SYNTHESIS AND CHEMISTRY OF (TRIALKYLSILYL)VINYLKETENES

SYNTHESIS AND CHEMISTRY OF (TRIALKYLSILYL)VINYLKETENES

byDAWN MARIE BENNETT

B. A., ChemistryBoston University, 1994

SUBMITTED TO THE DEPARTMENT OF CHEMISTRY IN PARTIALFULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

AT THE

MASSACHUSETTS INSTITUTE OF TECHNOLOGY

June 1999

© 1999 Massachusetts Institute of TechnologyAll Rights Reserved

Signature of Author................................ -.... .... .............

Department of ChemistryMay 19, 1999

C ertified by ............................................. ......... ........Rick L. DanheiserThesis Supervisor

A ccepted by............................ ........ ...............................Dietmar Seyferth

Department Committee on Graduate Studies

7%9;;" TUTE

M.I.T. I VWV

I LIBRARIES

This doctoral thesis has been examined by a committee of the Department of

Chemistry as follows:

Professor Stephen L. Buchwald.............. ....... ... ....

Chair

Professor Rick L. Danheiser........ ............. .......----

Thesis Supervisor

Professor Gregory C. Fu ............................ ---..........--

2

Acknowledgments

I would like to thank my family, friends, and colleagues for their unwavering supportthroughout my graduate school career. First and foremost, I must express my gratitude to myadvisor, Professor Rick Danheiser, for providing me with the opportunity to learn in hislaboratory. I promise to always remember that "criticism makes you stronger!" I am alsoindebted to my undergraduate advisor, Professor Warren Giering of Boston University, forinspiring me to pursue my love of chemistry at the highest possible level.

I wish to thank the National Science Foundation for a predoctoral fellowship from1994-1997. Also, thanks are due to Jeff Simpson and Mark Wall of the InstrumentationFacility (Spec Lab!!) as well as to Kevin Shea for their patience with me on the NMRmachines over the years.

I consider myself fortunate to have had such excellent co-workers over the past fiveyears and I would like to thank all Danheiser group members past and present for theircontributions to my intellectual growth. I would especially like to thank the current groupmembers for their support while I've been writing this thesis and for not teasing me too muchabout my Diet Coke addiction.

I would like to thank both Jason Diffendal and Matt Martin for helping me to sort outall kinds on computer snafus and for being such good sports about all the Star Wars jokes.Christophe Mellon has provided valuable discussions about chemistry (and life) and I wishhim the best in Cantata! Kevin Shea has helped me numerous times with NMR experimentsand has been a great resource for discussing ketene chemistry and Italian restaurants. I wouldlike to thank Greg Dudley for his last-minute help with MacSpartan calculations even thoughit meant he had to suffer through missing a day of group clean-up (Keanu Reeves rules!). Ialso wish to thank Dana Buske, Chris, Greg, and that blue guy for making my last night ofthesis writing so memorable! Best wishes to Ralf Demuth, Yoshi, Rob Niger, and ouryoungest group member, Dave Amos (TABOO!).

I would like to express my appreciation to Jason, Kevin, Matt, Greg, and Jeremy forproofreading this dissertation on extremely short notice. Special thanks are due to Greg whograciously accepted proofreading duties for the longest chapter plus all the associatedexperimentals! Most of all, I would like to thank Rick for reading each chapter in minutedetail and providing useful writing tips at the same time.

Thank you to my dear friend Elizabeth for all your sane advice on life and love whenthe world seemed upside down. I consider myself incredibly fortunate to have you as myfriend. Mint chocolate chip always!

To the members of the women's group (you know who you are), thank you from thebottom of my heart for helping me to recognize my own strengths. I will carry a part of eachof you with me forever.

Rob Lea, I wish you were still on this earth to share this happy moment with me. Ithink of you often and know you have found peace in heaven.

I have been truly blessed with a wonderful family. This thesis would have beenimpossible without the support of my Mom, Dad, and brother, Craig. Of course the recentaddition to my "family", Wile E., has helped immensely, too! I love you all dearly and can'tfind appropriate words to express my sincere gratitude for all of your kindness and support.

3

To my family

4

SYNTHESIS AND CHEMISTRY OF (TRIALKYLSILYL)VINYLKETENES

byDawn M. Bennett

Submitted to the Department of Chemistryon May 19, 1999 in partial fulfillment of the

requirements for the Degree of Doctor of Philosophy

ABSTRACT

The properties of silylketenes differ dramatically from those of ketenes. New

strategies for the synthesis of (trialkylsilyl)vinylketenes ("TAS-vinylketenes") were explored.

These robust vinylketenes undergo highly regioselective [4+2] cycloadditions in which the

ketene carbonyl dominates in controlling the regiochemical course of the reaction. TAS-

Vinylketenes have been shown to undergo Diels-Alder reactions with activated olefinic and

azcetylenic dienophiles. These vinylketenes also participate in a hetero [4+2] version of the

reaction in which carbonyl and imino dienophiles react with TAS-vinylketenes to afford

substituted (6)-lactones and lactams. TAS-Vinylketenes also react in a [4+1] annulation to

afford cyclopentenones.

Thesis supervisor: Rick L. Danheiser

Title: Professor of Chemistry

5

Table of Contents

Part I

Synthesis and Chemistry of (Trialkylsilyl)vinylketenes

Chapter 1

Introduction and Background: Vinylketenes and (Trialkylsilyl)ketenes

Chapter 2

Synthetic Approaches to Substituted TAS-Vinylketenes

Chapter 3

Applications of TAS-Vinylketenes as Dienes in the Diels-AlderReaction: Synthesis of Carbocyclic Compounds

Chapter 4

Applications of TAS-Vinylketenes as Dienes in the Hetero Diels-AlderReaction: Synthesis of Heterocyclic Compounds

Chapter 5

TAS-Vinylketenes as Four-Carbon Components in a New [4 + 1]Annulation Strategy

Part II

Experimental Section

General Procedures

Materials

Chromatography

Instrumentation

Experiments and Spectra

6

8

34

53

68

100

113

114

114

114

115

116

PART I

Synthesis and Chemistry of (Trialkylsilyl)vinylketenes

7

CHAPTER 1

INTRODUCTION AND BACKGROUND: VINYLKETENES AND

(TRIALKYLSILYL)KETENES

Introduction

The utility of ketenes in organic synthesis is well-established.' Studies in the

laboratory of Staudinger2 nearly one hundred years ago led to the first general methods for the

synthesis of alkyl- and arylketenes. Staudinger also examined the participation of ketenes in

addition reactions with nucleophiles as well as their dimerization and polymerization

reactions. Since his pioneering work, ketene chemistry has flourished and now includes

studies in theoretical, mechanistic, and synthetic areas.

Vinylketenes (a,P-unsaturated alkenylketenes) were believed to be intermediates in

chemical reactions as early as 1941 . Although these ketenes are difficult to isolate, a variety

of methods exist for generating them in situ for use in synthetic reactions. Vinylketenes have

the capacity to function as versatile four-carbon building blocks for the assembly of a variety

of carbocyclic systems. [2+2] Cycloadditions of vinylketenes are especially valuable

transformations which can serve as triggering steps for pericyclic cascades leading to the

formation of six- and eight-membered carbocyclic compounds (vide infra).

My research has focused on the chemistry of (trialkylsilylvinylketenes ("TAS-

vinylketenes"). We expected that these compounds would be more stable than vinylketenes,

allowing them to express their underlying reactivity as electron-rich conjugated dienes and

participate as four-carbon components in Diels-Alder and related cycloadditions. As

background to our work, this chapter includes a review of the synthesis and reactions of

vinylketenes, followed by a discussion of silylketenes, including their properties, common

1. For reviews of ketene chemistry, see: (a) Tidwell, T. T. Ketenes; Wiley: New York, 1991. (b) The

Chemistry ofKetenes, Allenes, and Related Compounds; Patai, S., Ed.; John Wiley and Sons: New

York, 1980 and references cited therein. (c) Hyatt, J. A.; Raynolds, P. W. Org. React. 1994, 45, 159.(d) Schaumann, E.; Scheiblich, S. In Methoden der organischen Chemie (Houben Weyl); Kropf, H.,Schaumann, E., Eds.; George Thieme: Stuttgart, 1993; Vol. E15/3, pp 2818-2880, 2933-2957.

2. (a) Staudinger, H. Chem. Ber. 1905, 38, 1735. (b) Staudinger, H.; Ott, E. Chem. Ber. 1908, 41,2208. (c) Staudinger, H.; Anthes, E.; Schneider, H. Chem. Ber. 1913, 46, 3539.

3. Smith, L. I.; Hoehn, H. H. J. Am. Chem. Soc. 1941, 63, 1181.

8

methods of preparation, and utility in organic synthesis. The remaining chapters will detail

our investigations of the synthesis and reactions of TAS-vinylketenes.

Generation of Vinylketenes

Three synthetically significant methods for the generation of vinylketenes have been

reported to date: (1) the dehydrohalogenation of a,p-unsaturated acid chlorides, (2) the

electrocyclic ring opening of cyclobutenones, and (3) the photochemical Wolff rearrangement

of a,p-unsaturated diazo ketones. In addition, several other methods have been used to

prepare particular vinylketenes. Most vinylketenes cannot be isolated due to their tendency to

dimerize and polymerize, and therefore are usually generated in situ and immediately reacted

with a ketenophile.

Dehydrohalogenation of acid chlorides has been used as a method for the generation

of vinylketenes since 1966 when Payne first discovered that reaction of 3-methyl-2-butenoyl

chloride with trimethylamine forms isopropenylketene. 4 Although this ketene itself is

unstable, it can be trapped by reaction with ethyl vinyl ether or isolated as the ketene dimer.

In subsequent years, a variety of vinylketenes have been prepared via the

dehydrohalogenation method5 including the isolable vinylketene 2 which Wuest found could

be purified by distillation (24 *C, 0.01 mm Hg, eq 1).5k

00 C11 Et3N 1

160 C, 8h

benzene

sealed tube

1 62% 2

4. Payne, G. B. J. Org. Chem. 1966, 31, 718.5. For examples of the generation of vinylketenes via dehydrohalogenation, see: (a) Gelin, R.; Gelin, S.;

Dohnazon, R. Tetrahedron Lett. 1970,3657. (b) Hickmott, P. W.; Miles, G. J.; Sheppard, G.; Urbani,

R.; Yoxall, C. T. J. Chem. Soc., Perkin Trans. 1 1973, 1514. (c) Rey, M.; Roberts, S.; Dieffenbacher,

A.; Dreiding, A. S. Helv. Chim. Acta 1970, 53,417. (d) Rey, M.; Dunkelblum, E.; Allain, R.;Dreiding, A. S. Helv. Chim. Acta 1970, 53,2159. (e) Dondoni, A. Heterocycles 1980, 1547. (f)Wuest, J. D. Tetrahedron 1980, 36,2291. (g) Huston, R.; Rey, M.; Dreiding, A. S. Helv. Chim. Acta

1982, 65,451. (h) Jackson, D. A.; Rey, M.; Dreiding, A. S. HeIv. Chim. Acta 1983, 66,2330. (i)Danheiser, R. L.; Martinez-Davila, C.; Sard, H. Tetrahedron 1981, 37, 3943. (j) Danheiser, R. L.;

Gee, S. K.; Sard, H. J. Am. Chem. Soc. 1982, 104, 7670. (k) Wuest, J. D.; Madonik, A. M.; Gordon,

D. C. J. Org. Chem. 1977, 42, 2111.

9

One major drawback of the dehydrohalogenation route to vinylketenes is the

production of an amine hydrochloride byproduct which is known to catalyze the

polymerization of ketenes. 6 In addition, this salt can cause isomerization of the 0,y-double

bonds of 2-vinylcyclobutanoneS 5h (products of [2+2] reactions of vinylketenes with olefins)

into conjugation with the carbonyl group.

The electrocyclic ring opening of cyclobutenones is a second general method for the

synthesis of vinylketenes. This process can occur under either thermal7 or photochemical8

conditions. In 1956, Jenny and Roberts discovered that heating optically active 2,4-dichloro-

3-phenylcyclobutenone results in racemization (eq 2).9 They proposed that this reaction

involves reversible formation of chloroketenes 4 via electrocyclic ring opening of the

cyclobutenone. In the presence of ethanol the ketene can be trapped as the ethyl ester 5,

providing evidence in support of a vinylketene intermediate.

C1 0 CHC 3 C C,0 C 0

' 1X100 C EtOH H OEt (2)Ph CI Ph CI / Ph

H H

3 4 5

Photochemical conditions can also effect the electrocyclic cleavage of cyclobutenones.

For example, methylprenylketene is generated via the photochemically induced electrocyclic

6. For examples, see: (a) Brady, W. T.; Waters, 0. H. J. Org. Chem. 1967, 32, 3703. (b) Hansford, W.E.; Sauer, J. C. Org. React. 1946, 3, 108.

7. For examples of thermal electrocyclic ring opening to generate vinylketenes, see: (a) Bellus, D.; Ernst,B. Angew. Chem., Intl. Ed Engl. 1988, 27, 797. (b) Wong, H. N. C.; Lau, K.-L.; Tam, K.-F. Top.Curr. Chem. 1986, 133, 83. (c) Moore, H. W.; Yerxa, B. R. In Advances in Strain in OrganicChemistry, Halton, B., Ed.; Jai Press: London, 1995, pp 81-162. (d) Marvell, E. N. In ThermalElectrocyclic Reactions, Academic Press: New York, 1980, pp 124-190. (e) Moore, H. W.; Decker,0. H. W. Chem. Rev. 1986, 86, 821.

8. For examples of photochemical electrocyclic ring opening to generate vinylketenes, see: (a) Barton, D.H. R. Helv. Chim. Acta 1959, 42, 2604. (b) Baldwin, J. E.; McDaniel, M. C. J. Am. Chem. Soc.1968, 90, 6118. (c) Chapman, 0. L.; Lassila, J. D. J. Am. Chem. Soc. 1968, 90, 2449. (d) Arnold, D.R.; Hedaya, E.; Merritt, V. Y.; Karnischky, L. A.; Kent, M. E. Tetrahedron Lett. 1972, 3917. (e)Huisgen, R.; Mayr, H. J. Chem. Soc., Chem. Commun. 1976, 55. (f) Mayr, H.; Huisgen, R. J. Chem.Soc., Chem. Commun. 1976, 57.

9. (a) Jenny, E. F.; Roberts, J. D. J. Am. Chem. Soc. 1956, 78, 2005. (b) Silversmith, E. F.; Kitahara, Y.;Roberts, J. D. J. Am. Chem. Soc. 1958, 80,4088. (c) Silversmith, E. F.; Kitahara, Y.; Caserio, M. C.;Roberts, J. D. J. Am. Chem. Soc. 1958, 80, 5840.

10

ring opening of 2,4,4-trimethylcyclobutenone. This vinylketene then undergoes [2+2]

cycloaddition with ethyl vinyl ether to afford a mixture of stereoisomeric cyclobutenones (eq

3 ).8f One advantage of this route is that vinylketenes can be generated at room temperature.

In our laboratory, Kollol Pal studied the generation of vinylketenes by irradiation of

cyclobutenones and in fact was able to observe the formation of several vinylketenes in good

yield by 'H NMR spectroscopy. 0

0 hv ro D O0 +0c tEto

(3)

6 7 8 9

72: 28

The advantages of the electrocylic ring opening for the generation of vinylketenes are

clear. This mild, efficient method produces no amine salt that can cause undesirable side

reactions. Also, because this is an intramolecular process requiring no reagents, the

vinylketene can be generated in very low concentration thus reducing the rate of dimerization.

A third method for the generation of ketenes, the photochemical Wolff rearrangement

of a,p-unsaturated diazo ketones, is useful for two reasons: the reaction can be run at room

temperature, and the only byproduct is nitrogen." In 1964, Roedig reported that irradiation of

diazo ketone 10 results in formation of vinylketene intermediate 11 which can be trapped by

reaction with N-(phenyl)benzaldimine to afford P-lactam 12 (eq 4).12 We have investigated

the generation of vinyl- and arylketene intermediates via the photochemical Wolff

rearrangement of a-diazo ketones in the context of the aromatic annulation strategy developed

in our laboratory.' 3

10. Pal, K., Ph. D. Thesis, Massachusetts Institute of Technology, Cambridge, MA, 1987, pp. 69-77.11. For reviews of the photo Wolff rearrangement, see: (a) Regitz, M.; Maas, G. Diazo Compounds -

Properties and Synthesis; Academic Press: New York, 1986. (b) Doyle, M. P.; McKervey, M. A.; Ye,T. Modern Catalytic Methods for Organic Synthesis with Diazo Compounds; Wiley: New York,1998. (c) Zollinger, H. Diazo Chemistry II: Aliphatic, Inorganic, and Organometallic Compounds;VCH: New York, 1995.

12. Roedig, A.; Fahr, E.; Aman, H. Chem. Ber. 1964, 97, 77.13. Danheiser, R. L.; Brisbois, R. G.; Kowalcyzk, J. J.; Miller, R. F. J. Am. Chem. Soc. 1990, 112, 3093.

11

0 N2 hv M0C C! h O

A CO 2Me benzene MeO 2C CN PhN MoO2C NPh

Cl ' CI -N2 CI1 CI Ph 28% CCI Ph

CI CI CI

10 11 12

Another method for the generation of vinylketenes involves the photolysis or

thermolysis of alkenylcarbene metal complexes.14 In this process, metal-ketene complexation

suppresses undesired side reactions such as dimerization. These metal-ketene complexes can

undergo nucleophilic addition, cycloaddition with alkenes, and intramolecular cyclization.

For example, reaction of cobalt complex 13 and 3-hexyne at 25 *C results in the formation of

cobalt-vinylketene complex 14;1' addition of sodium methoxide then yields ester 15 in 54%

yield (eq 5).

Ph Et --=-- Et Et C>Co(CO)SnPh3 Ph 3Sn(OC)2Co-

MeO benzene rt Ph

13 35% 14 OMe

NaOMeMeOH (5)

54%

0Et OMe

15 Et) PhOMe

In summary, three synthetically significant methods for the generation ofvinylketenes

have found use in recent years: (1). the dehydrohalogenation of acid chlorides, (2) the

electrocyclic ring opening of cyclobutenones, and (3) the photo Wolff rearrangement of

14. For reviews on alkenylcarbene metal complexes, see: (a) Hegedus, L. S. In Advances in Metal

Carbene Chemistry; Schubert, U., Ed.; Klumer Academic Publishers: Boston, 1989, pp 199-210, 233-246. (b) D~tz, K. H. New J. Chem. 1990, 14, 433.

15. Wulff, W. D.; Gilbertson, S. R.; Springer, J. P. J Am. Chem. Soc. 1986, 108, 520.

12

unsaturated a-diazo ketones. Other pyrolytic and thermolytic methods also can be used for

the preparation of specific ketenes. 16

Reactions of Vinylketenes

Vinylketenes participate in a variety of reactions including nucleophilic addition and

[2+2] and [4+2] cycloadditions. A brief summary of these reactions follows.

Nucleophilic addition of water, alcohols, amines, and carboxylic acids to vinylketenes

affords acids, esters, amides, and anhydrides. Nucleophiles trap vinylketenes very efficiently,

and this process is often cited as evidence that a ketene was generated in a reaction. For

example, in our laboratory Kollol Pal confirmed the intermediacy of a vinylketene generated

by the electrocyclic ring opening of 2,3-dimethylcyclobutenone by carrying out the reaction in

the presence of t-butylamine to afford amides 17 and 18 (eq 6).

o hv 00t-BuNH2 Nt-Bu + Nt-Bu (6)

100% A616 17 18

Vinylketenes have found their greatest utility reacting with alkenes and alkynes in

[2+2] cycloadditions leading to 2-vinylcyclobutanones and -cyclobutenones.!'" In

particular, our research group has extensively employed these cycloadditions as triggering

steps for pericyclic cascades that lead to the formation of 3-cyclohexenols, cyclooctadienones,

and highly substituted phenolic compounds (Scheme 1). In these reactions, vinylketenes

function as electron-deficient 2n components in the key [2+2] cycloaddition step.

16. See ref la, pp 52-149.17. For examples, see refs 5dh,ij, 8f, and 13.18. Danheiser, R. L.; Nishida, A.; Savariar, S.; Trova, M. P. Tetrahedron Lett. 1988, 4917.

13

Scheme 1

Rz REREO_ H' RE OH

D Rz [1,3) Sigmatropic ERearrangement Rz

D D

0 1 [3,31C"0 Sigmatropic

b 1,3-Dienes Rearrangement

Acetylenes R 1 0O 1) 4e~ Electrocyclic OHAceyleesCleavage R'

1 2 2) 69- ElectrocyclicR---- R R2 12)eClosure R2

3) Tautomerization

Scheme 1 outlines three strategies for the synthesis of carbocyclic compounds

developed in our laboratory based on vinylketenes. The first strategy (pathway a) provides a

method for the synthesis of substituted 3-cyclohexenols.5 i In this annulation, generation of

the vinylketene by dehydrohalogenation of an a,p-unsaturated acid chloride followed by a

[2+2] cycloaddition with an electron-rich alkene affords a 2-vinylcyclobutanone intermediate.

Addition of a nucleophile with subsequent alkoxy-accelerated [1,3] sigmatropic

rearrangement then yields a 3-cyclohexenol.

