture was stirred
under argon at room temperature until all the magnesium had
reacted (about 1 hour). The resulting golden liquid was
cooled to 0°C in an ice—water bath and S-iodo-2-methy1-2-
pentene, 22, (2.109, 0.01 mole) was added, followed immediately
by LiZCuCl4 (0.12m1, 0.1M, in THF). The reaction mixture
warmed and slowly turned black. After the solution had
stirred at 0°C for an additional 1 hour it was cast into
saturated aqueous NH4C1 (100ml) and ether (100ml). The
organic phase was separated and washed with water (100ml)
saturated brine (100ml), dried (NaZSO4) and concentrated 22
ygggg to yield a light yellow liquid. The crude product was
purified by chromatography on a column of silica gel (230-
400 mesh, 1009, 50mm o.d., ether; pet. ether 1:99, 30ml
fractions). Fractions 7-11 were combined and yielded 1.199

(73%) of 22 as a colorless liquid.

41

EI/MS (70eV): 164(M+,2), 149(3), 121(9.1), 108(8), 94(14),
82(base)
lHNMR (250MHz)J :7.29(t,J=2Hz,lH), 7.16(m,lH), 6.20(s,1H),
5.18(t,J=6Hz,lH), 2.38(t,J=6Hz,2H), 2.36-1.03
(m,4H), 1.64(s,3H), l.58(s,3H)
IR(neat): 2950, 2880, 1500, 1160, 1070, 1025, 905, 865,
780cm-1
2-methyl-6-(3-fury1)-hex-2-ene (43) - To activated
magnesium metal turnings (0.2439, 0.01 mole) covered by
anhydrous THF (10ml) was slowly added 1,2-dibromoethane
(1.889, 0.01 mole). The mixture was stirred at room termera—
ture for 1 hour and then heated under reflux until all the
magnesium had reacted (about 1/2 hour). This solution was
cooled to 0°C in an ice-water bath. To a solution of 2-(3-
furyl) ethyl-tri-(n-butyl) stannane, 22, (3.849, 0.01 mole),
cooled in a dry ice-isopropanol bath was added n-butyl
lithium (5.8ml, 0.01 mole, in hexane) over a period of 5
minutes. The solution was allowed to stir at -78°C for 10
minutes and transferred 222 cannula to the magnesium bromide-
THF solution prepared above. This mixture was allowed to
stir at 0°C for 1 hour then 4-chloro-2-methy1-but-2-ene, 22,
(1.049, 10 mmole) was added in one portion followed immediately
by LiZCuC14 (0.2ml, 0.1M in THF). After 5 minutes the result-
ing mixture was cast into saturated aqueous NH4C1 (150ml)
and ether (150ml). The organic phase was separated and

washed with water (100ml), brine (100ml), dried (NaZSO4),

and concentrated 22 vacuo to yield a yellow liquid. The

42

crude product was purified by chromatography on a column of
silica gel (230-400 mesh, 1009, 50mm o.d., pentane, 50ml
fractions) using the flash technique. Fractions 8—14 yielded
1.349, 82%, of 22 as a colorless liquid. For spectral data
see previous preparation.

2-methyl-4-(3-furyl)-2-epoxyrbutene (19) - To a stirred

 

solution of 2-methy1-4-(3—furyl)-but-2-ene, 22, (1.369, 0.01

mole) in methlene chloride (30ml) cooled to 0°C in an ice-

water bath was added a solution of m-chloroperbenzoic acid

(2.239, 0.011 mole, 85%) in methylene chloride (50ml) over

a period of 1/2 hour. The resulting mixture was allowed

to stir at 0°C for 1/2 hour. The resulting suspension was

suction filtered and the filtrate cast into 10% aqueous

sodium bisulfite (150ml) and ether (200ml). The organic

phase was separated and washed with saturated aqueous sodium

bicarbonate (100ml), water (100ml), brine (100ml), dried

(NaZSO4) and concentrated £2.22222 to yield a light yellow

liquid. The crude product was purified by chromatography

on a column of silica gel (60-230 mesh, 759, 50mm, o.d.,

ether-pet. ether 1:4, 40ml fractions) using the flash tech—

nique. Fractions 6-11 yielded 1.209, 79%, of 22 as a color-

less liquid.

EI/MS (70eV):,{152(M+,4.5) 137(base), 123(6.8), 108(29)

lHNMR (250MHz) :7.42(t,J=3Hz,1H), 7.27(s,1H), 6.30(s,1H),
2.89(t,J=6Hz,1H), 2.70(d of AB,J=6Hz,12Hz,2H)
1.42(s,3H), 1.40(s,3H)

13CNMR (69.8MH2) :144.4, 140.7, 122.0, 112.35, 64.94, 59.77,

26.24, 26.06

43

IR(neat): 2965, 2925, 1500, 1445, 1375, 1155, 1125, 1020,

870, 780, 760cm“1
2-methyl-4-(3-fu;yl)-l—epoxy-butene (20) - To a stirred
solution of 2-methyl—4—(3-furyl)-but-l-ene, 22, (1.39, 10 mmole)

in methylene chloride (30ml) cooled to 0°C in an ice-water

bath was added a solution of m-chloroperbenzoic acid (2.029,

10 mmole, 85%) in methylene chloride (50ml) over a period of

1/2 hour. The mixture was allowed to stir at 0°C for 1/2 hour.