Pathway b illustrates a [4+4] strategy for the synthesis of eight-membered

carbocycles. 5i In this method, generation of the vinylketene occurs by either 1,4-

dehydrohalogenation of an acid chloride or the thermal electrocyclic ring opening of a

cyclobutenone. [2+2] Cycloaddition of the vinylketene with a 1,3-diene yields a 2,3-

divinylcyclobutanone intermediate. Subsequent [3,3] sigmatropic rearrangement results in

formation of the 2,6-cyclooctadienone.

The aromatic annulation strategy described in pathway c is a powerful method for the

preparation of highly substituted phenolic compounds. 13,18 We have demonstrated the utility

of this strategy by developing direct synthetic routes to a variety of natural products, including

14

Dan Shen diterpenoid quinones,' 9 maesanin, 2 0 aegyptinones A and B, salvilenone,2 2 and

bergapten.2 Irradiation of an a-diazo ketone triggers a photochemical Wolff rearrangement

to produce a vinylketene which then reacts with an acetylene in a regiospecific [2+2]

cycloaddition. Continued irradiation or thermolysis effect the 4n electrocyclic opening of the

resulting 4-vinylcyclobutenone generating a dienylketene. The penultimate 67c

electrocyclization is followed by tautomerization to yield the aromatic product.

Although vinylketenes have a proclivity to undergo [2+2] cycloadditions, a few

examples of direct [4+2] cycloadditions of these compounds do exist,24 including cases

involving dimerization of vinylketenes. For example, reaction of two molecules of

vinylketene 19 affords the a-pyrones 20 and 21 resulting from a [4+2] cycloaddition (eq 7 ).5d

Generating vinylketenes in low concentrations in the presence of an excess of ketenophile

helps to prevent this undesired dimerization.

0 Et3N CW0 0 0

CICHC13 ''0 0 0 (C d3 is--I+ (7)rt

19 KI20 21

34% 31%

As summarized above, vinylketenes have been well-documented as intermediates in a

variety of reactions and their participation in [2+2] cycloadditions has been extensively

exploited in organic synthesis. If this normal reaction pathway could be suppressed,

vinylketenes might express their underlying reactivity as electron-rich conjugated dienes for

other useful tranformations. Based on the fact that silylketenes are remarkably more stable

19. (a) Danheiser, R. L.; Casebier, D. S.; Loebach, J. L. Tetrahedron Lett. 1992, 33, 1149. (b) Danheiser,R.L.; Casebier, D. S.; Firooznia, F. J. Org. Chem. 1995, 60, 8341.

20. Danheiser, R. L.; Cha, D. D. Tetrahedron Lett. 1990, 31, 1527.21. Danheiser, R. L.; Casebier, D. S.; Huboux, A. H. J. Org. Chem. 1994, 59, 4844.22. Danheiser, R. L.; Helgason, A. L. J. Am. Chem. Soc. 1994, 116,9471.23. Danheiser, R. L.; Trova, M. P. Synlett 1995, 573.24. For examples, see: (a) Day, A. C.; McDonald, A. N.; Anderson, B. F.; Bartczak, T. J.; Hodder, 0. J. R. J.

Chem. Soc., Chem. Commun. 1973,247. (b) Rees, C. W.; Somanathan, R.; Storr, R. C.; Woolhouse, A. D.J. Chem. Soc., Chem. Commun. 1976, 125. (c) Berge, J. M.; Rey, M.; Dreiding, A. S. Helv. Chim. Acta1982, 65,2230. (d) Collumb, D.; Doutheau, A. Tetrahedron Lett. 1997, 38, 1397.

15

that most other ketenes, we hypothesized that silylvinylketenes might exhibit the stability of

silylketenes, thus reducing their eagerness to participate in [2+2] reactions. In order to

understand the basis for this hypothesis, it is first necessary to review the synthesis and

chemistry of (trialkylsilyl)ketenes.

(Trialkylsilyl)ketenes

Silylketenes25 exhibit completely different properties from most other ketenes. The

remarkably stable parent compound (trimethylsilyl)ketene can be stored at room

temperature for several years without decomposition. In fact, this TAS-ketene can even be

purified by distillation at 82 *C. As will become apparent later in this chapter, TAS-ketenes

resist dimerization and undergo reactions at a much slower rate than their alkylketene

counterparts.

The silyl substituent has an extraordinary ability to stabilize ketenes and suppress their

natural tendency to dimerize and undergo [2+2] cycloadditions. Originally, TAS-ketenes

were believed to exist as a tautomeric mixture of (trimethylsilyl)ketene and

trimethylsilyloxyacetylene, accounting for their unusual stability. Spectroscopic studies later

showed that this was not true and that only the ketene structure is present.

Currently, the amazing stability of TAS-ketenes is believed to result from

hyperconjugative electron a-7t donation from the C-Si bond into the in-plane carbonyl 7r*-

orbital, a reasonable supposition considering that the power of the C-Si bond as a

hyperconjugative electron donor has been well documented. Also, Tidwell has proposed

that -the stability of TAS-ketenes is due to the electron-releasing ability of the electropositive

25. For a review on silylketenes, see: Pommier, A.; Kocienski, P.; Pons, J.-M. J. Chem. Soc., PerkinTrans. 1 1998, 2105.

26. Loebach, J. L.; Danheiser, R. L. In Encyclopedia ofReagentsfor Organic Synthesis; Paquette, L. A.,Ed.; Wiley: New York, 1995, pp 5266-5268.

27. For a discussion, see: (a) Patai, S. The Chemistry of Organic Silicon Compounds; Patai, S., Rappoport, Z.,Eds; Wiley: New York, 1989. (b) Weber, W. P. Silicon Reagentsfor Organic Synthesis; Springer-Verlag:Berlin, 1983. (c) Colvin, E. W. Silicon in Organic Synthesis; Butterworths: London, 1981. (d) Fleming, I.In Comprehensive Organic Synthesis; Jones, D. N., Ed.; Pergamon: New York, 1979; Vol. 3, pp 541-686.

(e) Armitage, D. A. In Comprehensive Organometallic Chemistry; Wilkinson, G., Ed.; Pergamon: New

York, 1982; Vol. 2, pp 1-203.

16

silicon group, allowing for a donation. The carbonyl group is rendered less electrophilic and,

as a result, less reactive toward nucleophilic addition reactions and cycloadditions.28

MeSi , Me3Si® _9C=O =-

H 2 1 H

A trimethylsilyl group attached to the C-2 of a ketene results in an upfield shift of the

13 C NMR resonance of C-1 as a result of partial sp character at C-1, providing some evidence

for silicon acting as a a-t donor in (trialkylsilyl)ketenes. The phenomenon of a-n donation

by silicon is well documented in other areas of silicon chemistry.

Alternatively, Runge has suggested that TAS-ketene stability is due to back-donation

from the ketene n-system to the d-orbitals of the silicon atom, resulting in a partial Si-C

double bond.29 Any major contribution by such a resonance form (see below) should result in

a longer C-C bond and shorter C-O bond. Ab initio molecular orbital calculations confirmed

that the C=O and C=C bond lengths of (trimethylsilyl)ketene are almost the same as the C=O

and C=C bonds of methylketene and therefore Runge's proposed structures probably do not

contribute significantly to TAS-ketene stability.

GMe 3 Si> M Me 3Si

>=C=0 =C= o 01 g C=H ZH

H

Generation of (Trialkylsilyl)ketenes

Silylketenes have been prepared by a wide variety of methods including the

thermolysis of alkoxyalkynes, dehydration of silylacetic acids, Wolff rearrangement of x-

diazo ketones, electrocyclic ring opening of cyclobutenediones, and reaction of lithiated

28. For a discussion, see ref Ia and the following: (a) Greg, L.; McAllister, M. A.; Tidwell, T. T. J. Am. Chem.

Soc. 1991, 113, 6021. (b) Allen, A. A.; Egle, I.; Janoschek, R.; Liu, H. W.; Ma, J.; Marra, R. M.; Tidwell,T. T. Chem. Lett. 1996, 45.

29. Runge, W. Prog. Phys. Org. Chem. 1981, 13,315.

17

silyldiazomethane with carbon monoxide. Shchykovskaya achieved the first synthesis of

(trimethylsilyl)ketene in 1965 by pyrolysis of an alkoxy(trialkylsilyl)acetylene, the most

commonly used procedure for the preparation of TAS-ketenes to date.3 0 Heating

ethoxy(trimethylsilyl)ethyne 22 at 120 *C results in liberation of ethylene, affording ketene 23

in 90% yield (eq 8). Initially, Shchukovskaya believed that the mechanism of the reaction

proceeds via an isomerization, but in 1975 Ruden reported that the reaction actually occurs

via a retro-ene 6-membered transition state.3 1

H H

120 *C H -C2H4 Me 3SiMe 3Si ---- OEt 0 ' C C=0 (8)

22 [ Me 3Si - 0 90% 23

Subsequent studies have shown that (trimethylsilyl)ketene (23) can be generated at

lower temperatures by the thermal decomposition of t-butoxy(trimethylsilyl)ethyne 24.

Heating alkyne 24 at temperatures as low as 50 *C resulted in the slow elimination of 2-

methylpropene. Nearly instantaneous conversion to (trimethylsilyl)ketene occurs at higher

temperatures (100-110 *C). The large t-butoxy group shields the triple bond and helps to

prevent side reactions (such as polymerization and nucleophile attack) that commonly occur

when the silylketene is generated from the less hindered ethoxy derivative 22. Therefore,

preparation of silylketenes from 24 can occur in the presence of nucleophiles which

subsequently react with the ketene. For example, (trimethylsilyl)ketene (generated by

pyrolysis of 24) in the presence of diphenylamine, afforded amide 25 in quantitative yield (eq

9). Unfortunately, the t-butoxy derivative 24 requires several steps to prepare whereas a one-

step procedure yields ethoxy derivative 22 from commercially available ethoxyacetylene.

30. Shchukovskaya, L. L.; Pal'chik, R. I.; Lazarev, A. N. Dokl. Akad Nauk SSSR (Engi. Transl) 1965, 164,

357.31. Ruden, R. A. J. Org. Chem. 1974, 39, 3607.32. Valenti, E.; Pericas, M. A.; Serratosa, F. J. Org. Chem. 1990, 55, 395.

18

Ph 2NHCHC13 0

50-55 *C, 11 h (Me3SI -= Ot-Bu 100 Me3Si -- NPh2 (9)100% N~ 2

24 25

Substituted (trialkylsilyl)ketenes can be prepared by a related method, the thermal

rearrangement of trialkylsilyloxyacetylenes.33 Treatment of 1-methoxy-3-methyl-i-butyne

with trimethylsilyl iodide affords intermediate 1-(trimethylsilyloxy)-3-methyl-i-butyne.

Rearrangement to the corresponding isopropyl-substituted ketene 28 occurs at room

temperature (eq 10).33a Synthesis of the parent (trimethylsilyl)ketene can occur via a similar

rearrangement in which reaction of ketene, trimethylsilyl triflate, and triethylamine affords

(trimethylsiloxy)acetylene as an intermediate which undergoes 1,3-silicon transfer at room

temperature to afford (trimethylsilyl)ketene in 63% yield. 4

i-Pr OM Me3SiI = OSiMe M 3SI >C=O (10)CH 3CN i-Pr

rt, 30 min 90%26 27 28

Lutsenko reported another pyrolytic method for the generation of (TAS)ketenes. 5

Heating (triethylsilyl)acetic anhydride affords (triethylsilyl)ketene in 80% yield (eq 11).

Condensation of two molecules of (triethylsilyl)acetic acid produces the necessary anhydride.

The requisite silylacetic acids can presumably be prepared in one step by the conversion of

acetic acid to the dianion by treatment with LDA followed by silylation with two equivalents

of the desired silyl chloride.36 Dehydrohalogenation of acid chlorides also forms TAS-

ketenes, but this method generally gives only low yields of the desired product?7

33. (a) Sakurai, H.; Shirahata, A.; Sasaki, K.; Hosomi, A. Synthesis 1979, 740. (b) Efinova, I. V.; Kazankova,M. A.; Lutsenko, I. F. J. Gen. Chem. USSR (Engi. Trans.) 1985, 55, 1465. (c) Pons, J.-M.; Kocienski, P.Tetrahedron Lett. 1989, 30, 1833.

34. Uhlig, W.; Tzschach, A. Z. Chem. 1988, 28, 409.35. Kostyuk, A. S.; Dudukina, 0. V.; Burlachenko, G. S.; Baukov, Yu. I.; Lutensko, I. F. JGen. Chem. USSR

(Engl. Transi.) 1969, 39, 441.36. Tsuge, 0.; Kanemasa, S.; Suzuki, T.; Matsuda, K. Bull. Chem. Soc. Jpn. 1986, 59, 2851.

19

0 0 A Et3Si C=O(11)Et 3SK~i,,k 0..J.SiEt3 80% H

29 30

In 1989, Olah reported an efficient preparation of silylketenes by the dehydration of

silyl acetic acids.38 Treatment of commercially available (trimethylsilyl)acetic acid 31 with

dicyclohexylcarbodiimide (DCC) in the presence of catalytic triethylamine affords

(trimethylsilyl)ketene in 63% yield (eq 12). However, all attempts in our laboratory to

accomplish the synthesis of (trimethylsilyl)ketene according to this procedure have thus far

been unsuccessful. Other (TAS)acetic acid precursors for this new approach to TAS-ketenes

can be prepared as mentioned above.

DCC

cat. Et3N

Me0Si . H C= (12)

63% H31 23

In 1985, Maas reported an extremely facile synthesis of TAS-ketenes by the Wolff

rearrangement of a-silyl-a-diazo ketones under photochemical conditions or with transition

metal catalysis. 3 9 Irradiation of a-diazo ketone 32 affords silylketene 33 in 94% yield (eq

13). Alternatively, treatment of 32 with copper triflate (4 mol% in benzene, 20 *C) results in

Wolff rearrangement, producing the desired ketene 33 in 93% yield. The requisite a-silyl-a-

diazo ketone can be prepared via silylation of an a -diazo ketone, thus allowing access to a

variety of aryl- and alkyl-substituted silylketenes.

37. Lutensko, I. F.; Baukov, Yu. I.; Kostyuk, A. S.; Savelyeva, N. I.; Krysina, V. K. J. Organomet. Chem.1969, 17, 241.

38. Olah, G. A.; Wu, A.; Farooq, 0. Synthesis 1989,568.39. (a) Maas, G.; Brackmann, R. J. Org. Chem. 1985, 50,2801. (b) Brtckmann, R.; Schneider, K.;

Maas, G. Tetrahedron 1989, 45,5517.

20

Ph XyN2Si(i-Pr)h

hvbenzenert 1.5h

94%

32

(i-Pr)3Si>-C=O

Ph

33

In recent years, Tidwell has done extensive studies on the preparation of

bis(trialkylsilyl)ketenes via the electrocyclic ring opening of cyclobutenediones under both

thermal and photochemical conditions.40 Reaction of bis(t-butyldimethyl-silyl)acetylene 34

with dichloroketene (generated from trichloroacetyl chloride with zinc-copper couple) affords

dichlorocyclobutenone 35. Hydrolysis with concentrated sulfuric acid yields the

cyclobutenedione 36 which undergoes a facile electrocyclic ring opening to produce bisketene

37 in excellent yield (Scheme 2).

Scheme 2

CC13CCOCIZn-Cu

Et 2O:DME

84%

34

A or hvCDC13

92%

t-BuMe2Si 0

9CIt-BuMe2Si C135

H2SO455 *C

44%

t-BuMe2Si 0

t-BuMe2Si 0

36

40. For recent reports on the synthesis and chemistry of bis(trialkylsilylketene), see: (a) Allen, A. D.; Liu,R.; Ma, J.; McAllister, M. A.; Tidwell, T. T.; Zhao, D.-C. Acc. Chem. Res. 1995, 28,265. (b) Allen,A. D.; Liu, R.; Ma, J.; McAllister, M. A.; Tidwell, T. T.; Zhao, D.-C. Pure Appi. Chem. 1995, 67,777. (c) Allen, A. D.; Colomvakos, J. D.; Egle, I.; Liu, R.; Ma, J.; Marra, R. M.; McAllister, M. A.;Tidwell, T. T. Can. J. Chem. 1996, 74,457. (d) Zhao, D.-C.; Tidwell, T. T. J. Am. Chem. Soc. 1992, 114,10980. (e) Zhao, D.-C.; Allen, A. D.; Tidwell, T. T. J. Am. Chem. Soc. 1993, 115, 10097.

21

(13)

t-BuMe2Si

III

t-BuMe2Si

t-BuMe 2Si XC :C SiMe 2t-Bu

37

Recently, Murai reported a remarkably efficient route to bis(silyl)- and

stannyl(silyl)ketenes by the reaction of lithiated (trimethylsilyl)diazomethane with carbon

monoxide, followed by trapping with a silyl chloride, silyl triflate, or stannyl chloride.4' Thus,

treatment of (trimethylsilyl)diazomethane with n-butyllithium followed by exposure to carbon

monoxide (1 atm) affords acyllithium intermediate 39. Extrusion of dinitrogen produces the

lithiated silylketene 40 which is trapped with triethylsilyl trifluoromethanesulfonate to

produce the bis(silyl)ketene 41 in 85% yield. Efforts to trap lithiated silylketene 40 with alkyl

halides or a proton source were unsuccessful.

Scheme 3

n-BuLi LiTHE Li COo

.0SI/=N2 TH ' ) N2 ,O 0= N2Me 3Si Me 3Si 3938 39

-N 2

00 Et3SiOTf 9C ofC

Me 3SSi SiEt3 Mo 3Si Li

41 40

Reactions of (Trialkylsilyl)ketenes

Silylketenes undergo a diverse assortment of reactions and the rich area of silylketene

chemistry has been extensively investigated. The remainder of this chapter will briefly

summarize the participation of TAS-ketenes in nucleophilic addition reactions as well as

[2+2] and [4+2] cycloadditions. Also, the formation of allenes, cyclopropanones, and

cyclobutanones from TAS-ketenes will be reviewed.

(1) Nucleophilic Addition

TAS-Ketenes readily undergo nucleophilic addition with a variety of compounds

including water, alcohols, amines, and certain carbon nucleophiles. Hydration of

41. Kai, H.; Iwamoto, K.; Chatani, N.; Murai, S. J Am. Chem. Soc. 1996, 118, 7634.

22

(trimethylsilyl)ketene under neutral conditions occurs at a significantly slower rate than the

addition of water to comparable alkylketenes. However, under acidic or basic conditions, the

rate of silylketene hydration is faster than that of alkylketenes.42 ,43 This interesting rate

difference is believed to result from silicon's unique ability to stabilize both a negative charge

a to the silicon (basic hydration conditions) and to stabilize a positive charge P to the silicon

(acidic hydration conditions).44

Other oxygen nucleophiles also react at a much faster rate with alkylketenes than with

TAS-ketenes. However, under appropriate conditions, simple alcohols do add to

(trimethylsilyl)ketene to afford esters in excellent yield.45 Alcohols also combine with the

more hindered (t-butyldimethylsilyl)ketene, although at a slightly reduced rate. TAS-Ketenes

react with sterically hindered alcohols very slowly unless a Lewis acid that can catalyze the

reaction is present. For example (eq 14), addition of n-butanol to TMS-ketene occurs rapidly

at -10 OC,30 but reaction of t-butanol with 23 is much slower (CC14 , 48 h, rt, 80%) unless

catalyzed by a Lewis acid (BF 3-OEt 2, 5 min, rt, 93%).31

t-BuOH0 BF 3-OEt n-BuOH

M e3S i -.-,A 1* rt 5 min Me 3Si C=o -10 C 0 Me3SIkOn-B. (14)Ot-Bu 93% H 85%

43 23 42

Zinc halides effectively catalyze the addition of alcohols to (trimethylsilyl)ketenes.46

Many functional groups tolerate the presence of this mild Lewis acid, including carbonyl

groups, acetals, olefins, and epoxides that would be sensitive to other Lewis acids such as

BF 3-OEt 2. Kita has thus found that zinc chloride catalyzes the addition of a-hydroxy ketones

42. Allen, A. D.; Tidwell, T. T. Tetrahedron Lett. 1991, 32, 847.43. For other studies on the hydration of silylketenes, see refs 30, 40e and the following: (a) Vodolazskaya, V.

M.; Baukov, Yu. I. J. Gen. Chem. USSR (Engl. Transl.) 1973, 43, 2076. (b) Ponomarev, S. V. Angew.Chem., Int. Ed Engl. 1973, 12,675.

44. Gong, L.; McAllister, M. A.; Tidwell, T. T. J. Am. Chem. Soc. 1991, 113, 6021.45. For examples of the reaction of silylketenes with alcohols, see refs 30, 31, 40e, 43a and the following:

(a) Ando, W.; Sekiguchi, A.; Migita, T.; Kammula, S.; Green, M.; Jones, M., Jr. J. Am. Chem. Soc.1975, 97, 3818. (b) Knauss, E. E.; Coutts, R. T.; Kazakoff, C. W. J. Chromatogr. Sci. 1976, 14, 525.(c) Ponomarev, S. V.; Erman, M. B.; Gervits, L. L. J. Gen. Chem. USSR (Engl. Transl.) 1972, 42,462.

46. Kita, Y.; Sekihachi, J.-i.; Hayashi, Y.; Da, Y.-Z. ; Yamamoto, M.; Akai, S. J. Org. Chem. 1990, 55,1108.

23

to (trimethylsilyl)ketene to afford functionalized a-silylacetates (eq 15), compounds that are

difficult to prepare by traditional methods. Yamamoto reported that a naturally occurring

lipase isolated from Rhizopusjaponicus also catalyzes this transformation.47

0 ZnC 2, CH 2C2 0)> c=0 + Ph)Y OH rt 45 mn Ph < O SiMe3 (15)

H Me 91% Me O

23 44 45

As mentioned earlier, TAS-ketenes react rapidly with amines to form amides.48

Ruden reported the quantitative conversion of (trimethylsilyl)ketene to the corresponding

amide 46 upon addition of diisopropylamine at room temperature (eq 16).31 Bromine also

adds to (trimethylsilyl)ketene to produce bromo(trimethylsilyl)acetylbromide. 30 In addition,

alkoxy(tributyl)tins combine with TAS-ketenes to afford a-stannyl esters.49

i-Pr2NHMeS C= Me3Si P (16)

H 100% M SA NiP)

23 46

Taylor reported a convenient one-pot preparation of coumarins based on the addition

of salicylaldehyde salts to TAS-ketenes.50 Condensation of o-acyl phenol 47 with ketene 23

affords a-silylcarboxylate 48, and subsequent cyclization and elimination yields coumarin 49

in 92% yield (eq 17).

47. Yamamoto, Y.; Ozasa, N.; Sawada, S. Chem. Express 1993, 8, 305.48. For examples of the reaction of silylketenes with amines, see refs 30, 31, and the following: Coutts, R.

T.; Jones, G. R.; Benderly, A.; Mak, A. L. C. J. Chromatogr. Sci. 1979, 17, 350.49. (a) Akai, S.; Tsuzuki, Y.; Matsuda, S.; Kitagaki, S.; Kita, Y. J. Chem. Soc., Perkin Trans. 1 1992,

2813. (b) Akai, S.; Tsuzuki, Y.; Matsuda, S.; Kitagaki, S.; Kita, Y. Synlett 1991, 911.50. Taylor, R. T.; Cassell, R. A. Synthesis 1982, 672.

24

Na O1

CHO Me 3Si H

47 23

DMF 0rt 2h N O

92% CHO SiMe 3

48 49

The addition of carbon nucleophiles to TAS-ketenes has also received considerable

attention. Organolithium reagents add to substituted TAS-ketenes; subsequent addition of

acid results in formation of a ketone, whereas treatment with a trialkylsilyl chloride affords a

silyl enol ether. For example, treatment of bis(trimethylsilyl)ketene with n-butyllithium

followed by acidic workup affords ketone 51.51 Alternatively, addition of n-butyllithium to

ketene 52 followed by trapping with trimethylchlorosilane produces the vinyl silyl ether 53 as

a single stereoisomer (eq 18).

1. n-BuLi, THFMe 3Si >C=O rt 8h NoMe 3Si 2. H30+

50 85%

1. n-BuLi, Et 20-78 *C 1h

2. Me 3SiCI-78 *C to rt 12 h

78%

0Me 3Si Y'knn-Bu

SiMe3

51

Me3Si OSiMe3

Et n-Bu

53

When a TAS-ketene containing a proton at C2 is treated with strongly basic

organolithium reagents, only proton abstraction results (no nucleophilic addition is

observed).53 Kita reported that the resulting lithiated silylketene can react with a

(trialkylsilyl)chloride at low temperature to afford bis(trialkylsilyl)ketenes in excellent yield.54

In the presence of a less basic organocerium reagent, nucleophilic addition to TAS-ketenes

51 . Woodbury, R. P.; Long, N. R.; Rathke, M. W. J. Org. Chem. 1978, 43,376.52. Baigrie, L. M.; Seiklay, H. R.; Tidwell, T. T. J. Am. Chem. Soc. 1985, 107, 5391.53. See ref 51 and the following: Sullivan, D. F.; Woodbury, R. P.; Rathke, M. W. J. Org. Chem. 1977,

42, 2038.

25

(17)

Me 3Si

Et

52

(18)

occurs, forming an enolate. 5 For example, addition of BuCeC 2 to (t-

butyldimethylsilyl)ketene affords enolate 55 which can be trapped with either aqueous

ammonium chloride or an alkyl halide to afford 56 and 57 (eq 19).

NH 4CI 0t-BuMe

2Si Bu_[ 091%t-BuMe2Si BuCeC 2 t-BuMe2Si 0 561C=0 N >=( (19)

H H Bu

54 55 Mel t-BuMe2SIy 0BHMPA Y,

Me76%

57

Enamines can react as ketenophiles, combining with silylketenes to form 1,3-diones. 56

For example, Shioiri reported that addition of enamine 58 to (t-butyldimethylsilyl)ketene 54 at

room temperature followed by hydrolysis of the imine during silica gel chromatograpy affords

dione 59 in 70% yield (eq 20).

1. PhH 0 0t-BuMe 2Si N rt, 18 h Sit-BuMe2H= + (5~ ~ IJ~(20)

H 2. Si02

70%54 58 59

(2) Cycloadditions of TAS-Ketenes

Because of the stabilizing effect of silicon substituents on ketenes, TAS-ketenes

undergo [2+2] cycloadditions only with very reactive, electron-rich olefms. In 1974, Zaitseva

reported the addition of 1,1 -diethoxyethene to (trimethylsilyl)ketene to afford the [2+2]

cycloadduct.57 Brady later reported that (trimethylsilyl)ketene reacts with tetramethoxyethene

54. Akai, S.; Kitagaki, S.; Naka, T.; Yamamoto, K.; Tsuzuki, Y.; Matsumoto, K.; Kita, Y. J. Chem. Soc.,Perkin Trans. 1 1996, 1705.

55. Kita, Y.; Matsuda, S.; Kitagaki, S.; Tsuzuki, Y.; Akai, S. Synlett 1991, 401.56. Takaoka, K.; Aoyama, T.; Shioiri, T. Synlett 1994, 1005.57. Zaitseva, G. S.; Baukov, Yu. I.; Mal'tsev, V. V.; Lutsenko, I. F. J. Gen. Chem. USSR (Engi. Transl.) 1974,

44, 1389.

26

to form cyclobutanone 61.58 Treatment of ketene 23 with the bulkier alkene 62 results only in

addition to the ketene (without subsequent ring closure) to yield acyclic ester 63 (eq 21).' 9

MoO OMe60

MoO OMeMe3Si

90*C 2h a MeI OMe

65% MeO OMe

61Me3Si (21)

H MoO OSiMo3 62 M0 3Si OSiMe323 H OSiMO3 H' CO 2SMe3

rt 24 h MeO

95% 63

In 1975, Zaitseva reported the first [2+2] cycloaddition of a TAS-ketene and an

aldehyde. Reaction of (trimethylsilyl)ketene with benzaldehyde in the presence of BF 3-OEt2

affords a 2:1 mixture of cis- and trans-4-phenyl-3-trimethylsilyloxetan-2-one (64) in 65%

yield (eq 22).60 Due to the decreased reactivity of TAS-ketenes in [2+2] cycloadditions, these

reactions usually require catalysis by Lewis acids. Since Zaitseva's discovery, this [2+2]

approach has become a common method for the synthesis of P-lactones. 61 Recent

developments in the area of oxetanone synthesis include the use of bulky Lewis acids for

preparation of exclusively cis-substituted P-lactones62 and chiral Lewis acids for the

formation of enantiomerically enriched products.63 Silylketenes do not participate in [2+2]

reactions with ketones.

58. Brady, W. T.; Saidi, K. J Org. Chem. 1980, 45, 727.59. Brady, W. T.; Saidi, K. J. Org. Chem. 1990, 55, 4215.60. Zaitseva, G. S.; Vinokurova, N. G.; Baukov, Yu. I. J. Gen. Chem. USSR (Engl. Transl.) 1975, 45, 1372.61. For a review on the synthesis of P-lactones via the [2+2] reaction of TAS-ketenes and carbonyl

compounds, see ref 25 and references cited therein.62. Maruoka, K.; Concepcion, A. B.; Yamamoto, H. Synlett 1992, 31.63. For a recent example, see: Dymock, B.; Pons, J.-M.; Kocienski, P. J. Chem. Soc., Chem. Commun. 1996,

1053.

27

PhCHOcat. BF 3-OEt

65%

Me 3Si 0

Ph

64

Although the well-known Staudinger reaction of ketenes and imines is an extremely

useful method for the synthesis of P-lactams, few examples of silylketene participation in this

type of reaction have been reported.64 In 1976, Brady reported the first [2+2] reaction of a

TAS-ketene with an imine. Reaction of bromo(trimethylsilyl)ketene 66 (generated in situ by

dehydrohalogenation of 65) with N-(t-butyl)benzaldimine provided P-lactam 67 in 56% yield

(eq 23).64a

N.,t-BuEt3N Me3S

Pheptane 3Br C= Ph H

reflux 16 h -56%

66

t-Bu 0'N

Ph! SiMe3Br

67

Schubert later reported the reaction of ethoxy(triphenylsilyl)ketene 69 (generated in

situ from metal complex 68) with N-(methyl)benzaldimine which produces a 74:16 mixture of

cis and trans-substituted P-lactams 70 and 71 (eq 24).64' At about the same time, Zaitseva

reported that (trimethylsilyl)ketene reacts with N-alkylsulfonyl chloralimines.64c

64. (a) Brady, W. T.; Owens, R. A. Tetrahedron Lett. 1976, 1553. (b) Kron, J.; Schubert, U. J. Organomet.Chem. 1989, 373, 203. (c) Novikova, 0. P.; Livantsova, L. I.; Zaitseva, G. S. J. Gen. Chem. USSR (Engl.Transi.) 1989, 59, 2350. (d) Zaitseva, G. S.; Novikova, 0. P.; Livantsova, L. I.; Petrosyan, V. S.; Baukov,Yu. I. J. Gen. Chem. USSR (Engl. Transi.) 1991, 61, 1268.

28

Me 3Si >

H

23

(22)

MOSi -)ABr

Br

65

(23)

(CO) 5MC(OEt)SiPh3

M = Cr, Mo, W

68

Et2O

-M(CO) 6

Me, 0N

P h /SiMe 3H OEt

7074%

Ph 3Si>=

EtO69IPh

)=N-MeH

Me 0N

PhXf I/EtH SiMe3

7116%

Usually TAS-ketenes do not participate in formal [4+2] cycloaddition reactions, but a

few examples have been reported. Shioiri recently described the reaction of (t-

butyldimethylsily)ketene with an electron-rich 1,3-diene. 65 Experimental evidence was

obtained suggesting that this reaction proceeds via a stepwise mechanism. Thus, addition of

diene 72 to ketene 54 affords a betaine intermediate (73) that can be trapped as the ester 74 by

addition of water. Increasing the time or temperature of the reaction results in cyclization to

afford the dihydropyranone 75 which subsequently isomerizes to pyranone 76 (Scheme 4).

Scheme 4

t-BuMe2Si OSiMO3C=0 +

H Oh.~.IIEO3

54 72

benzenereflux 3 d

a OSiMe3t-BuMe2SI OSiMe

73

'-Q COSiMe3

74

t-BuMe2 Si I An0 0

76

t-BuMe 2Si 0 077%

75

65. Ito, T.; Aoyama, T.; Shioiri, T. Tetrahedron Lett. 1993, 34, 6583.

29

(24)

In 1996, Aoyama and Shioiri reported another example of a TAS-ketene undergoing a

[4+2] cycloaddition.66 Reaction of (trimethylsilyl)ketene with acylisocyanate 77 affords the

[4+2] adduct, 78. Subsequent reaction with dimethylacetylene dicarboxylate and expulsion of

carbon dioxide then yields 2-pyridone 79 in nearly quantitative yield in an extremely efficient

one-pot procedure (eq 25).

0

R N=C=0 +

77 IMO 3SI

H R = (p-N0 2)C6 H4

23

o-dichlorobenzenereflux 3 h

0

[OMe 3SiO N R

78

.IDMAD3h

99%

CO 2MeCO2Me

O N RH

79

(3) Formation of Allenes, Cyclopropanones, and Cyclobutanones

Olefmation of (trimethylsilyl)ketene with stabilized phosphorus ylides has been used

for the preparation of allenes.67 For example, reaction of ylide 80 with (trimethylsilyl)ketene

at -5 'C affords allenic ester 81 in 85% yield (eq 26). Reaction of (trimethylsilyl)ketene with

unstabilized phosphorus ylides results only in formation of complex mixtures of products.

Combination of a variety of phosphorus ylides (such as Ph3P=C(Ph)(Et)) with

bis(trialkylsilyl)ketenes gives reasonable yields of silylallenes.

66. (a) Takaoka, K.; Aoyama, T.; Shioiri, T. Tetrahedron Lett. 1996, 37, 4973. (b) Takaoka, K.; Aoyama, T.;Shioiri, T. Tetrahedron Lett. 1996, 37, 4977.

67. See ref 25 and Kita, Y.; Tsuzuki, Y.; Kitagaki, S.; Akai, S. Chem. Pharm. Bull. 1994, 42, 233.

30

(25)

H

MeSPh 3 P C2Et Me 3Si HHO iC=O 80 HC== (26)H H CO 2Et

CH 2CI2, -5 *C23 85% 81

Silyl-substituted ketenimines can also be prepared using organophosphorous

chemistry. Barbaro and co-workers described the reaction of (trimethylsilyl)ketene with

[bis(trimethylsilyl)methylimino]triphenylphosphorane 82 at room temperature, affording

ketenimine 83 in moderate yield (eq 27).68 Another aza-Wittig-type reaction serves as a key

step in Molina's total synthesis of aaptamine in which treatment of a functionalized

iminophosphorane with (trimethylsilyl)ketene affords a silylketenimine intermediate.69

,CH(SiMe 3)2

Me 3Si Ph3P=N 82 Me 3Si (2)CH(SiMe)

H Et2O, 25*C H

23 42% 83

Zaitseva found that reaction of TAS-ketenes with diazomethane produces

cyclopropanones or cyclobutanones.70 The outcome of the reaction depends greatly on the

amount of diazomethane used. Treatment of (trimethylsilyl)ketene with one equivalent of

diazomethane at 130 *C results in exclusive formation of 2-(trimethylsilyl)-cyclopropanone

84. The cyclopropanone product can react with a second equivalent of diazomethane to

afford the ring expansion products 85 and 86 as a mixture of isomers. The cyclobutanones

can be obtained directly by treatment of 23 with two equivalents of diazomethane at -78 *C

(Scheme 5).

68. Barbaro, G.; Battagliz, A.; Giogianni, P.; Guerrini, A.; Seconi, G. J. Org. Chem. 1995, 60, 6032.69. Molina, P.; Vida, A.; Varquero, I. Synthesis 1996, 1199.70. (a) Zaitseva, G. S.; Bogdanova, G. S.; Baukov, Yu. I.; Lutsenko, I. F. J. Organomet. Chem. 1976, 121, Cl-

C22. (b) Zaitseva, G. S.; Bogdanova, G. S.; Baukov, Yu. I.; Lutsenko, I. F. J. Gen. Chem. USSR (Engl.Trans.) 1978, 48, 111.

31

1 equiv CH 2N2 M0 3Si-130 C

50%

Et 2

2 equiv CH 2N2 Me 3SiEt2 , -78 C

90% 85

84

quiv CH 2N2), -78 *C

0

Me 3Si 86

40: 60

(Trimethylsilyl)diazomethane also combines with TAS-ketenes in [2+1]

cycloadditions.7 1 Brady reported that (diethylmethylsilyl)ketene 87 reacts smoothly with

TMS-diazomethane to afford exclusively the cis-substituted cyclopropanone 88 (eq 2 8 ).',a

Reaction of TMS-diazomethane with (trimethylsilyl)ketene and other

(trimethylsilyl)arylketenes have also been reported.7 lb" Concurrent with our work (see

Chapter 5), Tidwell reported the [4+1] reaction of bis(trialkylsilylketenes) with TMS-

diazomethane. This reaction will be discussed in detail in Chapter 5.

Et 2MeSi

H

Me 3SiCHN 2Et2 O

-10 *C to rt

90%87

Et 2MeSi > 0Me3Si

88

(28)

Our laboratory reported the first synthesis and isolation of a silylvinylketene several

years ago. Since then, we have been interested in developing new routes to substituted TAS-

vinylketenes and systematically investigating their utility in organic synthesis. Chapter 2

71. (a) Brady, W. T.; Cheng, T. C. J. Organomet. Chem. 1977, 137,287. (b) Zaitseva, G. S.; Kisim, A.N.; Fedorenko, E. N.; Nosova, V. M.; Livantsova, L. I.; Baukov, Yu. I.; Lorberth, J. J. Gen. Chem.

USSR (Engl. Transi.) 1987, 57, 1836. (c) Zaitseva, G. S.; Lutsenko, I. F.; Kisin, A. V.; Baukov, Yu. I.;Lorberth, J. J. Organomet. Chem. 1988, 345, 253.

72. Colomvakos, J. D.; Egle, I.; Ma, J.; Pole, D. L.; Tidwell, T. T.; Warkentin, J. J. Org. Chem. 1996, 61,9522.

32

Scheme 5

Me 3Si

H

23

O

discusses new approaches for the synthesis TAS-vinylketenes, and Chapters 3-5 present our

results on TAS-vinylketene reactivity in [4+2] and [4+1] cycloadditions.

33

CHAPTER 2

SYNTHETIC APPROACHES TO SUBSTITUTED TAS-VINYLKETENES

Synthesis of (Trimethylsilyl)vinylketene by Dehydrohalogenation

One of the most widely used methods for the preparation of TAS-vinylketenes

involves the dehydrohalogenation of acid chlorides. Indeed, this approach has been used for

the synthesis of vinylketenes as discussed in Chapter 1. Several years ago our laboratory

reported the first synthesis of (trimethylsilyl)vinylketene (92) by this method, according to the

route outlined in Scheme 6.7 Treatment of 1-(trimethylsilyl)propyne (89) with

diisobutylaluminum hydride and methyllithium affords a vinylalanate intermediate75 that

subsequently reacts with carbon dioxide to form (Z)-2-(trimethylsilyl)-2-butenoic acid (90).

The potassium salt of 90 reacts with oxalyl chloride to afford acid chloride 91 that is used in

the next step without purification. Addition of 91 to a pentane solution of triethylamine

followed by heating at reflux (15 to 24 h) yields (trimethylsilyl)ketene. Purification by

distillation (25 *C, 1 mmHg) affords the ketene (92) as a yellow-green liquid (Scheme 6).

(Trimethylsilyl)vinylketene is moderately stable and can be stored in solution at 0 0C for 1-2

weeks without appreciable decomposition.

Scheme 6

1. DIBAL-MeLi 0

- SiMe3 Et20-hexane O Me3SI OH2. C0 2

89 90

KOH, MeOH1.1 equiv (COC) 2pentane, cat. DMF

0 *C to rt, 1.5 h

0.9 equiv Et 3NMeSiNC, pentane, reflux Me Si

C 15-24 h M3S C11

39-50%92 overall yield 91

73. Danheiser, R. L.; Sard, H. J. Org. Chem. 1980, 45, 4810.74. (a) Eisch, J. J.; Damasevitz, G. A. J. Org. Chem. 1976, 41, 2214. (b) Uchida, K.; Utimoto, K.; Nozaki, H.

J. Org. Chem. 1976, 41, 2215.75. Zweifel, G.; Steele, R. B. J. Am. Chem. Soc. 1967,89,2754.

34

Unfortunately, the limited thermal stability of this vinylketene restricts the scope of its

reactions. We expected that substituted TAS-vinylketenes would be more stable.

Unfortunately, although dehydrohalogenation of the aP-unsaturated chloride 91 provides

convenient access to (trimethylsilyl)vinylketene, this strategy is not well-suited for the

synthesis of more highly substituted TAS-ketenes. In some cases the requisite precursors are

not available, and regiochemical ambiguities can arise in dehydrohalogenation (1,4-

elimination) reactions involving p,p-disubstituted-a,p-unsaturated carboxylic acid derivatives

(eq 29).

R3Si C R3Si C" R3Si C"C00_ + (29)

R R2 R R2 R1 R2

Recent studies in our laboratory have demonstrated that the photochemical Wolff

rearrangement provides an efficient method for the synthesis of vinylketenes.13 In these

transformations, the vinylketene is not isolated, but is trapped in situ with an alkyne, initiating

a cascade of pericyclic reactions leading ultimately to the formation of a highly substituted

aromatic system. Maas and co-workers have shown that saturated TAS-ketenes can be

generated in a similar fashion by the Wolff rearrangement of a-silyl-a-diazo ketones.39 The

efficiency of these processes suggested that the rearrangement of a'-silyl-a'-diazo-a,P-

unsaturated ketones (94) might provide the basis for a new route to TAS-vinylketenes (eq 30).