The resulting suspension was suction filtered and the filtrate

cast into 10% aqueous sodium bisulfite (150ml) and ether

(200ml). The organic phase was separated and washed with

saturated aqueous sodium bicarbonate (100ml), water (100ml),

brine (100ml), dried (Na2804) and concentrated 22 22222_to

yield a light yellow liquid. The crude product was purified

by chromatography on a column of silica gel (60-230 mesh,

759, 50mm o.d., ether-pet. ether 1:4, 40ml fractions),

using the flash technique. Fractions 6—11 provided 0.389,

25% of 22 as a colorless liquid.

lHNMR (250MHz)J':7.21(t,J=2Hz,lH), 7.09(m,lH), 6.23(br s,lH),

2.53(m,4H), 1.83(m,2H), l.38(s,3H)

EI/MS (70eV):156(M+,54.6), 139(84.3), 121(43.13), 112(63.10)
93(48.7), 81(67.0), 55(base)

IR(neat): 2930, 2860, 1500, 1450, 1430, 1390, 1175, 1030,

890cm'l

2—methy1-5—(3-furyl)-2—epoxy-pentene (21) - To a solution

of 2-methyl-5-(3-fury1)-pent-2-ene, 22, (1.509, 10 mmole)

44

in methylene chloride (30ml) cooled to 0°C in an ice-water

bath was added a solution of m-chloroperbenzoic acid (2.029,

10 mmole, 85%) in methylene chloride (50ml) over a period of

1/2 hour. The mixture was allowed to stir at 0°C for 1

hour after the addition was completed. The resulting

suspension was suction filtered and the filtrate was cast

into 10% aqueous sodium sulfite (150ml) and ether (150ml).

The organic phase was separated and washed with saturated

aqueous sodium bicarbonate (100ml), water (100ml), brine

(100ml), dried (Na2804) and concentrated 22 22222 to yield

a light yellow liquid. The crude product was purified by

chromatography on a column of silica gel (60-230 mesh, 709,

50mm o.d., ether-pet 1:4, 30ml fractions) using the flash

technique. Fractions 8-13 provided 1.379, 83% of 228b as a

colorless liquid.

EI/MS (70eV):166(M+,7.1), 151(12), 33(10), 123(13.4), 108
(42.8), 95(39.4), 85(75.0), 81(83.4), 72(38.5),
59(base)

lHNMR (250MHz)J :7.39(t,J=2Hz,1H), 7.22(s,1H), 6.29(s,lH),

2.78(t,J=6Hz,lH), 2.56(m,2H), l.78(q,J=
8Hz,2H), l.32(s,3H), 1.21(s,3H)

IR(neat): 2980, 2940, 2880, 1500, 1440, 1380, 1160, 1115,

1025, 925, 875, 790cm'1

l3CNMR (69.8MHz)6':155.8, 141.4, 114.16, 110.00, 76.59,

37.81, 28,64, 21.20, 18.99, 24.1

2-methy1-6-(3-furyl)-2-epoxy-hexene (23) - To a solution

 

of 2-methyl-6-(3-fury1)-hex-2-ene, 22, (1.649, 10 mmole) in

methylene chloride (30ml) cooled to 0°C in an ice-water bath

was added a solution of m-chloroperbenzoic acid (2.029, 10

45

mmole, 85%) in methylene chloride (50ml) over a period of
1/2 hour. The mixture was allowed to stir at 0°C for 1/2 hour.
The resulting suspension was suction filtered and the filtrate
was cast into 10% aqueous sodium bisulfite (150ml) and ether
(150ml). The organic phase was separated and washed with
saturated aqueous sodium bicarbonate (100ml), water (100ml),
brine (100ml), dried (NaZSO4), and concentrated 22 22222 to
yield a light yellow liquid. The crude product was purified
by chromatography on a column of silica gel (60-230 mesh,
709, 50mm o.d., ether-pet. ether 1:4, 25ml fractions) using
the flash technique. Fractions 8-11 provided 1.409, 78%
of 22 as a colorless liquid.
EI/MS (70eV):180(M+,1.7), 151(7.4), 135(5.6), 121(14), 107
(11.3), 98(2), 94(base)
lHNMR (250MHz)é :7.28(t,J=2Hz,lH), 7.18(t,J=2Hz,1H), 6.21
(br s,lH), 2.67(t,J=6Hz,1H), 2.45(m,2H), 1.58
(m,4H), l.22(s,3H), 1.18(s,3H)
IR(neat): 2980, 2950, 2880, 1500, 1440, 1390, 1150, 1115,
1020, 915, 875, 790, 720cm"l

2-methyl-6-(3-furyl)-1-epoxy-hexene (24) - To a solution

 

of tri-(n-butyl)-stannyl methyl furan, 22, (1.859, 5 mmole)

in anhydrous THF (Sml) cooled to -78°C in a dry ice-
isopropanol bath was added n-butyl lithium (3.33m1, 5 mmole,

in hexane) over a period of 5 minutes. The solution was
allowed to stir at -78°C for 10 minutes, HMPA (0.8969, 5 mmole)
was then added in one portion and the mixture stirred at

-78°C for 10 minutes. The resulting solution was transferred
222 cannula to a solution of 5-iodo-2-methyl-l-epoxy-pentene,

48b, (1.129, 5 mmole) in anhydrous THF (10ml) cooled to

46

—25°C in a dry ice-carbon tetrachloride bath. Upon addition
of the organolithium reagent the initially colorless solution
turned deep red-brown. The cooling bath was removed and

the mixture allowed to stir at room temperature overnight.