1 R3SiOTf R1 0 N RSi C R3SiR N2 i-Pr2EtN 2 hv R1 (30)

R 3 R

93 94 95

35

Synthesis of Substituted TAS-Vinylketenes by Photochemical Wolff Rearrangement

The requisite a-diazo ketone precursors (93) were all prepared using the

"detrifluoroacetylative" diazo transfer strategy previously developed in our laboratory.76

Thus, treatment of methyl ketone 96 with 1.1 equivalents of lithium hexamethyldisilazide in

THF at -78 *C produces a lithium enolate, which is acylated by exposure to 1.2 equivalents of

trifluoroethyl trifluoroacetate at -78 *C. Treatment of the resulting a-trifluoroacetyl ketone

(97) with 1.5 equiv of MsN3 in the presence of 1.0 equivalents of water and 1.5 equivalents of

Et3N in acetonitrile (25 *C, 4 h) affords the desired diazo ketone 98 in 75-79% yield after

chromatographic purification (eq 31). It should be noted that a'-diazo-a,P-unsaturated

ketones generally cannot be prepared by addition of diazomethane to acyl chlorides because

of competing side reactions involving dipolar cycloadditions to the conjugated double bond."

0 .1 L0 0 1.5 equiv MsN31. 1.1 equiv LiHMDS 1.0 equiv H20 N

THF, -78 C CF3 1.5 equiv Et 3N N2

2. 1.2 equiv TFETFA CH 3CN, rt 4 h-78 *C, 5-10 min

75-79%96 97 98

Silylation of the diazo ketones was accomplished using a modification of the method

of Mass and co-workers. 78 In this procedure, diazo ketones are treated with a silyl triflate

reagent and diisopropylethylamine to afford the a-silyl-a-diazo ketone. As reported by Maas,

the silylated a-diazo ketones have a tendency to undergo protodesilylation during their

preparation and isolation caused by the presence of acidic trialkylammonium triflate

byproducts. Jennifer Loebach of our laboratory found that this problem can be minimized by

employing a solvent system consisting of equal parts Et2O and hexane. 9 The ammonium

76. (a) Danheiser, R. L.; Miller, R. F.; Brisbois, R. G.; Park, S. Z. J. Org. Chem. 1990, 55, 1959. (b)Danheiser, R. L.; Miller, R. F.; Brisbois, R. G. Organic Syntheses; Wiley: New York; 1998; Collect. Vol.

IX, 197.77. For a discussion and examples, see: (a) Regitz, M.; Maas, G. Diazo Compounds, Properties and Synthesis;

Academic Press: New York, 1986, pp 498-499. (b) Fink, J.; Regitz, M. Synthesis 1985, 569. (c) Itoh, M.;

Sugihara, A. Chem. Pharm. Bull. 1969, 17, 2105. (d) Regitz, M.; Menz, F.; Liedhegener, A. Justus

Liebigs Ann. Chem. 1970, 739, 174. (e) Harmon, R. E.; Sood, V. K.; Gupta, S. K. Synthesis 1974, 577. (f)Rosenquist, N. R.; Chapman, 0. L. J. Org. Chem. 1976, 41, 3326 and references therein.

78. See ref 39 and the following: Bruckmann, R.; Maas, G. Chem. Ber. 1987, 120,635.79. (a) Loebach, J. L., Ph. D. Thesis, Massachusetts Institute of Technology, Cambridge, MA 1995. (b)

Loebach, J. L.; Bennett, D. B.; Danheiser, R. L. J. Org. Chem. 1998, 63, 8380.

36

salts have reduced solubility in this medium, and the desired silylation products are obtained

in 10-30% higher yield as compared to reaction in Et2O alone. A variety of a'-trialkylsilyl-

c'-diazo-a,P-enones can be prepared in good yield by employing this protocol (Scheme 7).

Diazo ketones 100-102 were isolated as yellow to orange oils that can be stored in solution at

0 *C for several days without significant decomposition. Because a wider range of silyl

chloride reagents are available as compared to silyl triflates, we briefly examined the

suitability of silyl chlorides as substitutes for triflates in the silylation reaction. As a model

case, the reaction of chlorotriethylsilane with diazo ketone 98 was examined under the

standard silylation conditions (1.0 equiv i-Pr2EtN, Et2O:hexane, 0 *C to rt). Unfortunately,

this reaction afforded none of the desired product even after stirring for several days at room

temperature. Apparently, silyl chlorides are insufficiently reactive to effect the desired

transformation.

Scheme 7

0 R3SiOTf, i-Pr2EtN 0

RI N2 Et2 -hexane R0_0C toi

_t1

99 R',R2= (CH 2)4 R3= (i-Pr)3 100 75%

98 RI = R2 = CH 3 R3 =(i-Pr)3 101 86-89%

98 Ri= R2 =CH3 R3 = Et3 102 70-84%

Recently we have been interested in preparing TAS-vinylketenes with alkoxy- or

phenyl-substituted silyl groups. We anticipated that these particular silyl groups would permit

useful synthetic transformations following cycloaddition. The silyl triflates necessary for the

synthesis of these desirable TAS-vinylketenes were not readily available. However, these

highly reactive and extremely moisture sensitive compounds have been prepared previously

by treatment of the requisite silyl chloride with triflic acid80 or by reaction of the silyl chloride

with silver triflate.8 1 Alternatively, treatment of an aryl- or allylsilane with triflic acid results

80. Corey, E. J.; Cho, H.; Racker, C.; Hua, D. H. Tetrahedron Lett. 1981, 22, 3455.

81. Riediker, M.; Graf, W. Helv. Chim. Acta 1979, 62, 205.

37

in displacement of the phenyl or allyl group, affording a silyl triflate product.82 We first

prepared an aryl-substituted silyl triflate using the strategy outlined in eq 32. Reaction of

dichlorodiphenylsilane with 2.4 equivalents of ethylmagnesium bromide afforded the desired

silane 103 in excellent yield,83 and addition of an equimolar amount of triflic acid then

generated the desired silyl triflate 104 and benzene, which can be removed by distillation. The

crude triflate reagent (104) was used for silylation reactions without further purification.

2.4 equiv EtMgBr

Ph 2SiC12 Et 2O, reflux 4 h - Ph2M2 1.0 equiv TfOH PhEt2SIOTf (32)

87% 103 -PhH 104

Reaction of diazo ketone 98 with approximately one equivalent of the unpurified

diethyl(phenyl)silyl trifluoromethanesulfonate 104 under the standard silylation conditions

affords the desired silyl diazo ketone 105 in 56-74% yield (eq 33, not optimized).

PhEt 2SiOTf, i-Pr2EtNN2 E20-hexane N

0 C to rt, 1-2 h 10N 2 (33)SIEt2Ph

56-74%98 105

Unfortunately, synthesis of the alkoxy-substituted silyl triflate 107 was not

straightforward (Scheme 8). Dichlorodiethylsilane, when treated with isopropanol in

refluxing benzene, affords the alkoxysilyl chloride 106 in 75% yield after distillation.84

(Chloride 106, as expected, was not sufficiently reactive to effect silylation of diazo ketone

98). Reaction of chloride 106 with silver triflate afforded a mixture of products that could not

be separated by distillation. By a second route, dichlorodiethylsilane reacts with two

equivalents of silver triflate to afford the ditriflate 107 in 61% yield after distillation.

82. (a) Matyjaszewski, K.; Chen, Y. L. J. Organomet. Chem. 1988, 340, 7. (b) Uhligh, W. J. Organomet.Chem. 1993, 452, 29.

83. Kipping, C. J. Chem. Soc. 1927, 107.84. McCusker, P. A.; Greene, C. E. J. Am. Chem. Soc. 1948, 70,2807.

38

Unfortunately, subsequent treatment with isopropanol and base afforded a mixture of products

that decomposed upon attempted distillation.

Scheme 8

1.0 equiv i-PrOHbenzene, reflux 1.0 equiv AgOTf

5 min CH2C 2, rt 2 hEt2Si(Oi-Pr)CI

75% 106 A OTfEM2=1~2 Et2S 1 -\ 2.0 equiv AgOTf 1.0 equiv i-PrOH Et 2 Oi-Pr

CH 2C12 Et2SI(OTf)2 Et3 N 107

61% 108

An alternate route to 107 which was not investigated experimentally is outlined in eq

34. Treatment of dichlorodiethylsilane with an equimolar amount of isopropanol followed by

addition of allylmagnesium bromide was expected to afford intermediate 109, and subsequent

displacement of the allyl group with triflic acid was then anticipated to produce the desired

compound 107.

1) i-PrOH .O(I-Pr) 1.0 equiv TfOHEt 2 SiC2 - - - - - - - - - - - - - - Et2S - 107 (34)

2) . MgBr109

With efficient routes to a variety of a-trialkylsilyl-L-diazo ketones in hand, we were

able to carry out a systematic investigation of their conversion to the desired TAS-

vinylketenes. Initial studies by Loebach found that Wolff rearrangement occurred upon

irradiation at 300 nm using a Rayonet RPR-100 photochemical reactor at 30-35 *C (Scheme

9). Better yields were obtained in the synthesis of more highly substituted vinylketenes,

possibly because of the greater stability of the diazo ketone precursors.

39

Scheme 9

0*0hv (300 nm) R3Si C*

R1 N2 benzene imp-R2) SiR3 R

12R

100 R ,R2= (CH 2)4, R3= (i-Pr)3 110 70-74%101 R1 = R2 = CH 3, R3= (i-Pr)3 111 74-87%102 Ri = R2 = CH 3, R3 = Et3 112 59%

105 R1 = R2 = CH 3, R3 = Et 2 Ph 113 64%

TAS-Vinylketenes exhibit a number of interesting spectral characteristics. The

following page outlines the spectral features of the diethyl(phenyl)silyl-substituted ketene 113

(Figure 1). The IR spectrum of TAS-vinylketene 113 shows the expected strong diagnostic

stretch near 2100 cm-1 resulting from the symmetric stretching modes of the ketene backbone

(C=C=O). The 'H NMR spectrum is relatively uninformative because TAS-vinylketenes do

not contain a proton on the ketene double bond that would be diagnostic of the structure.

However, the 13C NMR spectrum has two notable features. As expected, the C-1 (C=C=0)

carbons of TAS-vinylketenes are extensively deshielded and give a low field signal near 185

ppm. The C-2 carbon (C=C=O) exhibits an unusually high field signal near 20 ppm, resulting

from contribution of the resonance structure illustrated below.

40

0 R3SR 3 Si C ______G

C

Figure 1. Spectral Characteristics of

(E)-2-(1-Methyl-1-propenyl)-2-(diethyl(phenyl)silyl)ketene (113)

IR (film): 3360, 2440, and 2060

'H NMR (500 MHz, CDCl3):

7.53-7.57 (m, 2H) 0.90-1.04 (m)

7.32-7.39 (m, 3H) Ph,SI C0

1.80 (s) H3C H 4.96

CH 3 1.51

(q, J= 6.8 Hz)

(d, J= 6.8 Hz)

I'If t

i 6 7 6 5 4 3

13C NMR (125 MHz, CDCl3):

226 200 180 i6 i48 126 1600 80 6 40 20 pPM

41

Li2 1 I ppe

7.9

136.2,135.1, Ph 4.8 o130.1, 128.6 S25.0 184.8

1123.8

19.2 H3C 119.3

CH 3

14.6

I .

cm~I

Maas previously reported the application of the transition-metal catalyzed Wolff

rearrangement for the preparation of saturated alkyl(TAS)ketenes, 39 and we hoped that this

method might be applicable for the synthesis of TAS-vinylketenes. Treatment of silylated a-

diazo ketone 100 with a 0.04 to 0.2 equivalents of CuOTf results in formation of the desired

ketene 110 (eq 35). Unfortunately, the yield of this reaction (22-26%) compares poorly with

the excellent yield (89%) obtained by the photochemically induced Wolff rearrangement of

the same diazo ketone, and this method was not pursued further.

cat. CuOTf (i-Pr)3Si C0

benzeneI(i-Pr)3 rt, 1-2.5 h (35)

22-26%100 110

Substituted TAS-vinylketenes are remarkably robust compounds, in dramatic contrast

to typical vinylketenes. The triisopropyl derivative (111) was recovered unchanged after

heating in C6D6 at 80 *C for four days, although some decomposition of the

(triethylsilyl)vinylketene 112 was observed after heating at this temperature for 10 h. Also

notable is the observation that these ketenes can be purified by conventional silica gel

chromatography without any detectable decomposition.

We were interested in preparing 3-alkoxy-substituted TAS-vinylketenes because these

electron-rich compounds are expected to exhibit increased reactivity in cycloaddition

reactions. As discussed earlier, TAS-vinylketenes are easily prepared by the photochemically

induced Wolff rearrangement of a'-trialkylsilyl-a'-diazo-a,p-enones. We expected that

application of this protocol to a-alkoxy-a'-trialkylsilyl-a'-diazo-a,p-enones might provide a

route to the desired alkoxy-substituted vinylketenes.

Dr. Yutaka Ukaji of our laboratory investigated this approach as outlined in Scheme

10. Lithiation of 3,4-dihydro-2H-pyran (114) with n-butyllithium followed by treatment with

N,N-dimethylacetamide afforded the aP-unsaturated methyl ketone 115 as previously

reported by Boeckman.85 Application of our standard diazo transfer conditions provided

85. Boeckman, R. K.; Bruza, K. J. Tetrahedron 1981, 37, 3997.

42

diazo ketone 116 in moderate yield. Subsequent silylation with triisopropylsilyl

trifluoromethanesulfonate in the presence of diisopropylethylamine provided silyl diazo

ketone 117 in 58% yield after purification by silica gel chromatography. Unfortunately,

irradiation of 117 in a degassed benzene solution yielded none of the desired ketene 118. It is

possible that alkoxy-substituted TAS-vinylketene 118 is an exceptionally sensitive compound

and may have formed under the photochemical conditions and then decomposed or

polymerized.

Scheme 10

1) n-BuLi, THF0 to 50 C

2) H3CCONMe 2 1

32-49%

0

115

R3SI C"

118

1) LiHMDS, THF-78 *C thenTFETFA

2) MsN 3H20, Et 3NCH 3CN, rt

46%

hv (300 nm)benzene

00 N2

116

(i-Pr)3SiOTfi-Pr2EtN

Et 2O:hexane0 0C to rt

58%

0

0 N2

SiR3

117

Synthesis of Substituted TAS-Vinylketenes by Electrocylic Cleavage of Cyclobutenones

In our previous studies on the application of vinylketenes in annulation routes to six-s8

and eight-5imembered rings, our laboratory has shown that the electrocyclic ring opening of

cyclobutenones provides an especially attractive method for the generation of these reactive

species. It therefore appeared likely that 2-trialkylsilylcyclobutenones might afford TAS-

vinylketenes in good yield upon heating. Advantages of this method include the fact that no

byproducts are formed and the reaction is reversible. In fact, Tidwell has recently shown that

86. See refs 5i, 13, and the following: Danheiser, R. L.; Gee, S. K. J. Org. Chem. 1984, 49, 1672.

43

0

114

V

A

silylated bisketenes can be generated in a similar fashion by the thermal or photochemical

ring opening of cyclobutenediones (see Scheme 2).

Our plan for the preparation of cyclobutenone precursors to TAS-vinylketenes called

for the [2+2] cycloaddition of dichloroketene with an activated alkyne followed by reductive

dechlorination. It is important to note that in order to obtain the requisite 2-silyl substituted

cyclobutenone, the alkyne must contain a donating substituent (such as an alkoxy or phenyl

group) to direct the regiochemical course of the reaction. In the case of

(trialkylsilyl)acetylenes with only alkyl substituents, mixtures of regioisomeric cycloadducts

are generally obtained. Earlier, in attempts to prepare (trimethylsilyl)vinylketene via this

method, Sard discovered that the [2+2] cycloaddition of (trimethylsilyl)acetylene with

dichloroketene affords almost exclusively the undesired P-silylcyclobutenone 120 (eq 36).87

Me3Si CC13COCI Me 3Si 0 0Zn(Cu)

' i Et2 O, reflux /Ci .J /CI

H9 C Me 3Si Cl

67% 119 120

1:99

Scheme 11 outlines the application of this strategy to the synthesis of TAS-

vinylketene 127. Addition of dichloroketene to the phenyl-substituted

(trialkylsilyl)acetylenes 121 and 122 provides the desired dichlorocyclobutenones 123 and

124 in high yield. The structure of compound 124 (a single regioisomer as determined by 300

MHz 'H NMR analysis) was assigned based on the known powerful directing effect of phenyl

groups in the [2+2] cycloadditions of ketenes and also by knowledge of the 13C NMR

spectrum of the previously characterized cyclobutenone 123 (Figure 2).86

87. Danheiser, R. L.; Sard, H. Tetrahedron Lett. 1983, 24, 23.

44

R3Si

IIIPh

121 (R= Me)122 (R = Et)

R3Si X o

Ph M

127 (R = Me)

CC13COCIZn(Cu)

Et2O, 35 c

benzene60 *C, 4.5 h

37%

R3Si 0

Ph iPh C1

123 (R = Me) 86-95%124 (R = Et) 92%

ZnAcOH-TMEDA

EtOH0 to 25 0C

R3Si 0

Ph

125 (R = Me) 64-74%126 (R = Et) 55%

150.7

Me 3Si 10182.9

180.,5

Ph /1/Cl91.2

123

149.8

EtASi1 183.4

PhC191.0

124

Dechlorination of 4,4-dichlorocyclobutenones does not proceed smoothly under

conventional conditions, but can be accomplished in good yield by employing the protocol

developed independently in our laboratory88 and that of Dreiding. 89 Upon heating in benzene

at 60 'C for 4.5 h, 125 undergoes smooth conversion to the desired TAS-vinylketene 127. As

expected, this (trimethylsilyl)vinylketene proved less stable to chromatography than the

triethyl- and triisopropylsilyl derivatives 110-112 and could only be obtained in 37% yield

after purification on Florisil. For this reason, it ultimately proved advantageous to generate

the ketenes from cyclobutenones 125 and 126 in situ for various cycloadditions rather than to

subject these sensitive compound to isolation and purification. Chapters 3 and 4 discuss the

"in situ generation" of TAS-vinylketenes from cyclobutenones 125 and 126 in detail.

88. (a) Danheiser, R. L.; Savariar, S. Tetrahedron Lett. 1987, 28, 3299. (b) Danheiser, R. L.; Savariar, S.; Cha,D. D. Organic Syntheses; Wiley: New York, 1993; Collect. Vol. VIII, pp 82-86.

89. Ammann, A. A.; Rey, M.; Dreiding, A. S. Helv. Chim. Acta 1987, 70, 321.

45

Scheme 11

Figure 2

Interestingly, in several reactions where cyclobutenone 125 was heated at elevated

temperatures (110-150 *C), a butenolide side product was isolated. We believe that this

byproduct results from reaction of TAS-vinylketene 127 with diradical oxygen. A related

reaction of oxygen with a bisketene has been previously reported by Tidwell,90 and a similar

mechanism may be operative in our examples. Scheme 12 outlines the mechanism we

propose for formation of butenolide byproduct 129. Addition of oxygen to ketene 127

followed by reaction with a second ketene molecule affords diradical intermediate 128. After

intersystem crossing, 128 can undergo cyclization to afford two molecules of butenolide 129.

After bubbling air through a refluxing toluene solution of cyclobutenone 125 for 18 h, the

butenolide (129) was isolated in 42% yield following purification by silica gel

chromatography.

Scheme 12

Me 3Si 0 toluene, reflux 002 48 h MeSI 3

be 3 0

Ph 42% 5125 Ph129

Me 3Si C0

Ph: 127

Ph02

0 . Me 3Si 0 Ph

M0 3Si M0 3Si C

t PhSiePh Ph 128 0

A variety of spectral data was collected and analyzed in order to substantiate the

structural assignment of 129. The 1H NMR spectrum exhibits a singlet at 4.94 ppm

corresponding to the two C-5 protons on the butenolide ring. The IR spectrum of 129 shows a

strong carbonyl stretching frequency at 1745 cm-1 indicative of the a,p-unsaturated lactone.

Further proof for the structure of 129 was obtained from a 13C NMR DEPT experiment. The

90. Zhao, D.-C.; Tidwell, T. T. J. Am. Chem. Soc. 1992, 114, 10980.

46

C-5 carbon appears as a methylene group at a characteristic resonance of 73.7 ppm. Also, the

carbonyl resonance is observed at 176.9 ppm and the two butenolide alkenyl carbons appear

at 174.0 ppm (C-4) and 133.4 ppm (C-3). The C-4 carbon is deshielded by conjugation with

the carbonyl group and therefore appears downfield in the 13C NMR spectrum.

We have shown that TAS-vinylketenes form easily by the electrocyclic ring opening

of 2-silylcyclobutenones and we therefore believed it would be possible to access alkoxy-

substituted TAS-vinylketenes (131) by this method (Scheme 13).76 One factor motivating our

interest in alkoxy-substituted TAS-cyclobutenones is the potential utility of these compounds

as synthetic intermediates. Addition of R to the carbonyl of 130 followed by hydrolysis

should provide an avenue to a wide array of substituted cyclobutanones (132) that could then

undergo electrocyclic ring opening to TAS-vinylketenes (133). Three methods were explored

for the formation of the requisite 3-alkoxy-2-(trialkylsilyl)cyclobutenones.

Scheme 13

A or hv R3Si C

R3Si 0 RO13113

RO 1)R + R3Si 0 R3Si C130 2) H30 A or hv R

R132 133

The first method we investigated involved the [2+2] cycloaddition of dichloroketene

with an alkoxysilane (eq 37). (Trimethylsilyl)ethoxyacetylene (22)' underwent smooth [2+2]

cycloaddition with dichloroketene (generated via dehalogenation of trichloroacetyl chloride)

to afford the desired cyclobutenone 134 in excellent yield. The next step in our scheme

required the reductive dechlorination of a 4,4-dichlorocyclobutenone, which is considerably

more difficult than in the case of the corresponding saturated derivatives, though it can be

accomplished using procedures developed independently in our laboratory and that of

Dreiding (vide supra). Unfortunately, the reductive dechlorination of 134 was unsuccessful

91. Ruden, R. A. J. Org. Chem. 1974, 39, 3607.

47

using the standard conditions and we were forced to abandon this approach. We suspect that

zinc coordination to the neighboring ethoxy group may prevent the desired reductive

dechlorination from occurring.

Me 3Si CC13COCI Me 3Si o Zn, EtOH Me3S 0

11 Zn(Cu) -I Ac0H-TMEDA4, (37Et 2O, reflux /CI

EtO EtO C1 EtO86%

22 134 135

Zaitseva has reported that reaction of TAS-ketenes with ketene acetals affords 3,3-

dialkoxy-2-(trialkylsilyl)cyclobutanones in 48-49% yield.57 Our second strategy for the

synthesis of 3-alkoxy-2-(trialkylsilyl)cyclobutenones was based on this precedent. We

anticipated that cyclobutanone 140 produced by the Zaitseva reaction might undergo

elimination in the presence of ZnCl2 to give the desired 2-silylcylobutenone 141 (Scheme 14).

We hoped that the regiochemistry of this elimination reaction could be controlled by

employing the procedure of Scheeren which allowed him to regioselectively deprotonate

cyclobutanone 136 at the more hindered position. Scheeren found that situ trapping with

(trimethylsilyl)chloride affords silyl enol ether 137, while without ZnCl2, the alternative

isomer 138 is isolated (eq 38). In our case, in the absence of a trapping agent, we expected to

observe elimination of ethoxide to provide the desired cyclobutenone 141 upon treatment of

140 with Et3N-ZnCl2.

Scheme 14

EtASi>C=0

H H 30 80 O*A _Et3Si 0 RA3 I 0

+ -(--Ir ------ ON-EtO 3> EtO EtO )

tEtOEtO139 140 141

48

OSiMe3 0 ZnC12 OSiMe3Et3N, Me3SCf EtN, Me 3SiCIrt 25 h rt 25 h (38)

Et Et Et4 IiEt EtO Et90% 90%

138 136 137

Zaitseva reported the preparation of our target intermediate (140) in 48% yield via the

reaction of diethylketene acetal (139, synthesized by dehydrobromination of

bromoacetaldehyde diethylacetal in t-BuOK/t-BuOH)92 and (triethylsilyl)ketene without

solvent at 80 *C. Unfortunately, in our hands, attempts to repeat Zaitseva's procedure

afforded none of the desired product 140. The reaction was attempted using freshly distilled

reagents, and a variety of solvents, temperatures, and reaction times were investigated, but in

no case could [2+2] cycloadduct be detected by 'H NMR. We have been unable to account

for these results.

Our third method investigated the generation of alkoxy-substituted TAS-vinylketenes

from squaric acid derivatives (Scheme 15). Liebeskind has extensively studied the chemistry

of squaric acid and its derivatives, reporting that addition of organolithium reagents to

diisopropylsquarate followed by treatment with catalytic amounts of acid is the best available

procedure for the preparation of squaric acid derivatives. 93 We found that reaction of

diisopropyl squarate (142) with dimethylphenylsilyllithium in THF at -78 *C followed by

treatment with concentrated aqueous HCl affords silyl-substituted cyclobutendione 144 in 84-

86% yield.

92. McElvain, S. M.; Kundiger, D. Organic Syntheses; Wiley: New York, 1955; Collect. Vol. III, pp 506-509.93. Liebeskind, L. S.; Fengl, R. W.; Wirtz, K. R.; Shawe, T. T. J. Org. Chem. 1988, 53, 2482.

49

i-PrO 0

i-PrO 0

142

PhMe 2SiLi i-PrO OH PTHF SiMe 2Ph conc. HCI

-78 *C, 30 mn iPr 84-86%

143

hMe2Si 0

i-PrO O

144

H

PhMe2Si 0

i-PrO H

145

Liebeskind also reported that selective reduction of the more reactive carbonyl group

of squaric acid derivatives occurs smoothly upon addition of lithium tri-(tert-

butoxy)aluminum hydride at reduced temperatures. 94 We studied the reduction of

cyclobutendione 144 with a variety of hydride reagents, as summarized in Table 1. Of the

reagents investigated, only the lithium tri-(t-butoxy)aluminum hydride successfully reduced

cyclobutenedione 144, but the reaction required a large excess of the hydride reagent (entry

6). Although this route appeared feasible, we were discouraged by the multiple steps required

to access the desired cyclobutenone intermediate and therefore this route was not studied

further.

94. (a) Edwards, J. P.; Krysan, D. J.; Liebeskind, L. S. J. Org. Chem. 1993, 58, 3942. (b) Krysan, D. J.;Gurski, A.; Liebeskind, L. S. J. Am. Chem. Soc. 1992, 114, 1412.

50

Scheme 15

PhMe 2Si 0- PhMe2S! 0

)q 90H

i-PrO i-PrO H

144 145

Conditions

i-Bu 2AlH, CH 2C12, -78 *C, 1 h

Li s-Bu3BH , THF, 0 'C, 5 min

NaBH 4, i-PrOH, rt, 5 min

LiAl(t-BuO) 3H, THF, 0 *C to rt, 2 days

LiAl(t-BuO) 3H, THF, -78 *C to rt, 2 days

LiAl(t-BuO) 3H, THF, -40 *C, 45 min

Red-Al, Et2O, 0 *C, 2 h

Complex mixture

Complex mixture

Complex mixture

No reaction observed

No reaction observed

-30%

Complex mixture

Alternative Routes to TAS-Vinylketenes: Preliminary Studies

Currently in our laboratory, Kevin Shea is investigating the preparation of TAS-vinyl-

and alkynylketenes via the cross coupling reactions of stannyl- and halo(silylketenes) as

outlined in Scheme 16. Murai previously reported a synthesis of

tributylstannyl(trimethylsilyl)ketene 146 from (trimethylsilyl)diazomethane as discussed in

Chapter 1. We expect that ketene 146 may undergo Stille-type reactions with alkenyl- and

alkynyl halides. Alternatively, reaction of bromo(trialkylsilyl)ketenes6 4a (148) with alkenyl-

or alkynylstannanes should afford unsaturated ketene products as well. This method would

allow efficient access to a wide array of unsaturated TAS-ketenes (147), including the elusive

TAS-alkynylketenes.

51

Table I

Entry

1

2

3

4

5

6

7

Results

1.2

1.2

1.6

1.2

2.3

5.0

1.2

equiv

equiv

equiv

equiv

equiv

equiv

equiv

n-BuLi;CO;Bu 3SnCI

Me3SiCHN 2

X R'

C R2

Me 3Si SnBu3

146

0 Br2; 0C Et 3N

R3Si H R3Si Br148

R3Si C -

R3Sn.., R'

I R'R2 R

147

We have developed efficient methods for the synthesis of TAS-vinylketene and

continue to investigate new, improved routes to these compounds. In what kind of

synthetically significant reactions do these novel compounds participate? The remainder of

this thesis explores this provocative question. Chapters 3 and 4 detail the [4+2]

cycloadditions of TAS-vinylketenes, and Chapter 5 presents the application of these

compounds in a novel [4+1] annulation strategy for the synthesis of cyclopentenones.

52

Scheme 16

CHAPTER 3

APPLICATION OF TAS-VINYLKETENES AS DIENES IN THE DIELS-ALDER

REACTION: SYNTHESIS OF CARBOCYCLIC COMPOUNDS 95

Substituted cyclohexenones are important synthetic targets and are most commonly

prepared via the Robinson annulation. In recent years, considerable research has been

directed toward the development of alternative routes to cyclohexenones based on the Diels-

Alder reaction. Most of these methods have been based on the application of highly

functionalized dienes as 47t components, 96 and depending on the choice of diene, routes to

cyclohexenones with electron-withdrawing groups at C-4, C-5, and C-6 are now available.

Danishefsky has developed an elegant route to cyclohexenones in which oxygen

substituted dienes participate in [4+2] cycloadditions with electron-deficient dienophiles.97

The oxygen substituents impart a high degree of reactivity and regiospecificity on the dienes

which combine with activated dienophiles to produce cyclohexene intermediates. Subsequent

treatment of the enol ether with acid hydrolyzes the vinylogous hemiketal to produce

substituted cyclohexenones. Note that this methodology produces cyclohexenones in which

the dienophile activating (W) group ends up at the C-4 position of the cyclohexenone ring.

OR OR

WJ7 W +

RO - RO

Even moderately reactive dienophiles undergo cycloadditions with these unhindered,

electron-rich dienes. For example, 2-methylcyclohexenone reacts with Danishefsky's diene at

200 *C to afford the substituted cyclohexanone intermediate 151. Exposure of the reaction

mixture to a 3:1 THF:0.005 N HCl solution then produces the cis-fused cyclohexenone 152 in

95. Portions of this chapter are published in ref 79b.96. (a) Fringuelli, F.; Taticchi, A. Dienes in the Diels-Alder Reaction; Wiley: New York, 1990. (b) Petrzilka,

M.; Grayson, J. I. Synthesis 1981, 753. (c) Danishefsky, S. Acc. Chem. Res. 1981, 14, 400.

53

47% yield (eq 39). Subsequently, Danishefsky reported the synthesis of chiral l-alkoxy-3-

trimethylsilyloxy-1,3-dienes for asymmetric Diels-Alder reactions. 98' 99 In a related method,

Rawal has reported that chiral l-amino-3-silyloxy-1,3-dienes participate in Diels-Alder

reactions with a,p-unsaturated aldehydes and esters to afford substituted cyclohexenones in

high ee.100

HMe3S!O xylene Me 3SiO

+ 200 *C 20 h

47%OMe 0 MoO 0

149 150 151

I HCI (39)

H

00

152

Trost has investigated a novel Diels-Alder approach for the preparation of

cyclohexenones in which 2-alkoxy-3-thio-substituted dienes undergo [4+2] cycloadditions

with electron-poor dienophiles to afford cycloadducts that can be converted to

cyclohexenones by enol ether hydrolysis, oxidation of the sulfide to a sulfoxide, and thermal

elimination. This methodology generates cyclohexenones in which the dienophile W-group is

ultimately found at the C-5 position of the new ring.' Trost has employed this approach for

the synthesis of carvone (156) as illustrated in Scheme 17.

97. (a) Danishefsky, S.; Kitahara, T. J. Am. Chem. Soc. 1974, 96, 7807. (b) Danishefsky, S.; Kitahara, T. J.

Org. Chem. 1975, 40, 538.98. (a) Bednarski, M.; Danishefsky, S. J Am. Chem. Soc. 1983, 105, 6968. (b) Danishefsky, S.; Bednarski,

M.; Izawa, T.; Maring, C. J. Org. Chem. 1984, 49, 2290. (c) Bednarski, M.; Danishefsky, S. J. Am.

Chem. Soc. 1986, 108, 7060.99. For other examples of chiral alkoxy dienes, see: (a) Trost, B. M.; O'Krongly, D.; Belletire, J. L. J. Am.

Chem. Soc. 1980, 102, 7595. (b) Lubineau, A.; Queneau, Y. J. Org. Chem. 1987, 52, 1001. (c) Gupta, R.

C.; Larsen, D. S.; Stoodley, R. J.; Slawin, A. M. Z.; Williams, D. J. J. Chem. Soc., Perkin Trans. 1 1989,739. (d) Tripathy, R.; Carroll, P. J.; Thornton, E. R. J. Am. Chem. Soc. 1991, 113, 7630. (e) Boehler, M.

A.; Konopelski, J. P. Tetrahedron 1991, 47, 4519.100. Kozmin, S. A.; Rawal, V. H. J. Am. Chem. Soc. 1997, 119, 7165.101. Trost, B. M.; Vladuchick, W. C.; Bridges, A. J. J. Am. Chem. Soc. 1980, 102, 3554.

54

Oy W RO ZQW RO + W

Scheme 17

0 0MeO A 2h MeO

PhS 75% PhS

153 154 (4:1 regioisomers)155

1) Ph 3P=CH 2 , THF2) HCO 4, aq CH 3CN3) NaH, Mel, THF4) mCPBA5) (MeO) 3P, A

21%

carvone 0156

A third objective of interest in this area of Diels-Alder based cyclohexenone synthesis

has been the development of vinylketene equivalents capable of participating as diene

components in [4+2] cycloadditions. In contrast to the Diels-Alder strategies summarized

above, these reactions would produce cyclohexenones in which the dienophile W-group

would appear at the C-6 position of the new cyclohexenone ring. Unfortunately, the tendency

of vinylketenes to form only [2+2] cycloadducts with alkenes (eq 40) and the intrinsic

instability of these ketenes generally precludes their direct use as [4+2] enophiles, as

discussed in Chapter 1.24

55

W 0, W

[2+2] C + [+ (40)

Vinylketene acetals have found use as vinylketene surrogates; however, the scope of

this methodology is somewhat limited because of the sensitivity of these compounds and the

fact that the s-cis conformation required for cycloaddition is usually disfavored in these Z-

substituted dienes. Substitution by an electron-donating alkoxy- or silyloxy-group at C-3

dramatically increases the reactivity of vinylketene acetals, and these activated dienes readily

participate in [4+2] cycloadditions with a variety of dienophiles.C"lO 2 For example, the

oxygenated diene 157 combines rapidly with dimethylacetylene dicarboxylate (DMAD) in

refluxing benzene to afford substituted phenol 159 in 89% yield (eq 41).

OMe CO 2Me benzene OH

+ 80 *C 30 min- OP CO2Me (41)

Me 3SiO CO2Me 89% HO CO2 Me

157 158 159

In 1980, our laboratory reported that (trimethylsilyl)vinylketene participates as a

reactive enophile in [4+2] cycloadditions with standard electron-deficient Diels-Alder dienes.

This chemistry is summarized later in this section. Several years later, Ghosez reported that

N-silylated silylvinylketenimines can function in a similar manner as vinylketene equivalents

in Diels-Alder reactions with olefinic and acetylenic dienophiles.103 In an example of this

interesting cycloaddition, ketenimine 160 combines smoothly with methyl acrylate at 150 'C,

and subsequent protodesilylation with potassium fluoride in refluxing methanol yields

dihydroaniline 162 (eq 42). Hydrolysis to the cyclohexenone 163 can then be effected using

1.2 N HCl. Ketenimine 160 also reacts in [4+2] cycloadditions with methyl crotonate,

various acetylenic esters, and napthoquinones.

102. Savard, J.; Brassard, P. Tetrahedron 1984, 40, 3455.103. Differding, E.; Vandevelde, 0.; Roekins, B.; Van, T. T.; Ghosez, L. Tetrahedron Lett. 1987, 28, 397.

56

Si(i-Pr)h,N 1)150*C NH2(i-Pr)Si C CO 2Me 2) KF-MeOH (i-PrhSi C02MO

70%

160 161 162

1.2 N HCICCi 4

(42)

70%

0(i-Pr)hSi CO 2Me

163

[4+2] Cycloadditions of TAS-Vinylketenes

Several years ago, our laboratory reported the first synthesis and isolation of

(trimethylsilyl)vinylketene (92)." As expected, 92 fails to engage in typical ketene [2+2]

cycloadditions and instead participates as a diene component in [4+2] cycloadditions with

electron-deficient alkenes and alkynes. The ketene carbonyl group is predicted to donate

electron density into the diene system (see resonance hybrid, below), activating this system

for [4+2] cycloadditions with standard electron-deficient dienophiles.

6 OR3Si CRSI C

The trimethylsilyl group confers only a weak directing effect on Diels-Alder reactions

of 1- and 2-(trimethylsilyl)-1,3-dienes 10 4 and is not expected to contribute significantly to the

regiochemical course of TAS-vinylketene reactions. As mentioned above, the carbonyl

oxygen of the vinylketene is predicted to function as an electron-donor substituent and

104. (a) Fleming, I.; Percival, A. J. Chem. Soc., Chem. Commun. 1976, 681. (b) Fleming, I.; Percival, A. J.Chem. Soc., Chem. Commun. 1978, 178. (c) Jung, M. E.; Gaede, B. Tetrahedron 1979, 35, 621. (d) Batt,D. G.; Ganem, B. Tetrahedron Lett. 1978, 3323. (e) For a review on the application of silyl-substitutedconjugated dienes in synthesis, see: Luh, T.-Y.; Wong, K.-T. Synthesis 1993, 349.

57

therefore is expected to control the regiochemical outcome of these cycloadditions. In fact,

early experiments by Howard Sard in our laboratory confirmed the expectation that [4+2]

cycloadditions with TAS-vinylketenes would be highly regioselective.7 3 For example, methyl

propiolate (164) reacts smoothly with (trimethylsilyl)vinylketene (92) to afford 3-

(trimethylsilyl)salicylate (165) as a single regioisomer (eq 43). Note also that the cycloadduct

undergoes facile protodesilylation when treated with trifluoroacetic acid at room temperature

to yield 166.

, CO toluene OHMe3Si C 2 95 "C 63 h Me3Si CO2Me

+ |

45%92 164 165

TFA, CH 2Cl225 *C 24h (43)

78%

OH

CO 2 Me

166

(Trimethylsilyl)vinylketene also reacts with olefins such as diethyl fumarate and N-

phenylmaleimide to afford cyclohexenones. Unfortunately, the scope of these reactions is

restricted by the limited thermal stability of this vinylketene (which begins to decompose at

temperatures above 95 *C), and ketene 92 does not undergo cycloadditions with less reactive

dienophiles in satisfactory yield.

Originally, substituted TAS-vinylketenes were not available by the

dehydrohalogenation route. However, development of the photo Wolff rearrangement of a-

silyl-a-diazo ketones made the more robust, substituted ketenes available. We expected that

these derivatives would engage in [4+2] cycloadditions with a broader range of substrates

than (trimethylsilyl)ketene, providing a useful synthetic route to highly substituted

cyclohexenones and phenols. We therefore undertook a systematic investigation of Diels-

Alder reactions of these substituted TAS-vinylketenes, focusing on the scope and

58

stereochemical course of the process. Initial experiments by Jennifer Loebach of our

laboratory confined the feasibility of these reactions and, subsequently, my research goal has

been to extend the scope of this interesting methodology by investigating the use of different

substituted TAS-vinylketenes, other reactive dienophile partners (including heterodienophiles,

see Chapter 4), and Lewis acid catalysts.

The relative reactivity of different TAS-vinylketenes was explored initially by

examining their cycloadditions with the reactive dienophile, dimethyl acetylenedicarboxylate

(DMAD). In a typical reaction, a degassed toluene solution of TAS-vinylketene 111 and 1.0

equiv of DMAD was heated at 150 *C for 16 h in a threaded Pyrex tube sealed with a Teflon

cap. Concentration and purification by chromatography on silica gel provided the expected

phenol 167 in nearly quantitative yield (eq 44). Protodesilylation of this cycloadduct

proceeded readily (80% yield) upon exposure to 20 equiv of TFA in dichloromethane at room

temperature for 2 h.

'0 CO 2Me 150 C OH(+-Pr)3Si C toluene 16 h (i-Pr)3 Si , CO 2Me

+I ICO 2Me 98% CO 2Me

111 158 167

20 equiv TFACH2CI2 0rt 2 h

80%

OH

CO 2MO

CO2M.

168

Interestingly, cycloaddition with the corresponding triethylsilyl derivative 112

afforded a mixture of products tentatively assigned as the silyl ether 169 and phenol 170 (eq

45). In this case, the acidity of the phenolic product apparently promotes desilylation of the

cycloadduct at the elevated reaction temperature. This process is not as facile in the case of

the triisopropysilyl derivative 111 because of the steric bulk of the silyl group.

59

EtS "*0 C 2Me 150 0C OSEt3 OHEt3SirC,0 CO2Metoluene 16 h CO02Me O CO 2Me

+ C02| + C0M (45)

CC2Me

112 158 169 170

29% 20%

Further experiments investigated by Loebach confirmed that a wide range of

substituted TAS-vinylketenes would react with DMAD. As discussed in Chapter 2,

cyclobutenones undergo 4n-electrocyclic ring opening to afford TAS-vinylketenes. We were

interested in using this method to generate ketenes in situ from cyclobutenones for reaction

with DMAD. As illustrated in eq 46, heating 2-silylcyclobutenones 125 and 126 in toluene at

reflux in the presence of DMAD affords phenols 171 and 172 in good yield. Presumably,

cyclobutenones 125 and 126 undergo an in situ electrocyclic ring opening to generate TAS-

vinylketenes which then react in [4+2] cycloadditions. Lower yields of the desired phenols

were obtained when the reactions were conducted at higher temperatures and butenolide side

products were isolated (see Chapter 2 for a discussion). It is interesting to note that in both

cases, none of the rearranged products (akin to 169 and 170) were observed.