The solution was cast into saturated aqueous NH Cl (100ml)

4

and ether (100ml). The organic phase was separated and washed

with water (100ml), brine (100ml), dried (NaZSO4), and concen-

trated 22_22222 to yield a yellow liquid. The crude product

was purified by chromatography on a column of silica gel

(230-400 mesh, 759, 40mm o.d., ether-pet. ether 1:4, 25ml

fractions) using the flash technique. Fractions 14-18

yielded 0.619, 68% of 22 as a light yellow liquid.

EI/MS (70eV):180(M+,12), 163(11), 149(14.4), 135(28), 121

(18.7), 108(60), 82(base)

lHNMR (250MHz)é :7.36(t,J=3Hz,1H), 7.21(t,J=3Hz,lH), 6.24
(s,lH), 3.90(t,J=9Hz,1H), 3.28(m,lH),
2.58(m,2H), 2.42(t,J=9Hz,2H), 1.66-1.38(m,4H),
1.31(s,3H)

IR(neat): 3010, 2990, 2925, 1540, 1500, 1445, 1380, 1150,

1110, 1070, 900, 805, 780cm'1
7,7-dimethyl-6-hydroxy-4,5-6,7-tetrahydrobenzofuran (27)

- To a solution of 2-methyl-5-(3-fury1)-2-epoxy-pentene, 22,

(0.19, 0.60 mmole) in methylene chloride (10ml) cooled to

-25°C in a dry ice-carbon tetrachloride bath was added

freshly distilled boron trifloride etherate (0.0289, 0.20

mmole). The mixture was quenched (-25°C) with saturated

47

aqueous NH4CL (Sml) and allowed to warm to room temperature.
The two phase mixture was cast into ether (50ml) and
saturated aqueous NH4C1 (50ml). The organic phase was
separated and washed with water (50ml), brine (50ml), dried
(NaZSO4), and concentrated 22 22222_to yield a dark red
liquid. The crude product was purified by chromatography
on a column of silica gel (230—400 mesh, 409, 30mm o.d.,
ether-pet. ether 1:], 10ml fractions) using the flash
technique. Fractions 14-17 provided 47mg, 47% of 22 as
a light yellow liquid.
EI/MS (70eV):166(M+,40.4). 151(9.4), 133(4.80), 122(base)
lHNMR (250mm); :7.26(d,J=l.8Hz,lH), 6.14(d,J=1.8Hz,lH),
3.83(br s, 1H), 3.40(t of d, Jt=8Hz, Jd=
6Hz,2H), l.92(m,2H), l.38(s,3H), l.22(s,3H)
13CNMR (69.8MHz) 5:155.8, 141.4, 114.1, 109.9, 76.3, 37.7,
28.0, 25.5, 21.1, 18.9
IR(neat): 3435(br), 2900, 1620, 1500, 1470, 1385, 1360,
1280, 1150, 1120, 1085, 1045, 890, 780cm”l

2-methy1-5-(3-furyl)-1-epoxy-pentene (22) - To a solution

 

of tri-(n-buty1)-stanyl methyl furan 22 (1.859, 5 mmole)

in anhydrous THF (Sml) cooled to -78°C in a dry ice-iso-
propanol bath was added n-butyl lithium (3.33ml, 5 mmole,
in hexane) over a period of 5 minutes. The solution was
allowed to stir at -78°C for 10 minutes and then HMPA
(0.8969, 5 mmole) was added in one portion and the mixture
allowed to stir at -78°C for an additional 10 minutes. The

resulting solution was transferred via cannula into a solution

48

of 4-ioeo-2-methyl-l-epoxy-butene, 222, (1.069, 5 mmole) in
THF (10ml) which was cooled to -25°C in a dry ice-carbon
tetrachloride bath. Upon addition of the organo-lithium
reagent the colorless solution turned deep red-brown. The
cooling bath was removed and the mixture allowed to stir at
room temperature overnight. The solution was cast into
saturated aqueous NH4C1 (100ml), and ether (100ml). The
organic phase was separated and washed with water (100ml),
brine (100ml), dried (NaZSO4), and concentrated 22 22222 to
yield a yellow liquid. The crude product was purified by
chromatography on a column of silica gel (230-400 mesh, 759,
40mm o.d., etherpet. ether, 1:4, 25ml fractions) using the
flash technique. Fractions 12-17 provided 0.669, 80%, of 22
as a light yellow liquid.
EI/MS (70eV):l66(M+,2.3), l49(8.1), 141(19). 135(8.6),
129(7.8), 121(12.0), 109(17.6), 94(base)
lHNMR (250MHz)J :7.32(t,J=3Hz,1H), 7.20(m,lH), 6.22(m,1H),
3.18(m,2H), 2.76-2.50(m,2H), 1.77-1.51(m.2H),
l.32(s,3H)
IR(neat): 2925, 2860, 1500, 1450, 1390, 1160, 1070, 1025,
905, 975, 890cm'1