R3Si o toluene 2. R0 C2Me OH3 reflux 42-55 h R3 RSi CO2Me

Ph Ph X CO2 Me Ph CO2Me

125 (R = Me) 158 171 (R = Me) 63%

126 (R = Et) 172 (R = Et) 55%

The scope of TAS-vinylketene cycloadditions encompasses other reactive dienophiles

such as ethyl cyanoacrylate, nitroethylene, and cyanoallene. Since ethyl cyanoacrylate bears

two electron-withdrawing groups, the cycloadditions of this dienophile were not informative

with regard to the stereochemical course of the Diels-Alder reactions of TAS-vinylketenes.

Simple acrylate derivatives proved insufficiently reactive, so we turned our attention to

reactions of nitroalkenes. TAS-Vinylketene 111 combines with nitroethylene in a

regioselective Diels-Alder reaction, but because of the high acidity of the doubly activated a'

60

proton in the cycloadduct, the product is expected to undergo equilibration. Therefore, the

identity of the major product does not necessarily provide insight into the stereochemical

course of the cycloaddition stop. No such ambiguity is associated with the reaction of 2-

nitropropene with ketene 111 (eq 47), and indeed this cycloaddition produces a single

cycloadduct in 35% yield whose structure was established by Loebach as 174 by analysis of

'H NMR coupling constant data and an NOE study.

0 toluene O(i-PraSi C"' N02 reflux 116 h (i-Pr)3 Si N02

+ y/CH 3 (47)35%

CH3

111 173 174

This result indicates that as in other Diels-Alder reactions, cycloadditions involving

TAS-vinylketenes follow the Alder endo rule and prefer transition states in which the

dienophile activating group adopts an endo orientation (175) relative to the diene system (eq

48). Disappointingly, vinylketene 111 failed to react with less reactive dienophiles such as

N-phenylmaleimide and chloroacrylonitrile; in each case, complex mixtures of products were

obtained.

0 2N

(i-Pr)Si C=0 endo , 174 (48)

175

Other work by Loebach established that monoactivated allenes are excellent partners

for [4+2] cycloadditions with TAS-vinylketenes, providing access to phenolic products that

could otherwise be produced only by cycloadditions with less reactive acetylene dienophiles.

For example, addition of cyanoallene to TAS-vinylketene 111 in toluene at 150 *C gave a

single product in good yield, which NOE studies revealed to be not the expected o-

(triisopropylsilyl)phenol, but rather the isomeric silyl ether 178 (Scheme 18). Isomerization

61

of the presumed initial cycloadduct 175 to 178 may proceed via the intermediacy of the a-

silyl ketone 177. 1,3-Silyl shifts of a-silyl ketones to form silyl enol ethers are well-known

processes.105

Scheme 18

.0(i-Pr)3Si C'

[4+2]cycloaddition

0R3Si CN

I -

H2C=C=: CNH

150 "Ctoluene 31 h

67%

0RASi CN

(i-Pr)hSiOCN

178

1, 3-silylshift

0R3Si CN

I

- 175 176 177

We anticipated that other reactive allenes might also participate in cycloadditions with

TAS-vinylketenes and so we next examined the reaction of acetylallene with the vinylketene

generated in situ by electrocyclic ring opening of 125. Thus, a degassed toluene solution of

acetylallene (179) and cyclobutenone 125 was heated at reflux for 20 h, and the desired

cycloadduct was isolated in 37-45% yield (eq 49). None of the silyloxy product resulting

from. 1,3-silyl rearrangement was observed as in the cyanoallene case. Disappointingly,

reaction of acetylallene with the triisopropylsilyl-substituted ketene 111 led to even lower

yields (25-26%) of the phenolic product.

105. See: Munschauer, R.; Maas, G. Angew. Chem., Int. Ed. Engl. 1991, 30, 306 and references therein.

62

o [ 0 OH OMe3Si 0 toluee Me 3Si Me 3SI+ ~reflux 20h (9

Ph ' J437-45%9)

L125 179 180 181

The acetylallene for this study was prepared in two steps by the method of Buono (eq

50).10 First, reaction of 2,4-pentanedione 182 with triphenylphosphine-bromine in

dichloromethane affords a mixture of (E/Z)-2-bromo-4-oxo-2-pentene 183 and a

triphenylphosphine oxide hydrobromide salt, which is precipitated by the addition of ether.

After concentration of the filtrate, the resulting vinyl bromide 183 is used directly in the next

step without further purification. Dehydrobromination of 183 occurs upon treatment with

triethylamine in ether. The solvent is removed by distillation and pure acetylallene (179) is

isolated in 59-65% yield following careful distillation at reduced pressure. The colorless

liquid (bp 62 *C, 80 mmHg) polymerizes readily and must be stored over a small amount of

hydroquinone to inhibit this undesired reaction.

(000 0 PPh3-Br2 Br 0Et3N If<(0

57-65%overall

182 183 179

Rigby has reported the application of the [4+2] cycloadditions of vinyl isocyanates

and benzyne107 as a useful method for the preparation of phenanthridinones (eq 51).'O8 We

expected that this highly reactive acetylene species might react with TAS-vinylketenes in

good yield to afford interesting polycyclic products.

106. (a) Buono, G. Synthesis 1981, 872. (b) Constantieux, T.; Buono, G. Org. Synth., in press.107. For reviews on the preparation and [4+2] cycloadditions of arynes, see: (a) Gilchrist, T. L. In The

Chemistry ofFunctional Groups; Patai, S.; Rappoport, Z., Eds.; Wiley: New York, 1983, pp383-419. (b)Hart, H. In The Chemistry of Triple-Bonded Functional Groups, Supplement 12; Wiley: New York, 1994,pp 1017- 1134.

63

0NH 2 Pb(OAc)

4 C 1N% CH 2CI 2 rt N HN(

N I:5 .C a x (51)N 58%

184 185 186 187

Benzyne can be generated by a variety of methods. Rigby obtained the best results

with Diels-Alder reactions of vinyl isocyanates when the benzyne was generated by Pb(OAc) 4

oxidative decomposition of 1-aminobenzotriazole (184). With this method for benzyne

preparation, reaction can occur over a broad range of temperatures (-78 *C to 55 *C).

However, side reactions sometimes occur in the presence of the lead tetraacetate, a powerful

oxidizing agent, and acetic acid is a byproduct of the reaction. Generation of benzyne with

lead tetraacetate is believed to proceed through a nitrene intermediate (188, eq 52).

Fragmentation of the ortho-fused benzene derivative expels two molecules of nitrogen, lead

diacetate, and acetic acid, none of which compete for benzyne.

NH 2 N:I I

N Pb(OAc)4 N (

// N - -,*-.-/ (52)N Pb(OAC) 2 N - 2N2

2 AcOH184 188 186

1-Aminobenzotriazole (184) is commercially available but is quite expensive, so for

our study this benzyne precursor was prepared by amination of benzotriazole according to the

procedure of Campbell and Rees (eq 53).109 Thus, treatment of triazole 189 with

hydroxylamine-o-sulfonic acid in aqueous potassium hydroxide affords 1-aminobenzotriazole

in 13% yield (lit. 38%).

64

108. (a) Rigby, J. H.; Holsworth, D. D.; James, K. J. Org. Chem. 1989, 54, 4020. (b) Rigby, J. H.; Holsworth,D. D. Tetrahedron Lett. 1991, 32, 5757.

109. Campbell, C. D.; Rees, C. W. J Chem. Soc. C 1969, 742.

H

NN

N

189

NH 2OSO 3H

aq KOH

13%

NH2

NN

N

184

Unfortunately, attempted reaction of TAS-vinylketene 190 with benzyne (generated as

described above by the lead tetraacetate method) led only to unreacted starting material and

the benzyne dimer biphenylene (192, eq 54). Apparently, reaction of TAS-vinylketenes with

benzyne occurs more slowly than benzyne dimerization.

Et 3S

Ph

190

OH

+ 0 'a i (,Et3Si )

186 191

192

In an attempt to expand the scope of the [4+2] cycloadditions of TAS-vinylketenes to

include dienophiles of low reactivity, the application of Lewis acids as catalysts for the

reaction was examined. It should be noted that Lewis acids have found considerable use as

catalysts for [2+2] cycloadditions of TAS-ketenes with aldehydes." Unfortunately, attempts

to promote the cycloaddition of vinylketene 111 with methyl acrylate or DMAD using either

Me2AlCl or BF 3-OEt2 at various temperatures proved unsuccessful (see Table 2). In all cases,

extensive decomposition of 111 was observed and none of the desired cycloadduct was

detected.

110. See ref 58 and the following: Brady, W. T.; Saidi, K. J. Org. Chem. 1979, 44, 733.

65

(53)

Et3Si V

184Pb(OAC)4

CH 2C2

reflux

38%126

(54)

Attempted Lewis Acid Catalysis of Diels-Alder Cycloadditions of TAS-Vinylketenes

Entry

1

Ketene

111

E

2 111

3 111

4 111

5 111

6 111

ienophile Conditions

DMAD 1.0 equiv Me2AlClCH2CL2, 0 *C to rt

DMAD 1.0 equiv Me2A1ClCH2Cl2, -78 *C to rt

DMAD 0.1 equiv BF3-OEt2CH2Cl 2, 0 'C to rt

DMAD 0.1 equiv BF 3-OEt2

CH2 Cl 2 , -78 *C to rt

methyl 1.0 equiv Me2AlClacrylate CH2 Cl 2, 0 *C

methyl 0.1 equiv BF 3-OEt2acrylate CH2Cl 2, 0 'C

Result

rapid decompositionof the ketene






TAS-vinylketenes bearing electron-donating substituents are expected to exhibit

increased reactivity in [4+2] cycloadditions, expanding the scope of this methodology to

include less activated dienophiles. Our initial attempts to prepare 3-alkoxy TAS-vinylketenes

have thus far been unsuccessful (as discussed in Chapter 2), however, further studies are

planned to explore this important area.

R3Si C 0.

RO

We have shown that TAS-vinylketenes undergo highly regioselective [4+2]

cycloadditions with reactive olefinic and acetylenic dienophiles to produce highly substituted

cyclohexenones and phenols. The ketene carbonyl dominates in controlling the regiochemical

course of the reaction and the stereochemical course of these cycloadditions follows the Alder

66

Table 2

Entry

endo rule. The syntheses of oxygen and nitrogen heterocycles via the hetero Diels-Alder

reactions of TAS-vinylketenes is described in Chapter 4.

67

CHAPTER 4

APPLICATION OF TAS-VINYLKETENES AS DIENES IN THE DIELS-ALDER

REACTION: SYNTHESIS OF HETEROCYCLIC COMPOUNDS

The hetero Diels-Alder reaction is an extremely important method for the synthesis of

heterocyclic compounds."1 A wide variety of heterodienophiles participate in [4+2]

cycloadditions, including carbonyl compounds, imines, iminium salts, and azo- and nitroso

compounds. This methodology often allows the rapid construction of complex molecules

from simple starting materials and has been applied for the synthesis of many heterocyclic

natural products. In recent years, asymmetric versions of hetero Diels-Alder reactions have

also been reported.

As discussed in the preceding chaper, we have shown that TAS-vinylketenes take part

in [4+2] cycloadditions with olefins and acetylenes, and we anticipated that carbonyl

compounds and imines might also function as dienophiles providing efficient routes to

interesting heterocycles. This chapter describes the results of our investigation of the hetero

Diels-Alder reactions of TAS-vinylketenes, including a discussion of the scope,

stereochemical course, and mechanism of the process.

Diels-Alder Reactions with Carbonyl Dienophiles: Background

In 1949, Gresham and Stedman reported the first example of the oxa Diels-Alder

reaction of a carbonyl dienophile with a conjugated diene.1 1 2 Specifically, heating 2,4-

dimethyl-1,3-butadiene and formaldehyde affords 5,6-dihydropyran 194 in 60% yield (eq 55).

Unfortunately, similar reactions of diene 193 with higher aldehydes leads only to low yields

of products.

111. For reviews on the hetero Diels-Alder reaction, see: (a) Tietze, L. F.; Kettschau, G. Top. Curr. Chem.

1997, 189, 1. (b) Weinreb, S. M. In Comprehensive Organic Synthesis; Trost, B. M.; Fleming, I., Eds.;Pergamon Press: Oxford, 1991; Vol. 5, pp 401-449. (c) Boger, D. L. In Comprehensive OrganicSynthesis; Trost, B. M.; Fleming, I., Eds.; Pergamon Press: Oxford, 1991; Vol. 5, pp 451-512. (d) Boger,D. L.; Weinreb, S. M. Hetero Diels-Alder Methodology in Organic Synthesis; Academic Press: San

Diego, 1987. (e) Waldmann, H. Synthesis 1994, 535.112. Gresham, T. L.; Stedman, T. R. J. Am. Chem. Soc. 1949, 71, 737.

68

0+ H ) H

193

A6,

60%

(55)

194

Later findings confirmed that simple 1,3-dienes react best with electron-deficient

carbonyl compounds. For example, diethyl oxomalonate (196) combines with 2-methyl-1,3-

butadiene to afford the pyran cycloadduct 197 in 80% yield (eq 56).113 Less reactive carbonyl

compounds will participate in [4+2] cycloadditions in the presence of Lewis acid catalysts or

under the application of high pressure.

-Ic +

195

0

EtO2C CO 2Et

196

CH 3CN130 *C 4 h

80%

(56)

CC2Et197

Vinylketene equivalents such as 198 have been shown to undergo Lewis-acid

catalyzed Diels-Alder reactions with aldehydes. For example, as illustrated in eq 57, electron-

rich diene 198 and amino aldehyde 199 combine in the presence of diethylaluminum chloride

to afford a 92:8 mixture of lactones 200 and 201.1 14

OSIMe3 0 EtAICI 0

OMe + H i-Pr 2 0-l i-Pr

MeO NHBoc 67-83% MeONHBoc

198 199 200

0

+0 -Pr (57)MeO&

MHBoc201

92: 8

113. Bonjouklian, R.; Ruden, R. J. Org. Chem. 1977, 42, 4095.

69

Diels-Alder Reactions of TAS-Vinylketenes with Carbonyl Dienophiles

We anticipated that TAS-vinylketenes would participate in oxa Diels-Alder reactions

with activated carbonyl compounds to afford aP-unsaturated 8-valerolactones. The proposed

oxa Diels-Alder reaction of TAS-vinylketenes was initially investigated by examining

cycloadditions with the highly reactive and commercially available oxadienophile diethyl

oxomalonate. In a typical reaction, an acetonitrile solution of silylketene 111 and 1.5 equiv of

diethyl oxomalonate (196) was heated at reflux for 15 minutes. Concentration and

purification by chromatography on silica gel afforded the desired a,P-unsaturated 6-

valerolactone 202 in 94% yield (eq 58). As shown in previous studies (Chapter 3), the ketene

carbonyl functions as an electron-donor substituent and directs the regiochemical course of

the reaction.

0(i-Pr)hSi C.0 CH 3CN (i-Pr)3SI

0 reflux 15 min 0 (58)A3 2 CO 2Et

EtO2C CO2Et 94% Ha CO 2Et

111 196 202

1H NMR spectral data supports the structural assignment for compound 202 and

includes a diagnostic resonance at 3.21 ppm (quartet) that corresponds to proton Ha on the

lactone ring. The IR spectrum contains a strong carbonyl stretching frequency at 1715 cm1

due to the lactone carbonyl. In addition, the 13C NMR spectrum exhibits characteristic

resonances for the aP-unsaturated lactone carbonyl carbon at 168.9 ppm, two alkenyl

carbons (166.6 and 122.0 ppm), and the C-2 carbon at 85.0 ppm.

The reaction of oxomalonate 196 with the cyclohexenyl TAS-ketene 110 also

proceeds smoothly to afford lactone 203 in 77% yield (eq 59). With this more sterically

hindered ketene, the reaction rate is slower and requires two hours in refluxing acetonitrile to

reach completion.

70

114. (a) Midland, M. M.; Afonso, M. M. J. Am. Chem. Soc. 1989, 111,4368. (b) Midland, M. M.; Koops, R.W. J. Org. Chem. 1990, 55, 4647. (c) Midland, M. M.; Koops, R. W. J. Org. Chem. 1990, 55, 5058.

0(i-Pr)3Si C"0 CH3CN (i-Pr)3SI

+ Oreflux 2 h 0 (59)

Et02C AC02Et 77% C2Et

110 196 203

TAS-Vinylketenes generated by the electrocyclic ring opening of cyclobutenones also

participate in oxa Diels-Alder reactions. For example, addition of oxomalonate 196 to a

refluxing acetonitrile solution of cyclobutenone 126 affords the predicted cycloadduct 205 in

excellent yield (eq 60).

Et3SI CH 3CN E C~ 1960 Et3SI%.

E P I reflux 15 h E C 196 Et00.. CO 2 Et (60)

[L Ph J 92% Ph CO 2 Et

126 204 205

As expected, unactivated carbonyl compounds such as pivaldehyde do not participate

in cycloaddition reactions with TAS-vinylketenes. Lewis acids have found considerable use

as catalysts for [2+2] cycloadditions of TAS-ketenes with aldehydes and we hoped that Lewis

acids might promote the oxa Diels-Alder reaction of TAS-vinylketenes.106 Unfortunately,

attempts to promote the cycloaddition of vinylketene 111 with pivaldehyde using ZnCl2

proved unsuccessful, resulting only in slow decomposition of the ketene; none of the desired

cycloadduct was observed. Previous attempts to catalyze Diels-Alder reactions of TAS-

vinylketenes with other Lewis acids gave similarly disappointing results (see Chapter 3).

Finding effective conditions for protodesilylation of the lactone cycloadducts required

a substantial amount of investigation. A variety of acidic conditions were explored for the

protodesilylation of lactone 202 (Table 3), but none resulted in clean removal of the

triisopropylsilyl group. However, protodesilylation of the triethylsilyl derivative 205

proceeds more easily. Thus, treatment of lactone 205 with 5 equivalents of methanesulfonic

acid in refluxing dichloromethane affords the desired lactone 206 in quantitative yield (entry

6, eq 61).

71

Protodesilylation of a-Silylcyclopentenones

EiEResult

no reaction, recoveredstarting material in 88% yield

no reaction

complex mixture

no reaction

complex mixture

100%

Cycloadditions with Imino Dienophiles: Background

In addition to carbonyl compounds, we were also interested in exploring the utility of

imines as dienophiles in [4+2] cycloadditions with TAS-vinylketenes. An enormous amount

of research has been devoted to the study of electron-deficient imines such as N-sulfonyl and

N-acylimines as well as iminium salts as dienophiles in the Diels-Alder reaction. These [4+2]

cycloaditions comprise an important strategy for the synthesis of nitrogen heterocycles.

Neutral imines generally react in [4+2] cycloadditions only with highly activated

dienes in the presence of a Lewis acid catalyst. Grieco has demonstrated that iminium salts

react with dienes in intermolecular [4+2] cycloadditions.115 The iminium salts are generated

in aqueous solution by the reaction of a primary amine with aqueous formaldehyde or by the

115. (a) Larsen, S. D.; Grieco, P. A. J. Am. Chem. Soc. 1985, 107, 1768. (b) Grieco, P. A.; Larsen, S. D.;Fobare, W. F. Tetrahedron Le#. 1986,27, 1975.

72

ritry Lactone

1 202

2 202

3 202

4 205

5 205

6 205

0Et3SI 0

Ph C02CO 2Et

205

Conditions

20 equiv TFACH 2Cl2, rt, 48 h

15 equiv MsOHMeOH, rt, 24 h

5 equiv TBAFTHF, rt, 1 h

20 equiv TFACH 2Cl2, rt, 5 days

1 equiv TBAFTHF, rt, 30 min

5 equiv MsOHCH2Cl2, reflux, 15 h

5.0 equiv MsOHCH2C 2

reflux 15 h

Et100%

0

Ph CO 2EtCO 2Et

Table 3

206

(61)

reaction of an aldehyde with an ammonium chloride solution. Unfortunately, iminium ions

derived from higher aldehydes are generally much less reactive.

Unactivated imines were of particular interest to our group as heterodienophiles for

Diels-Alder reactions with TAS-vinylketenes. Simple, uncharged imines are usually not

reactive as heterodienophiles in Diels-Alder cycloadditions unless partnered with reactive,

electron-rich dienes in the presence of a Lewis acid.' 16 For example, Brassard's oxygenated

diene 198 (a vinylketene equivalent) combines with chiral imine 207 with the aid of a Lewis

acid to form the lactam 208 in good yield (eq 62).117

OMe

OSiMe3 +

MeO 1

198

i-Pr

N Y1CO2t-Bu

Ph H

207

Et2AICICH 2CI2

-78 C to rt

84%

95% d.e.

o CO 2t-Bu

N I-Pr

MeOL74Ph

208

Danishefsky employed the imino Diels-Alder reaction as a key step in the total

synthesis of ipalbidine (Scheme 19).118 Reaction of vinylketene equivalent 210 with 1-

pyrroline in the presence of BF 3-OEt 2 affords the lactam intermediate 212. Reduction of the

lactam carbonyl and subsequent dehydration provides ipalbidine (213) in 78% yield.

116. For recent reviews on the imino Diels-Alder reaction, see ref 11 lb and: Boger, D. L.; Weinreb, S. M.

Hetero Diels-Alder Methodology in Organic Synthesis; Academic Press: Orlando, 1987; pp 35-71.