Attempted cyclization of 21 with ethylaluminum dichloride

 

- To a solution of 2-methy1—5-(3-fury1)-2-epoxy-pentene 22
(0.19, 0.60 mmole) in methylene chloride (10ml) cooled to
-78°C was added ethylaluminum dichloride (l.22m1, 1.8 mmole,
1.47M in hexane) and the mixture warmed slowly to -25°C.

The solution was allowed to stir at -25°C for 1/2 hour and

49

then quenched by addition of saturated aqueous NH Cl (10ml).

4
The reaction was warmed to room temperature and the resulting

two phase mixture was cast into saturated aqueous NH Cl (50ml)

4
and ether (50ml). The organic phase was separated and washed
with 1N aqueous hydrochloric acid (50ml), water (50ml), brine
(50ml), dried (NaZSO4) and concentrated 22 22222 to yield

a yellow liquid. Flash chromatography of the crude product
provided 0.0169, 16%, of 22 as a light yellow liquid.

Attempted cyclization of 21 with diethylaluminum chloride

 

- To a solution of 2-methyl-5-(3-furyl)-2-epoxy-pentene 22
(0.19, 0.60 mmole) in methylene chloride (10ml) cooled to
0°C was added diethylaluminum chloride (1.22m1, 1.8 mmole,
1.48M in hexane) and the mixture immediately turned yellow.
The solution was allowed to stir at 0°C for 1 hour and then
quenched by the addition of saturated aqueous NH4C1 (10ml).
The resulting two phase mixture was cast into saturated
aqueous NH4C1 (50ml), and ether (50ml). The organic phase
was separated and washed with 1N aqueous hydrochloric acid,
(50ml), water (50ml), brine (50ml), dried (NaZSO4) and
concentrated 22 ygggg to yield a yellow liquid. Flash
chromatography of the crude product provided 0.0229, 22%,
of 22 as a light yellow liquid.

Attempted cyclization of 21 with alumina - To a solution

 

of 2-methy1-5-(3-furyl)-2-epoxy-pentene, 22 (0.19, 0.60 mmole)
in hexane (15ml, distilled from calcium hydride) was added

basic alumina (2.09, activity I) and the suspension stirred

50

at room temperature for 24 hours. Methanol (10ml) was added
and the mixture suction filtered and the alumina rinsed with
methanol (15ml). The solvent was removed 22 22222 to yield
a colorless liquid. Flash chromatography of the crude product
provided 0.0329, 32% of 22 as a light yellow liquid.

Attempted cyclization of 21 with magnesium bromide - To
a solution of 2-methy1-5-(3-fury1-2-epoxy-pentene 22 (0.109,
0.60 mmole) in methylene chloride (10ml) was added magnesium
bromide-THF complex (0.3179, 1.2 mmole) followed immediately
by triethyl amine (0.0619, 0.60 mmole, distilled under nitrogen
from calcium hydride). The mixture was allowed to stir at
room temperature for 24 hours. The reaction was quenched
by addition of 1N aqueous hydrochloric acid (10ml) and the
resulting two phase mixture cast into 1N aqueous hydrochloric
acid (50ml) and ether (50ml). The organic phase was sepa-
rated and washed with water (50ml), brine (50ml), dried
(NaZSO4), and concentrated 22 y2222_to yield a light yellow
liquid. Flash chromatography of the crude product provided
0.0439, 43% of 22 as a light yellow liquid.

Triisopropoxytitanium chloride - To a solution of ti-

 

tanium tetraiopropoxide (6.399, 22.5 mmole) in methylene
chloride (40ml) was added titanium tetrachloride (1.429,
7.5 mmole). This mixture was stored at -20°C.

Attempted cyclization of 21 with triisoprqpoxytitanium
chloride - To a solution of 2-methyl—5-(3-fury1)-2-epoxy-
pentene, 22, (0.109, 0.60 mmole) in methylene chloride
(10ml) cooled to 0°C in an ice water bath was added tri-

isopropoxytitanium chloride (2.40m1, 1.8 mmole, 0.75M in

51

methylene chloride). The mixture was allowed to stir at

0°C for 1 hour and then at room temperature for 1 hour.

The reaction was quenched by the addition of saturated
aqueous NH4C1 (10ml) and the resulting two phase mixture was
cast into saturated aqueous NH4C1 (50ml) and ether (50ml).
The organic phase was separated and washed with 1N aqueous
hydrochloric acid (50ml), water (50ml), brine (50ml),

dried (NaZSO4) and concentrated 22 22222 to yield a light
yellow liquid. Flash chromatography of the crude product
provided 0.0789, 78%, of 22 as a light yellow liquid.

Zinc Iodide - To zinc metal (3.269, 50 mmole, 30 mesh)
covered by ether (30ml), was added a solution of iodine
(12.79, 50 mmole) in ether (80ml) over a period of 1 hour.
The resulting mixture was heated to reflux until all brown
color had disappeared (about 2 hours). The solvent was
removed 22.22222 and the residue dried in a dessicator over
phosphorous pentoxide at reduced pressure to provide (19.09,

99%) of zinc iodide-ether complex.

Attempted cyclization of 21 with Zinc iodide - To a

 

solution of 2-methyl-5-(3-furyl)-2-epoxy—pentene, 22,

(0.10,g, 0.60 mmole) in methylene chloride (10ml) was added
anhydrous sodium acetate (50mg, 0.60 mmole) followed immediately
by zinc iodide-ether complex (0.479, 1.2 mmole) and the mix-
ture allowed to stir in the dark for 3 hours. The reaction

was quenched by addition of saturated aqueous NH4C1 (10ml)

52

and the resulting two phase mixture cast into saturated

aqueous NH4C1 (50ml) and ether (50ml). The organic phase

was separated and washed with 10% aqueous sodium bisulfite

(50ml), water (50ml), brine (50ml), dried (NaZSO4) and

concentrated 22_y2222 to yield a light yellow liquid. Flash

chromatography of the crude product provided 0.0719, 71%,

of 22 as a light yellow liquid.

7—methy1-7-hydroxmethyl-4,5-6.7-tetrahydrobenzofuran (28) -

To a solution of 2—methyl-5-(3-fury1)-l-epoxy-pentene, 22,

(0.109, 0.60 mmole) in methlene chloride (10ml) cooled to

0°C was added stock titanium catalyst (2.40m1, 1.8 mmole,

0.75M in methylene chloride). The mixture was allowed to

stir at 0°C for 1 hour and then at room temperature for 1

hour. The reaction was quenched by the addition of saturated

aqueous NH4C1 (10ml) and the resulting two phase mixture

cast into saturated aqueous NH4C1 (50ml) and ether (50ml).

The organic phase was separated and washed with 1N aqueous

hydrochloric acid (50ml), water (50ml), brine (50ml), dried

(NaZSO4) and concentrated 22'y2222_to yield a light yellow

liquid. The crude product was purified on a column of silica

gel (230-400 mesh, 40mm o.d., ether-pet. ether 1:1, 25ml

fractions) using the flash technique. Fractions 13-17

provided 0.0899, 89%, of 22 as a light yellow liquid.

EI/MS (70eV):166(M+, 8.8), 149(4.4), 135(base)

lHNMR (250MHz)5 :7.21(d,J=1.8Hz,lH), 6.15(d,J=8Hz,lH),
3.52(s,2H), 2.38(m,2H), 1.96(m,2H), 1.24
(s,3H)

IR(neat): 3440, 2940, 1500, 1380, 1205, 1160, 1040, 890,

7400m’1

53

8,8-dimethy1-7-hydroxy-4,5-7,8-tetrahydrocyclohepta-

 

6H-[b] furan 29 - To a solution of 2-methy1-6-(3-furyl)—

 

2—epoxy-hexene (0.109, 0.55 mmole) in methylene chloride (10ml)

was added anhydrous sodium acetate (0.0459, 0.55 mmole)

followed immediately by zinc iodide-ether complex (0.6489,

1.5 mmole) and the mixture allowed to stir in the dark for

3 hours. The reaction was quenched by the addition of

saturated aqueous NH4C1 (10ml) and the resulting two phase

mixture cast into saturated aqueous NH4C1 (50ml) and ether

(50ml). The organic phase was separated and washed with

10% aqueous sodium bisulfite (50ml), water (50ml), brine

(50ml), dried (NaZSO4) and concentrated 22 22222 to yield

a light yellow liquid. The crude product was purified on a

column of silica gel (230-400 mesh, 709, 40mm o.d., ether-

pet. ether 1:1, 24ml fractions) using the flash technique.

Fractions 15-20 provided 0.0889, 88%, of 22 as a light yellow

liquid.

EI/MS (70eV):l66(M+,32.2), 151(12.6), l49(17.9), 133(7.4),

122(base)

lHNMR (250MHz) ,5 :7.24(d,J=l.98Hz,1H), 6,l3(d,J=l.98Hz,1H)
3.73(d of d, J=4.21Hz,1H), 2.47(m,2H),
l.9l(m,6H), 1.30(s,3H), l.22(s,3H)

IR(neat): 3430, 2980, 1750, 1620, 1500, 1470, 1380, 1360,

1285, 1160, 1115, 1090, 1030, 890, 730, 705cm-l

Fractions 11-12 provided 9mg, 9%, of a mixture of allylic

alcohols

lHMNR (60MHz)5 :7.39(t,J=3Hz,2H), 7.21(t,J=3Hz,2H), 6.24
(s,2H), 4.85(s,lH), 4.80(s,1H), 4.10(m,1H),