117. Waldmann, H.; Braun, M.; Drager, H. Tetrahedron: Asymmetry 1991, 2, 1231.

118. Danishefsky, S. J.; Vogel, C. J. Org. Chem. 1986, 51, 3915.

73

(62)

Scheme 19

LDA, THF-HMPT

CI2Et t-BuMe 2SiCI OSt-BuMe2 + No

MeO ~100% MeO / OEt

209 210 211

BF 3-Et 2O, CH 2C 2-78 OC to rt

40-45%

N 1) LiAIH 4-AIC 3 NN 2) BBr3, CH 2C 2 N

MO 78% Meo 0

ipalbidine213

We envisioned that the [4+2] cycloadditions of TAS-vinylketenes with imines might

provide another direct route to substituted 5,6-dihydro-2(lIH)-pyridones (a,p-unsaturated 3-

valerolactones).

It is well known that imines of the type R2C=NH are generally not stable.

Benzophenone imine (Ph2C=NH), however, is stable and is commercially available, and we

hoped that this ketimine would react in [4+2] cycloadditions with TAS-vinylketenes.

Unfortunately, when ketene 111 was heated with benzophenone imine in refluxing

acetonitrile, only a complex mixture of inseparable. products was obtained and none of the

desired lactam cycloadduct could be observed in the 1H NMR spectrum.

. Unlike N-H substituted imines, N-(trimethylsilyl)imines are quite stable under

anhydrous conditions. These imines are easily prepared and some can even be purified by

distillation at reduced pressure. Note that the cycloadduct derived from an N-

(trimethylsilyl)imine is expected to undergo protodesilylation under very mild conditions (e.g.

aqueous workup or silica gel chromatography) to afford the same products that would have

resulted from cycloaddition of the unstable imines of the type R2C=NH.

Few examples of [4+2] cycloadditions of N-(trimethylsilyl)imines have been reported

to date. Barluenga has reported that N-(trimethylsilyl)imines react as dienophiles in ZnCl 2 -

74

catalyzed [4+2] cycloadditions with 2-amino-1,3-dienes to afford tetrahydropyridones and 4-

piperidinones.119 For example, reaction of N-(trimethylsilyl)benzaldimine 215 with the

electron-rich chiral diene 214 affords enamine 216 which is converted on silica gel to the

enantiomerically enriched tetrahydropyridone 217 (eq 63). As suggested above, in these

reactions N-silylimines function as masked forms of the elusive imines of ammonia. Most

imines that are not substituted at nitrogen rapidly trimerize to the corresponding triazines.

OMe

ZnC " 1)NaHCO 3

NiMe3 THF N 2) SiO 2

Ph H 0Nme Ph 70%

82% e.e.

215 216

OMe

' NH

0 V Ph

217

The reaction of imines with ketenes has been exploited extensively for the synthesis of

P-lactams in the well known Staudinger reaction. N-(Trimethylsilyl)imines have also been

employed as participants in these [2+2] cycloadditions. In 1977, Birkofer described that

addition of N-(trimethylsilyl)benzophenone imine to diphenylketene affords the acyclic

intermediate 220 in quantitative yield; none of the P-lactam product was observed (Scheme

20). Heating N-(trimethylsilyl)benzaldimine with diphenylketene at 160 *C and subsequent

hydrolysis of the N-trimethylsilyl group provides P-lactam 222 in poor yield.2

119. (a) Barluenga, J.; Aznar, F.; Valdds, C.; Cabal, M.-P. J. Org. Chem. 1993, 58, 3391. (b) Barleunga, J.;

Aznar, F.; Valdds, C.; Martin, A. An. Quim. 1993, 115. (c) Barluenga, J.; Aznar, F.; Vald6s, C.; Martin,

A.; Garcia-Granda, S.; Martin, E. J. Am. Chem. Soc. 1993, 115,4403.

75

OMe

+

N

OMe

214

(63)

N' SIMe benzene+ reflux 30 min

100%

219

N'SiMe3 xylenePh H 160 C

1 12%

215

Me 3SiON

Ph /PhPh Ph

220

C

Ph Ph

218

0

Ph Ph

218

Panunzio reported that the [2+2] cycloadditions of ketenes (generated in situ from an

acid chloride) with N-(trimethylsilyl)imines is a useful method for the direct preparation of N-

H substituted azetidinones. For example, reaction of N-(trimethylsilyl)imine 224 with the

ketene derived from acid chloride 225 affords the acyclic azadiene intermediate 226 which

can be observed by analysis of the crude reaction mixture. Upon heating in refluxing toluene

for 6 h, cyclization occurs to afford exclusively the trans-substituted [-lactams 227 and 228 in

59% yield as a 80/20 mixture of the anti and syn isomers (Scheme 21).121

76

Scheme 20

0 SiMe0 N *Se3 H20

Ph f~Ph Ph

221

0NH

Ph ffPh Ph

222

120. Birkofer, L.; Schramm, J. Liebigs Ann. Chem. 1977, 760.121. Bandini, E.; Maretili, G.; Spunta, G.; Panunzio, M. Synlett. 1996, 1017.

Scheme 21

OSi(i-Pr)3 LiHMDSH heptane 0 *C

0

223

OSi(i-Pr)3H

N, SmN4SiMe

224

00 0

Ph 225

Et3N, 0 *Ctuen ~

0 rO OSI(i-Pr)3

PhNH

0 80:;227

20-

(i-Pr)3S1O

Ph N , N- ePh OSIlMe3

226

toluenereflux

59%

0 -O OSi(i-Pr)3

PhN H

0228

Panunzio found that syn selectivity can be obtained in the [2+2] cycloadditions of N-

(trimethysilyl)imines with ketenes by appending an alkoxy substituent on the ketene. For

example, reaction of imine 215 with the ketene derived from acid chloride 229 affords the cis-

substituted P-lactam 231 in 81% yield.122 Panunzio postulates that the cis-diastereoselectivity

results from the formation of a weak electrostatic bond between the imine carbon and the

ether oxygen. This interaction stabilizes the azadiene intermediate 230. A similar

electrostatic interaction had previously been proposed in an example of the classical

Staudinger reaction of an alkoxy-substituted ketene with an N-alkyl imine. 23

SiMe3

Ph" H

215

0Ph y0%)

CI229

Et3 N, 0 'Ctoluene

81%[.Bn

0, +H'Ph I

NMe 3SiO

230

BnO \Ph

0)NH

231

122. Bacchi, S.; Bongini, A.; Panunzio, M.; Villa, M. Synlett 1998, 843.123. Bose, A. K.; Chiang, Y. H.; Manhas, M. S. Tetrahedron Lett. 1972, 4091.

77

(64)

Imino Diels-Alder Reactions with TAS-Vinylketenes

We anticipated that N-(trimethylsilyl)imines might participate as dienophiles in Diels-

Alder reactions. Described in the following section is our systematic investigation of the

imino Diels-Alder reactions of TAS-vinylketenes, including our studies of the scope,

stereochemical course, and mechanism of the process.

Reactions with N-(Trimethylsilyl)benzaldimine

To test the feasibility of the proposed silylimine [4+2] cycloadditions, we initially

examined the reaction of TAS-vinylketenes with the easily prepared N-

(trimethylsilyl)benzaldimine. To our immense delight, we found that the combination of

silylketene 111 and 1.5 equivalents of imine 215 in refluxing acetonitrile provides the desired

cycloadduct (tentatively assigned as the N-silyl isomer 232) and no catalyst is needed to

promote the reaction! The bond to silicon is easily hydrolyzed from the crude product 232

during purification on silica gel to afford lactam 233 in excellent yield (eq 65). Similarly, in

subsequent cases, protodesilylation of the initial crude products occurs during silica gel

chromatography.

,0 CH3CN 0(i-Pr)3Si C NSiMe 3 reflux 1.5 h (i-Pr)Sai NSiMe3

Ph H Ph

111 215 232

Si02 (65)79-83%

0(i-Pr)3Si NH

Ph

233

78

The N-(trimethylsilyl)benzaldimine substrate (215) was prepared using the procedure

of Hart (eq 66).124 Thus, treatment of benzaldehyde with lithium hexamethyldisilylamide in

hexane affords the desired silyl imine 215 in 90% yield (lit. 89%) after distillation at reduced

pressure. This yellow liquid can be stored at 0 *C indefinitely under anhydrous conditions.

Note that this reaction proceeds via a heteroatom variant of the Peterson olefination,

presumably via the four-membered transition state 234.

Li hexane+ 0 C to rt 10 min

Me 3SI 'SiMe 390%

LiO- ;SiMe 3

Ph H SiM03 ]234-234

'I-Me 3SiOLi (66)

NSiMe 3

Ph H

215

'H NMR spectral data supports the structural assignment of compound 233, including

a resonance at 4.77 ppm due to the C-6 proton and a multiplet at 2.21 ppm corresponding to

the C-5 proton. The N-H proton is observed as a broad singlet at 5.51 ppm. These

assignments are consistent with data previously reported by Marson for the related lactam 235

in which the C-5 and 6 substituents are trans rather than cis (Figure 3).114

124. Hart, D. J.; Kanai, K.; Thomas, D. G.; Yang, T.-K. J. Org. Chem. 1983, 48, 289.

79

0

Ph AH

5.51 (br s)

0+(i-Pr)3Si H

PhH

Hb CHa4

4.77 (d, J= 3.6 Hz)

5.62 (br s)

0

NH56Ph2 "H

3 aH3C Hb 1 a

14.28 (dd, J= 11 and 7 Hz)2.59 (in)

235

Proof of the cis-stereochemistry of 233 was obtained from 'H NMR coupling constant

data and an NOE study (Figure 4). The proton at C-6 (Ha) is assigned to be on the same side

of the ring as proton Hb due to the observed equatorial-axial coupling constant (Jab = 3.6 Hz).

Also, irradiation of Ha results in a 9% enhancement of Hb, providing further confirmation that

Ha and Hb are on the same side of the ring. Although the trans isomer was not available for

comparison of NOE data, later NOE studies on an analogous compound support this

assignment.

Figure 4

Irradiate

9% +H3C Ha Si(i-Pr)3

Hb NHC Ph 'H3C

233

0%

'IH3C Ha Si(i-Pr)3

Hb tN.

H3C

Irradiate

80

Figure 3

2.21 (m)

233

The same cycloaddition was run with 1.0 equiv of imine 215 rather than 1.5 equiv as

in eq 65. However, in this case the time for the reaction to reach completion increased from

1.5 h to 3 h and the yield decreased slightly to 72%. Therefore, in most of the [4+2]

cycloadditions examined subsequently, a slight excess of imine was employed.

The rate of the [4+2] reaction was found to be noticeably faster in acetonitrile than in

toluene. Reactions in toluene were only approximately 50% complete by TLC analysis after

the same period of time in which the acetonitrile based reactions were finished. This increase

in rate in the more polar solvent suggests that the mechanism of the reaction proceeds via a

transition state involving more charge separation than the ground state (vide infra). We later

discovered that the hetero Diels-Alder reactions of TAS-vinylketenes proceed rapidly and at a

lower temperature in the absence of solvent. For example, ketene 112 combines with imine

215 (1.5 equiv) at room temperature in two hours (no solvent) to afford the cis-substituted

lactam 236 in 76% yield (eq 67). Again, an NOE study confirmed the cis assignment of the

substituents on this triethylsilyl-substituted lactam: irradiation of the C-6 proton resulted in a

6% enhancement of the neighboring C-5 proton.

0EtNSi C 0 no solvent EtSI Et 3i ,SiMe 3 rt 2 h; (67

+ N 0 X 6(67)A Si02 5 6

Ph H 7.Ph76%

112 215 236

The Diels-Alder reactions of N-(trimethylsilyl)benzaldimine and two other TAS-

vinylketenes were explored. Silylketene 110 undergoes cycloaddition with imine 215 to

afford the exo-substituted product 237 (as identified by an NOE study) in 91% yield (eq 68).

Irradiation of the C-1 proton (Ha) results in no enhancement of the C-8a proton (Hb).

However, one C-8 proton (He) exhibits a 7% enhancement, confirming the trans assignment

of lactam 237 (Figure 5).

81

CH 3CNN 'SiMO3 reflux 25 h;

Ph H Si0 2

91%

215

0(i-Pr)3S* NH

. Ph8

237

Assuming that these reactions follow a concerted cycloaddition mechanism, it would

appear that the steric effect of the cyclohexenyl substitutent in ketene 110 leads to a

preference for the transition state in which the phenyl group is exo (238, eq 69), whereas the

less hindered ketenes 111 and 112 prefer the transition state in which this group has the endo

orientation.

Me3Si,N

(i-Pr)3Si -.Ph

H

exo 237 (69)

238

82

(i-Pr)3Si C"O

110

Figure 5

(68)

7% Irradiate

H Ha SI(i-Pr)3

NHd Ph'HHb

t 2370%

[4+2] Cycloaddition of silylimine 215 with the TAS-vinylketene generated via in situ

electrocyclic ring opening of cyclobutenone 126 also was successful and afforded the lactone

239 in 84% yield after purification by column chromatography (eq 70).

SIM CH 3CNPN 0SIM3 reflux 45 min;

Ph ) H Si0 2

84%

215

0EtASi H

Ah[:Ph

239

(2) Reactions with N-(Trimethylsilyl)cinnamaldimine

Clearly, the imino Diels-Alder reactions of TAS-vinylketenes have a broad scope with

respect to the ketene. We next were interested in investigating the scope of the reaction with

regard to imino dienophiles. We began this study with a reaction of a vinyl-substituted imine,

N-(trimethylsilyl)cinnamaldimine (240), which was prepared according to the procedure of

Colvin.12 5 Treatment of commercially available (and pleasantly aromatic) cinnamaldehyde

with LiHMDS followed by trapping with chlorotrimethylsilane affords imine 240 as a yellow

liquid in 59% yield (lit. 95%) after distillation at reduced pressure.

Reaction of TAS-vinylketenes 111 and 110 with aldimine 240 occurs almost

instantaneously at room temperature. Under these solvent-free conditions, the cycloadducts

241 and 242 were obtained in 78% and 73% yield, respectively (Scheme 22).

125. Colvin, E. W.; McGarry, D.; Nugent, M. J. Tetrahedron 1988, 44, 4157.

83

Et 3SI 0

Ph

126

(70)

no solvent+ MO3S I N rt 15min;

SiO 2

78%240 Ph

no solvent+ MO3S I N rt 10min;

H SiO2

Ph 73%240

0(i-Pr)3Si NH

156

Ph241

0(i-Pr)3Si NH

181

Ph242

Support for the sterochemical assignments for the cis isomer 241 and the trans isomer

242 was obtained by NOE experiments. Irradiation of the C-6 proton (Ha) of 241 resulted in a

9% enhancement of the corresponding C-5 proton (Hb), indicating that the two protons have a

cis orientation. However, irradiation of the C-1 proton (Ha) of 242 resulted in no

enhancement of the C-8a proton (Hb) and instead showed an 8% enhancement of H0 , thereby

indicating that the lactam ring is trans-substituted (Figure 6).

Figure 6

9% Irradiate

I Ha Si(i-Pr)3HbH3C N 0

H3C HPh

241

8% Irradiate

He Ha Si(i-Pr)3

Hd Hb | H

t Ph

0% 242

84

Scheme 22

(i-Pr)3Si

111

110

(3) Reactions with N-(Silyl)ketimines

To further investigate the scope of this interesting cycloaddition, we next turned our

attention from aldimines to ketimines. Would this class of more highly substituted imines be

effective partners for cycloadditions with TAS-vinylketenes? We first undertook a study of

the reaction of TAS-vinylketenes with N-(trimethylsilyl)benzophenone imine (219), prepared

via the procedure of Rochow.126 Addition of benzonitrile to an ethereal solution of

phenyllithium and subsequent treatment with chlorotrimethylsilane affords the N-silyl imine

219 in 77% yield (lit. 55%) after purification by distillation at reduced pressure (eq 71).

PhLi, Et2Ort 15 min;

Ph-C=N Me 3SiCI

rt 22 h

243 77%

,'S!Me3NSIO

Ph Ph

219

We were delighted to discover that ketimine 219 reacts with ketene 111 to afford the

lactam 244 in 79% yield (eq 72). This cycloaddition proceeds at a significantly slower rate

(refluxing CH 3CN, 18 h) than the comparable reaction with aldimine 215 (refluxing CH3CN,

15 min), probably due to the increased steric demands of 219.

(i-Pr)3S! C

111

M03Si, I+ M3 hN

Ph Ph

219

CH 3CNreflux 18 h;

Si0 2

79%

0(i-Pr)3S1 NH

Ph

244

The cyclohexenyl-substituted ketene 110 also combines smoothly with imine 219

(refluxing acetonitrile, 24 h) to provide the desired lactam 245 in 66% yield (eq 73).

126. Chan, L.-H.; Rochow, E. G. J. Organomet. Chem. 1967, 9,-231.

85

(71)

(72)

(i-Pr)3 Si C 0

110

Me3SiN

Ph Ph

219

Silylketenimines were also studied as potential heterodienophiles in Diels-Alder

reactions with TAS-vinylketenes. Silylketenimines 247 and 249 were prepared by the method

of Watt. 127 Thus, treatment of the requisite nitrile with lithium diisopropylamide and

subsequent trapping with t-butyldimethylchlorosilane affords the ketenimine products in

excellent yields (Scheme 23). Unfortunately, reaction of these imines with ketene 111

afforded only a complex mixture of products.

Scheme 23

LDA, THF-78 *C;

t-BuMe 2SiCI-78 *C to rt 25 h

91%

Ph>

H3C Sit-BuMe2

247

G- C=N

248

LDA, THF-78 *C;

t-BuMe2SiCI '=NiSit-BuMe2-78 *C to rt 16 h

84% 249

(4) Reaction with Alkyl-Substituted N-(Silyl)Imines

In all successful cases thus far, the azadienophile component has had one or more

stabilizing phenyl substitutents. We next investigated the reaction of TAS-vinylketenes with

alkyl-subsituted N-silylimines, which are notably less stable than the aromatic-substituted

derivatives. In 1987, Cainelli reported the preparation of enolizable N-(trimethylsilyl)imines

127. Watt, D. Synth. Commun. 1974, 4, 127.

86

CH 3CNreflux 24 h;

SiO 2

66%

0(i-Pr)3S" NH

PhPh

245

(73)

Ph>

H3C

246

which must be generated and trapped in situ at temperatures of -30 *C or below."' For

example, addition of isobutyraldehyde to a -30 *C solution of LiHMDS generates N-

(trimethylsilyl)isobutyraldimine (250, eq 74).

SLiHMDS N.SiM0 3

THF -40 0CH H (74)

250

Attempts to use this procedure for in situ reactions with TAS-vinylketenes led to

disappointing results. Generation of the imine by the aforementioned procedure and

subsequent addition of ketene 111 at -78 0C followed by slow warming to room temperature

afforded a complex mixture of inseparable products. At low temperatures, the vinylketene

and imine probably do not undergo cycloaddition and as the temperature increases, the imine

likely isomerizes to the more stable enamine which may react with the ketene by alternative

pathways initiated by nucleophilic addition.

In contrast to this result, we have found that nonenolizable alkyl-substituted imines

serve as good partners for [4+2] cycloadditions with TAS-vinylketenes. Reaction of

pivaldehyde with LiHMDS at ambient temperature affords N-(trimethylsilyl)pivaldimine

251115 in 46-48% yield after distillation at reduced pressure. Treatment of ketene 110 with

imine 251 using our standard cycloaddition conditions (CH 3CN, reflux, 24 h) afforded none

of the desired lactam product. However, heating the reaction mixture in a sealed tube at 110

*C for 90 h did provide the desired cycloadduct in 56% yield (eq 75). This imine appears to

slowly decompose at these elevated temperatures, and therefore it was necessary to utilize

three equivalents of the dienophile which were added in two portions of 1.5 equiv each.

128. (a) Cainelli, G.; Giacomini, M.; Panunzio, M.; Martelli, G.; Spunta, G. Tetrahedron Lett. 1987, 28, 5369.

(b) Cainelli, G.; Giacomini, D.; Galletti, P.; Gaiba, A. Synlett 1996, 657.

87

(i-Pr)3Si C' 0

110

Me3Si'N

H t-Bu

251

0CH 3CNI

110C 90h (i-Pr)3 S* NHSiO2 -I8a~

t-Bu8.

56%252

Not surprisingly, the combination of hindered ketene 110 with the sterically

encumbered imine 251 results in formation of the trans-substituted product. This

stereochemistry was confirmed by an NOE study in which irradiation of the C-1 proton (Ha)

results in a 5% enhancement of one C-8 proton (He). No enhancement of the C-8a proton

(Hb) is observed, confirming the trans assignment (Figure 7).

Figure 7

5% Irradiate

H Ha Si(i-Pr)h

Hd H H0Hb H

t0% 252

(5) Reactions with N-Alkyl-Substituted Imines

. We next chose to study the cycloadditions of N-alkyl imines and TAS-vinylketenes.

This class of imines is generally prepared by addition of an amine to an aldehyde or ketone.