3.62(br m 2H), 3.28(m,2H), 2.60(m,4H),

54

2.44(t,J=9Hz,4H), l.31(s,6H).
8-methy1—8-hydroxymethyl-4,5-7,8-tetrahydrocyclohepta-

6H-[b]-furan 30 - To a solution of 2-methy1-6-(3-furyl)-l-

 

epoxy-hexane (0.109, 0.55 mmole) in methylene chloride (10ml)
was added anhydrous sodium acetate (0.0459, 0.55 mmole)
followwed immediately by zinc iodide-ether complex (0.6489,
1.5 mmole) and the mixture stirred in the dark for 6 hours.
The reaction was quenched by the addition of saturated
aqueous NH4C1 (10ml) and the resulting two phase mixture
cast into saturated aqueous NH4C1 (50ml) and ether (50ml).
The organic phase was separated and washed with 10% aqueous
sodium bisulfite (50ml), water (50ml), brine (50ml), dried
(NaZSO4) and concentrated 22,22222 to yield a yellow liquid.
The crude product was purified on a column of silica gel
(230-400 mesh, 709, 40 mm o.d., ether- pet. ether 1:1,
30ml fractions) using the flash technique. Fractions
14-16 provided 0.023 g, 23% of 22 as a light yellow liquid.
EI/MS (70eV):l80(M+,10.0), 150(11.7), l49(base)
lHNMR (250MHz)c5:7.17(d,J=l.7Hz,lH), 6.12(d,J=1.7Hz,1H), 3.79
d,J=1l.le,1H), 3.58(d,J=11.1Hz,1H), 2.47
(m,2H), 1.96-l.31(br m,6H), l.22(s,3H)
l3CNMR (69.8MHz)6 :155.6, 141.1, 113.9, 109.8, 76.6, 37.7,
28.0, 25.7, 21.3, 19.0
Fractions 9-12 provided 52mg, 52% of a mixture of allylic
alcohols.
lHNMR (250MHz) :7.28(t,J=2Hz,2H), 7.16(s,2H), 6.18(s,2H),
4.96(s,1H), 4.80(s,1H), 3.90(br s, 1H),

55

3,80(t,J=6Hz,2H), 3.38(m,4H), 2.28(m,4H),
2.10-108(m,6H)
3-l3-hydroxy—pallesencin-A (51) - To a solution of 2.6-
dimethy1-9-(3-furyl)-2,6 nonadiene-oxide-2 (0.209, 0.85 mmole,
prepared according to procedure of S.P. Tanis) in methylene
chloride (10ml) cooled to 0°C was added stock titanium
catalyst (3.4m1, 2.55 mmole, 0.75M in methylene chloride).
The mixture was allowed to stir at 0°C for 1 hour and then
at room temperature for 1 hour. The reaction was quenched
by the addition of saturated aqueous NH4C1 (10ml) and the
resulting two phase mixture cast into saturated aqueous NH4C1
(10ml) and the resulting two phase mixture cast into saturated

aqueous NH Cl (50ml) and ether (50ml). The organic phase

4
was separated and washed with 1N aqueous hydrochloric acid
(50ml), water (50ml), brine (50ml), dried (NazSO4) and
concentrated 22 ygggg to yield a yellow liquid. The crude
product was purified on a column of silica gel (230-400 mesh,
709, 50mm o.d., ether-pet. ether 1:3, 25 m1 fractions) using

the flash technique, fractions 16—19 provided 0.1189, 59%

of 22_as a white solid. m.p. 120—122°c (lit. m.p.3°=122-122.5°c)
EI/MS (70eV):234(M+,46.6), 219(82), 201(base)

lHNMR (250MHz) 6 :7.13(d,J=1.6Hz,lH), 6.02(d,J=l.6Hz,lH),
3.31(m,3H), 3.43(m,4H), 2.22(m.lH), 1.5-2.1
(m,4H), l.18(s,3H), l.07(s,3H), U.89(m,3H)

IR(neat): 3490, 2900, 1500, 1450, 1370, 1200, 1080, 1020cm-l

56

Attempted cylization of 19 with triisoPropoxytitanium
chloride - To a solution of 2-methyl-4-(3-fury1)-2-epoxy-
butene, 22, (0.109, 0.66 mmole) in methylene chloride (10ml)
cooled to 0°C in an ice—water bath Was added triisopropoxy-
titanium chloride (2.40m1, 1.8 mmole, 0.75M in methylene
chloride). The mixture was allowed to stir at 0°C for 1
hour and then at room temperature for 3 hours. The reaction

was quenched by the addition of saturated aqueous NH Cl (10ml)

4
and the resulting two phase mixture was cast into saturated
aqueous NH4C1(40ml) and ether (40ml). The organic phase
was separated and washed with 1N hydrochloric acid (50ml),
water (50ml), brine (50ml), dried (NaZSO4), and concentrated
22 22222_to provide a light yellow liquid. The crude product
was purified by chromatography on a column of silica gel
(230-400 mesh, 409, 30mm o.d., ether—pet. ether 1:1, 10ml
fractions) using the flash technique. Fractions 10-14
provided 80mg, 80% of a mixture of allylic alcohols.
lHNMR (250MHz)d’:7.34(t,J=3Hz,2H), 7.24(s,2H), 6.28(s,2H),
4.90(s,1H), 4.79(s,1H), 3.60(d,J=6Hz,2H),
2,48(m,4H), 1.53(s,3H), 1.48(s,3H).
Attempted cyclization of 20 with triisopropoxytitanium
chloride - To a solution of 2-methyl-4-(3-fury1)-l-epoxy—
butene, 22, (0.109, 0.66 mmole) in methylene chloride (10ml)
cooled to 0°C in an ice-water bath was added triisopropoxy-
titanium chloride (2.40m1, 1.8 mmole, 0.75M in methylene

chloride). The mixture was allowed to stir at 0°C for 1 hour

and then at room temperature for 3 hours. The reaction was

57

quenched by the addition of saturated aqueous NH4C1 (10ml)

and the resulting two phase mixture was cast into saturated

aqueous NH4C1 (50ml) and ether (50ml). The organic phase

was separated and washed with 1N hydrochloric acid (50ml,

water (50ml), brine (50ml), dried (NaZSO4), and concentrated

22 22222 to yield a yellow liquid. The crude product was

purified by chromatography on a column of silica gel (230-

400 mesh, 409, 30mm o.d., ether pet-ether 1:1, 10ml fractions)

using the flash technique. Fractions 9-12 provided 0.0729,

72%, of a mixture of allylic alcohols.

lHNMR (60MHz)6 :7.26(t,J=2Hz,2H), 7.18(M,2H), 5.48(t,J=8Hz,1H),
4.94(s,1H), 4.83(s,1H), 4.0(br s,2H), 3.49
(s,2H), 3.12(d,J=6Hz,4H), 2.40(m,8H), 1.86
(2,3H), l.63)s,3H).

2-(3-fury1)-ethyl-tri-(n—butyl) stannane 53 - To

activated magnesium turnings (0.2439, 0.01 mmole) covered

by THF (15ml) was added (3-fury1) chloromethane (1.169, 0.01

mole) and the mixture allowed to stir at room temperature

until all the magnesium had reacted (about 1 hour). The

resulting golden liquid was cooled to 0°C in an ice-water

bath and iodomethyl-tributyl stannane22

(3.239, 7.5 mmole) was
added in one portion followed immediately by LiZCuC14

(0.2m1, 0.1M in THF). The reaction mixture immediately

became solid. The mixture was diluted with saturated aqueous
NH4C1 (100ml) and cast into pentane (100ml). The organic
phase was washed with water (100ml), brine (100ml),

dried (NaZSO4), and concentrated 22 22222 to yield a yellow

liquid. Distillation gave 2.659, 92%, of 22 as a pale

58

BIBLIOGRAPHY

yellow liquid. B.P.(0.5mm)=150°C.

EI/MS (70eV):384(M+,1.2), 325(7.2), 329(30), 291(33),
235(34), 201(13), 177(95.5), 121(80), 81(62.39),
41(base)

lHNMR (250MHz)é’:7.24(m,1H), 7.18(m,1H), 6.23(s,1H), 2.8-

2.4(m,2H), 2.0-0.7(m,29H)

IR(neat): 3000, 2960, 2890, 1500, 1470, 1380, 1175, 1070,

1035, 880, 780cm“1

59

1)

2)

3)

4)

5)

6)

7)

8)a)
b)
C)

d)

9)

10)

ll)

12)a)

b)

BIBLIOGRAPHY

Nakanishi, K., 22 22, eds., "Natural Products Chemistry,"
Kodansha Ltd., Tokyo, 1974.

Kubo, I.; Lee, Y.W.; Balogh-Nair, V.: Nakanishi, K;
Chapya, A., J. Chem. Soc. Chem. Comm., 1972, 949.

Ferrari, M.; Pelizzoni, F.; Ferrari, G., Phytochemistry,
1974, 22, 208.

 

Appel, H.H.; Connolly, J.D.; Overton, K.H.; (in part)
Bond, R.P.M., J. Chem. Soc., 1969, 4685.

Kupchan, S.M.; Hemingway, R.J.; Werner, D.; Karim, A.;
McPhail, J.T.; Sim, G.A., J. Amer. Chem. Soc., 1968,
22, 3596.

 

Kupchan, S.M.; Eakin, M.A.; Thomas, A.M., J. Med. Chem.,

 

1977, 14, 1147.
rvevx -——

Cook, C.E.; Whichard, L.P.; Turner, B.; Wall, M.E.;
Egley, G.H., Science, 1966, 154, 1189.

Gianturco, M.A.: Friedel, P., Can J. Chem., 1966,

22, 1083; "CV‘
Turner, J.A.; Herz, W.; J. Org. Chem., 2912, 22, 1900;
McMurry, J.E.; Donovan, S.F., Tetrahedron Lett.,

1917, 2869.

Inomata, K.: Aoyama, S.; Kotake, H.; Bull. Chem. Soc.

Japan, 1978, 22, 930.

 

 

 

 

Tanis, S.P., Tetrahedron Lett., in press.

Akita, H.; Naito, T.; Oishi, T., Chem. Pharm Bull.,
1980, 28, 2166.

W'—

Schulte, G.; Scheuer, P.J.; McConnel, O.J., Helv. Chim.
Acta., 1980, 22, 2159.

 

 

Johnson, W.S.; Harbert, C.A.; Ratcliffe, B.E.
Stipanovic, R.D., J. Amer. Chem. Soc., 1976,
6188; "
Brot, F.E.: Johnson, W.S.; Ratcliffe, B.E.; Stelling,
G.D., Bioorg. Chem., 1977, 2, 257;

98,

 

 

6O

13)

14)

15)

16)

17)

18)

19)

20)

21)

22)
23)

24)

25)

26)

C)
d)

e)

f)
9)

h)

Johnson, W.S., Acc. Chem. Res.,/22~2, 2, 1;

Goldsmith, D.J.; Phillips, C.F., J. Amer. Chem. Soc., 1969,
22, 5862;
vanTamelen, E.E.; Milne, G.M.; Suffness, M.I.:
RuderEhavven, M.C.; Anderson, R.J.; Achini, S.,

J. Amer. Chem. Soc. 197Q, 22, 7202;

Goldsmith, D.J., J. Amer. Chem. Soc., lgggJ 22, 3913;
vanTamelen, F.E.: Willet, J.; Schwartz, M.; Nadeau, R.,
J. Amer. Chem. Soc., 1966, 22, 5937;

vanTamelen, E.E.; James, .R., J. Amer. Chem. Soc.,
1977, 99, 950.

AM—

 

 

 

 

 

Boekman Jr., R.K.; Bruza, K. J.; Heinrich, G.R.,
J. Amer. Chem. Soc., 1978, 100, 7101.

 

Baldwin, J.E.; Thomas, R.C.; Kruse, L.I.; Silberman,
L.; J. Org. Chem., 1977, 22, 3846.

 

Boyd, M.P.; Harris, T.M.; Wilson, B.J., 2ynthesis,
1971, 545.
W

 

Collington, E.W.; Meyers, A.I., J. Org. Chem., 1971,
22, 3044. '

 

Watson, S.L.; Eastham, J.F., J. Organomet. Chem., 1967,
9, 165.

 

Tamura, M.; Kochi, J., Synthesis, 1971, 303.

 

Sharpless, K.B.; Michelson, R.C., J. Amer. Chem. Soc.,
2973, 22, 6138.

Payne, G.B.: Williams, P.H., J. Org. Chem., 1961, 26,

651. " " ’—

 

 

Johnson, C.R.; Dutra, G.A., J. Amer. Chem. Soc., 1973,
22, 7777.

 

Still, W.C., J. Amer. Chem. SOC., 1978, 100, 1481.

 

Marino, J.P.: Abe, H., 2ynthesis, 1920, 872.

 

 

Sherman, B.; Amstutz, E.D., J. Amer. Chem. Soc., 1950,
22, 2195.

Snider, B.B.; Rodini, D.J.; Karras, M.; Kirk, T.C.;
Deutsch, E.A.; Cordova, R.: Price, R.T., Tetrahedron,
1981, 22, 3927.

 

Duza, J.P.; Joseph, J.P.; Bernsetin, S., J. Amer. Chem.
Soc., 2922" 22, 3908 and Cohen, N., Acc. Chem. Res.,
1976, 2, 412.

 

61

27)

28)

29)

30)

31)a)

b)

C)

32)

33)

34)

35)

36)

37)

38)

Rust and Spialter, U.S. Patent Number 2,709,174.

Feld, R.; Cowe, P.L., "The Organic Chemistry of Titanium"
Butterworth Inc., Washington, D.C., 1965.

Reetz, M.T.; Huttenhain, S.; Hubner, F., gyn. Comm.,
1981, 22, 217.

 

Tanis, S.P., unpublished work; for previous cyclization
of epoxy-dendrolasin see Nasipuri, D; Das, G., J. Chem.
Soc. Perkin I, 1979, 2776.

 

Snider, B.B.; Rodini, D.J.; vanStraten, J., J. Amer.
Chem. Soc., 22224 222, 5872;

Sutherland, J.K., Chem. Soc. Rev., 1 80, 2, 265,
Hashimoto, S; Itoh, A.; Kitagawa, ., Yamamoto, H.:
Nozaki, H., J. Amer. Chem. Soc., 2321: 22, 4192.

 

 

 

Still, W.C.; Mitra, A.; Khan, M., J. Org. Chem., 2222!
22, 2923.

 

Mori, K.: Kobayashi, S.; Matsui, M., Agric. Biol. Chem.,
22fifiL, 39, 1889.

Moppett, C.E.; Sutherland, J.K., J. Chem. Soc. C, 1968,
3040.

 

 

Rogan, J.B., J. Org. Chem., 1962, 22, 3910.

 

McCormick, J.P.; Barton, D.L., J. Chem. Soc. Chem. Comm.,
1975, 303.

 

 

Braude, E.A.; Evans, E.A., J. Chem. Soc., 1955, 3324.

Burka, L.T.; Felice, L.J.; Jackson, S.W., Phytochem.,
1981, 22, 647.

 

62