For example, treatment of benzaldehyde with methylamine in refluxing benzene with water

separation via a Dean-Stark trap provides N-methyl benzaldimine (253) in 89% yield (lit. 87-

95%, eq 76).129

129. Moffett, R. B. Organic Syntheses; Wiley: New York, 1963; Collect. Vol. IV, pp 605-608.

88

(75)

benzene

H2N-CH3 + 0 5 *C to reflux NCH3 (76)Ph H Dean-Stark Ph H

89%253

Reaction of ketene 111 with N-alkyl imine 253 does proceed successfully, but affords

an inseparable 3:1 mixture of diastereomeric lactams 254 and 255 in 71% yield (eq 73).

2530 CH 3CN00

(i-PrhSi C'' 120 0C 42 h (i-Pr)3S H3 (i-Pr)3SI CH36 N 6 W(77)

71% Ph Ph

111 254 3:1 255

SYBYL Calculations of the ground state energy of the cis and trans-substituted

lactams 254 and 255 were performed in an attempt to determine which conformational isomer

has a lower ground state energy for each product. For the modeling studies, the

triisopropylsilyl group of lactams 254 and 255 was replaced with a proton in order to simplify

the calculations and because the silyl group probably does not significantly influence the

difference in ground state energies of lactams 254 and 255. These molecular modeling

studies indicate that for the cis-substituted lactam, it is conformer 255a (with the phenyl

group in the equatorial position) which is the thermodynamically most stable conformation.

The difference in energy between 255a and 255b (AG = 1.46 kcal/mol) should result in

approximately a 90:10 ratio of conformational isomers. The trans-substituted lactam 254

slightly favors a conformation in which the phenyl group adopts an equatorial position (254a).

The small difference in energy between 254a and 254b (AG = 0.388 kcal/mol) indicates that

compound 254 exists as approximately a 2:1 ratio of conformational isomers 254a and 254b.

89

H3C Ha Si(i-Pr)3H3C N>

I Ph *CH3Hb P

4.205 kcal/mol

254a

H3C C H3 (-P)

Hb I rP*

Ph

majorcycloadduct

H3 C Ha S1(1-Pr) 3

Hb N

I3 Ph CH3

3.451 kcal/mol

255a

minorcycloadduct

4.593 kcal/mol

254b

Hb IHHCH Si(i-Pr)h

Ph H

4.914 kcal/mol

255b

The identity of the major cycloadduct was confirmed by an NOE study (Figure 8).

Irradiation of the C-5 methyl group of the major isomer (254, favored conformation shown)

results in a 5% enhancement of the proton at C-6, confirming that these two groups reside on

the same side of the ring (trans-substituted). However, irradiation of the C-6 proton (Ha) of

the minor isomer gives a 7% enhancement of the C-5 proton (Hb). Thus, the minor isomer

does in fact have both protons on the same side of the ring (cis-substituted). Although no

enhancement of Ha is expected upon irradiation of the C-5 methyl group, a 3% enhancement

is in fact observed, but this value is not considered to be a significant NOE enhancement.

Figure 8

Irradiate

7% 4H3C Ha Si(i-Pr)3

Hb? N, 0

HC Ph CH3

minor isomer

255

3%

1H3C Ha SI(i-Pr)3

Hb? N,C6

HCPh CH3HC0

IIrradiate

90

5%

Irradiate 4H3C a Si(i-Pr)h

H3 C-tN.wPh CH 3

Hbmajor isomer

254

In order to determine experimentally which isomer is the thermodynamically favored

product, we investigated the base-induced isomerization of the mixture of products. Thus,

treatment of the 75:25 (trans:cis) mixture of isomers 254 and 255 with potassium t-butoxide

in t-butanol results in isomerization to provide predominately the cis-substituted isomer in a

72:28 cis:trans ratio (eq 78), indicating that the cis isomer is the thermodynamically favored

product in agreement with our molecular modeling calculations. In fact, based on the

calculated difference in energies between 254a and 255a (AG = 0.754) we would predict to

observe a similar ratio of approximately 75:25.

0 0(i-Pr)3Si ,CH 3 (i-Pr)3Si ,CH 3 KOt-Bu

N + N t-BuOH5II 5 6 W 254 + 255 (78)

Ph rt 18h

28: 72254 255

75:25

Unfortunately, other more hindered alkyl-substituted imines did not participate in

[4+2] cycloadditions with TAS-vinylketenes. For example, reaction of ketene 111 with

enantiomerically enriched imine 258 (prepared from benzaldehyde and (S)-a-methyl-

benzylamine) at 125 to 150 *C afforded mostly unreacted ketene and imine starting materials.

None of the desired cycloadduct was observed by 'H NMR. Reaction of 2-phenyl-1-pyrroline

2590 with ketene 111 (CH 3CN, 110 *C, 35 h) also gave disappointing results, providing a

complex mixture of products.

N Ph

Ph H Ph

258 259

130. A sample of this imine was kindly provided by Marcus Hansen. For details on the preparation of thiscompound, see: Sorgi, K. L.; Maryanoff, C. A.; McComsey, D. F.; Maryanoff, B. E. Organic Syntheses

1997, 75, pp 215-222.

91

(6) Cycloadditions with Other Azadienophiles

In the next phase of our study, we turned our attention to investigating electron-rich

and electron-deficient imine derivatives as partners in Diels-Alder reactions with TAS-

vinylketenes. First, electron-rich heteroatom-substituted imines such as amidines were

studied. Visiting scientist Iwao Okamoto examined the [4+2] cycloadditions of these

nucleophilic imines without success. For example, treatment of ketene 111 with amidine 260

at room temperature resulted in some disappearance of the starting material but no

cycloadduct could be observed by 'H NMR (eq 79). Preliminary experiments have been

carried out to survey the possible utility of other azadienophiles, however these compounds

(shown in Figure 9) also resulted in no reaction with ketene 111.

.0 0(i-Pr)3SI C' Me (i-Pr)3SI ,Me

+N (79)

Me 2N H

111 260 261

Figure 90

N'NM2 N Ph N11i1

Ph H Me 2N H 0

262 263 264

Electron-deficient imine derivatives also do not participate in [4+2] cycloadditions

with TAS-vinylketenes. Addition of glyoxylate 268 to ketene 111 in refluxing acetonitrile for

22 h afforded none of the desired product. Preparation of this imine was accomplished by a

two-step procedure as described by Colvin (Scheme 24).12' First, silylation of glycine methyl

ester hydrochloride by treatment with base and then a silyl chloride affords the N-silylamine

266. Subsequent chlorination of 266 with t-butyl hypochlorite followed by base induced

elimination forms the imine 268.

92

Scheme 24

1) Et 3N, DMAP

2) t-BuMe 2SiC

26%

,SiMe 2t-BuN DBU

MeO 2C H 77%

268

H% NSiMe 2t-Bu

'Co2Me

266

'It-BuOCI

C1. ,SiMe 2 t-BuN

CO2Me

267

Mechanism of Imino Diels-Alder Reactions

What is the mechanism of the Diels-Alder reaction of TAS-vinylketenes and imines?

One can envision alternate pathways ranging from a Diels-Alder concerted mechanism to a

true stepwise process, as illustrated in Scheme 25. In the stepwise process, the imine would

first add to the ketene to afford an acyclic intermediate that would then undergo electrocyclic

closure to afford the lactam.

Scheme 25

[R3SI Ci

+ N

RASI G..0R'N R

Concerted

R a RR3SI 8+/R

Asynchronousconcerted

R331 R

Np

L Stepwise

IR3S N 0R

I93

H eNH3CI

CO2Me

265

]

Marson has reported a convenient method for the formation of the 5,6-dihydro-2(1IH)-

pyridone ring system by a [5+1] condensation of a 3-alkenamide with an aldehyde or ketone

that occurs via an intermediate similar to the one proposed above for the stepwise mechanism

for the [4+2] cycloaddition. 3 1 For example, amide 269 and benzaldehyde combine in the

presence of polyphosphoric ester (PPE) to form intermediate 270 which cyclizes to afford

exclusively the trans-substituted lactam 235 in 63% yield (eq 80). Recall that our related

[4+2] cycloaddition of ketene 111 with imine 215 affords only the cis-substituted product

233.

PPE [ O-P 0

NH2 + 0 35*C 24h IrO H (80)

H .1k Ph Ph - :Ph

269 270 235

With respect to the stepwise mechanism (which would involve nucleophilic attack of

the imine on the ketene), it is important to note that amines are much more nucleophilic than

imines. In order to examine whether a stepwise mechanism is feasible, we first studied the

nucleophilic addition of amines to (TAS)ketenes. We found that diethylamine adds rapidly to

(triisopropylsilyl)ketene at room temperature (CH 3CN, 20 min). TAS-Vinylketenes also react

quickly with diethylamine, however the rate of this addition is slower than the addition of

amine with TAS-ketenes. For example, TAS-vinylketene 112 reacts with diethylamine in

acetonitrile when heated at 50 *C for 20 min to afford a mixture of isomers (271-273) that

result from nucleophilic addition of the amine to the TAS-vinylketene (eq 81). Because

nucleophilic addition to TAS-vinylketenes does occur under conditions comparable to those

employed in our imine cycloadditions, we cannot immediately rule out the possibility of a

stepwise mechanism in the [4+2] cycloaddition in which the imine would act as a nucleophile

and attack the ketene, followed by cyclization in a second step.

131. Marson, C. M.; Grabowska, U.; Fallah, A. J. Org. Chem. 1994, 59, 291.

94

Et 2 NH 0

______ Et3SI NEt 2 +CH 3CN +

50 *C 20min

112 271

0Et3SI N R2

272

0Et3S1 NEt2 (81)

273

Also relevant to the mechanism question are the [2+2] cycloadditions of ketenes with

imines. It is well-established that alkyl- and aryl-ketenes react with imines to form P-lactams

in a process commonly known as the Staudinger reaction.132 In a typical example, acid

chloride 274 undergoes in situ dehydrohalogenation to form methylketene which

subsequently reacts with imine 275 to afford the cis-substituted P-lactam 276 in good yield

(eq 82).133 The stereochemical outcome of these [2+2] cycloadditions is sometimes difficult

to predict and depends greatly on the structure of the imine and ketene, the solvent, base,

mode of ketene generation, and temperature as well as other factors.

H COPh

WA

Et 3N

80-90%

275

COPh

0N0 R e

276

The Bose reaction, a variation of the Staudinger reaction in which an aminoketene

reacts with an imine, affords 3-amino-substituted P-lactams. For example, combination of an

amino-substituted ketene (generated in situ from anhydride 277) reacts with imine 278 in a

[2+2] cycloaddition to form 3-amino-2-azetidinone 280 after hydrolysis (eq 83).34

132. For a review of the formation of P-lactams by [2+2] cycloaddition, see: (a) The Organic Chemistry of D-Lactams; Georg, G. I., Ed.; VCH: New York, 1993; Ch. 6, pp 295-368. (b) Dugat, D.; Just, G.; Sahoo, S.Can. J. Chem. 1987, 65, 88.

133. (a) Paloma, C.; Aizpura, J. M.; Lopez, M. C.; Aurrekoetxea, N.; Oiarbide, M. Tetrahedron Lett. 1990, 31,6425. (b) Paloma, C.; Ontoria, J. M.; Odriozola, J. M.; Aizpurua, J. M.; Ganboa, I. J. Chem. Soc., Chem.Commun. 1986, 161.

134. Sain, B.; Baruah, J. N.; Sandhu, J. S. J. Heterocyci. Chem. 1984, 21, 257.

95

Et 3Si C-

0 cl

274

(82)

0OOEt Ph H RHN Ph H21 PhRHN + Oyt P EtN PTSA (83)

0 0 N N0 NDMB 0 DMB DMB

277 278 279 280R = -C(CH)CHCO(OEt)

The mechanism of these important [2+2] cycloadditions has been the subject of

intense scrutiny and the reactions are generally believed to proceed via a stepwise mechanism.

In fact, experiments have been reported in which the intermediate zwitterion was trapped with

ethanol. 35 The substitutents on the ketene presumably determine the direction of attack by

the imine based on steric considerations. Because the trans conformation of the imine is

preferred, formation of the zwitterionic intermediate and subsequent conrotatory ring closure

generally affords the cis-substituted P-lactams.

As mentioned previously, [2+2] cycloadditions of ketenes and N-

(trimethylsilyl)imines are known to proceed via a stepwise mechanism because often the

acyclic intermediate can be observed spectroscopically or, in some cases, isolated. In these

examples, addition of a TAS-ketene to an imine first generates a characterizable acyclic

intermediate that, upon heating, cyclizes to an azetidinone. Surprisingly, no reactions of

(trialkylsilyl)ketenes with N-(trimethylsilyl)imines have been reported to date. We believed

that the rate of such a reaction would provide an important value for comparison with our

[4+2] cycloadditions of (trialkylsilyl)vinylketenes with N-(trimethylsilyl)imines. In fact, we

found that heating (triisopropylsilyl)ketene (281) with N-(trimethylsilyl)benzaldimine 215 at

70 'C resulted in slow reaction and was only approximately 30% complete (by TLC analysis)

after 3 days. However, when the reaction temperature was increased to 140 'C for 3 h, we

were able to isolate the [2+2] adduct 283 in 72% yield (eq 84).

2150 no solvent 0 SiMes3C1 140 *C 3 ha (O %+,SiMe3 N (4

A~ N(i-Pr)3S1 H 72% (i-Pr)hSi Ph (i-Pr)3S' Ph

281 282 283

96

Presumably, this [2+2] reaction proceeds by a stepwise mechanism as do known

reactions of TAS-imines with ketenes. The initially formed trans, cis-zwitterionic compound

isomerizes to zwitterionic intermediate 282 which then undergoes conrototory ring closure to

afford the trans-substituted P-lactam 283. Since the conrotatory ring closure in this

cycloaddition affords the trans-substituted P-lactam, we would expect (and in fact do observe)

formation of the cis-substituted product from the disrotatory ring closure in the [4+2]

cycloaddition of ketene 111 with imine 215 (eq 65).

Because the [4+2] cycloadditions of imines with TAS-vinylketenes are much faster

than the related stepwise [2+2] cycloaddition of imines with TAS-ketenes it is possible that

the TAS-vinylketene reactions are proceeding via a completely different mechanism, a

concerted process. As noted earlier, the reactions proceed at a faster rate in polar than in non-

polar solvents, and thus the reaction would follow an asynchronous concerted mechanism in

which a significant separation of charge occurs in the transition state (Figure 10). This

charged intermediate would be stabilized by the presence of a polar solvent.

Figure 10

8 *0R3SI #"-SiM03

'N 8S

R

Further investigations are underway in our laboratory to determine the mechanism of

the [4+2] cycloaddition and to explain the interesting stereochemical observations made in the

studies described in this chapter.

Protodesilylation of Lactam Cycloadducts

Multiple attempts were made to protodesilylate the triisopropylsilyl-substituted

lactams 233 and 252 (Table 4). Although some of these conditions effectively removed the

triisopropylsilyl group, the desired product could not be isolated in >95% purity (entries 1, 5,

6).

135. Moore, H. W.; Hernandez, L.; Chambers, R. J. Am. Chem. Soc. 1978, 100, 2245.

97

Table 4

Entry

1

2

3

4

Table 5 details our attempts to protodesilylate the triethylsilyl-substituted lactams 239

and 236 which cleaved more readily than the triisopropylsilyl derivatives. Clean removal of

the triethylsilyl-group can be accomplished by treatment with 5 equiv of methanesulfonic acid

in refluxing dichloromethane (entries 3 and 4, eq 85).

Table 5 Protodesilylation of (Triethylsilyl)-Substituted Lactams

Entry Cycloadduct Conditions Result

1 239 20 equiv TFA 14% yield; 239 wasCH 2 C 2 , rt 25h recovered in 86% yield

2 239 5 equiv TBAF no reactionTHF, rt 3 days

3 239 5 equiv MsOH 55% yield of 245CH 2Cl 2 , reflux 4 h

4 236 5 equiv MsOH 83% yield of 285CH 2C 2, reflux 6 h

98

5

6

7

8

9

Protodesilylation of

Cycloadduct

233

233

233

233

233

233

252

252

252

(Triisopropylsilyl)-Substituted Lactams

Conditions Result

20 equiv TFA complete conversionCH 2C12, rt 20 h to 246, <95% purity

5 equiv TBAF complex mixtureTHF, reflux 1 h

1.1 equiv CsF no reactionCH 3CN, reflux 20 h

20 equiv MsOH no reactionMeOH, reflux 4 h

20 equiv MsOH complete conversionCH 2Cl 2, reflux 1.5 h to 246, <95% purity

5 equiv MsOH complete conversionCH 2Cl 2, reflux 17 h to 246, <95% purity

10% aq. HCl no reactionTHF, rt 2 h

oxalic acid no reactionMeOH, rt 22 h

1.1 AlCl 3 no reactionCH 2C 2, rt

MsOHCH2C12

Et 3Si NH reflux 4 h HI NHPh Ph 55% Ph Ph

239 284(85)

o MsOHEt 3Si CH 2C12

NH reflux 6 h NH

Ph 83% Ph

236 285

Conclusion

We have shown that TAS-vinylketenes participate in oxa and imino Diels-Alder

reactions with a variety of heterodienophiles. As shown in previous work, the ketene

carbonyl dominates in controlling the regiochemical course of these reactions. Particularly

interesting is the stereochemical course of these cycloadditions: with less hindered dienes the

endo product is obtained and with more bulky vinylketenes the exo product forms.

Investigations are ongoing to determine whether the mechanism of the imino [4+2]

cycloaddions proceeds via a stepwise mechanism or an asynchronous concerted transition

state which is stabilized in the presence of a polar solvent. Overall, we have shown that TAS-

vinylketenes function as interesting four-carbon dienes in Diels-Alder cycloadditions to afford

X, P-unsaturated 6-valerolactones and -lactams.

99

CHAPTER 5

TAS-VINYLKETENES AS FOUR-CARBON COMPONENTS

IN A NEW [4+1] ANNULATION STRATEGY

Introduction

The invention of new methods for the construction of five-membered rings continues

to be a problem of considerable importance in organic synthesis.136 Many biologically active

natural products incorporate a cyclopentenone ring as a major structural feature and

cyclopentyl systems serve as valuable synthetic intermediates in routes to many classes of

compounds. Five-membered rings can be synthesized by a variety of different methods,

including the Nazarov cyclization,'3 7 the Pauson-Khand cyclization of olefins with alkynes,13 8

[3+2] coupling reactions,139 the [3+2] annulation,'40 intramolecular aldol condensations, and

vinylcarbene insertion into C-H bonds.14' Only a few [4+1] annulation strategies have been

reported to date, one example being the methodology base on oxyanion and carbanion-

accelerated vinylcyclopropane rearrangements developed in our laboratory.142 43 This

chapter outlines our work on the stereoselective [4+1] annulation strategy for the synthesis of

136. For reviews, see: (a) Paquette, L. A. Top. Curr. Chem. 1984, 199, 1. (b) Ramaiah, M. Synthesis 1984,529. (c) Hudlicky, T.; Price, J. D. Chem. Rev. 1989, 89, 1467.

137. (a) Denmark, S. E. In Comprehensive Organic Synthesis; Trost, B.M.; Fleming, I., Eds.; Pergamon Press:New York, 1990; Vol. 5, p 751. (b) Habermas, K. L.; Denmark, S. E. In Organic Reactions; Paquette, L.A., Ed.; Wiley: New York, 1994; Vol. 45, pp 1-158.

138. (a) Schore, N. E. In Comprehensive Organic Synthesis; Trost, B. M.; Fleming, I., Eds.; Pergamon Press:New York, 1990; Vol. 5, p 1037. (b) Schore, N. E. In Organic Reactions; Paquette, L. A., Ed., Wiley:New York, 1991; Vol. 40, pp 1-90.

139. (a) Noyori, R.; Shimizu, F.; Fukuta, K.; Takaya, H.; Hayakawa, Y. J. Am. Chem. Soc. 1977, 99, 5196. (b)Noyori, R.; Yakoyama, K.; Hayakawa, Y. J. Am. Chem. Soc. 1973, 95, 2722.

140. (a) Danheiser, R. L.; Dixon, B. R.; Gleason, R. W. J. Org. Chem. 1992, 57, 6094. (b) Danheiser, R. L.;Carini, D. J.; Basak, A. J. Am. Chem. Soc. 1981, 103, 1604. (c) Danheiser, R. L.; Carini, D. J.; Fink, D.M.; Basak, A. Tetrahedron 1983, 39,935. (d) Danheiser, R. L.; Fink, D. M. Tetrahedron Lett. 1985, 26,2513. (e) Danheiser, R. L.; Takahashi, T.; Bert6k, B.; Dixon, B. R. Tetrahedron Lett. 1993, 34, 3845.

141. (a) Karpf, M.; Huguet, J.; Dreiding, A. S. Helv. Chim. Acta 1982, 65, 13. (b) Williamson, B. L.;Tykwinski, R. R.; Stang, P. J. Am. Chem. Soc. 1994, 116, 93.

142. Reviewed in: Bronson, J. J.; Danheiser, R. L. Charge-Accelerated Small Ring Rearrangements. InComprehensive Organic Synthesis; Paquette, L. A., Ed.; Pergamon Press: Oxford, U. K., 1991; Vol. 5, pp999-1035. A related intramolecular [4+1] annulation based on thermal vinylcyclopropane rearrangementshas been developed by Hudlicky; see: Hudlicky, T.; Reed, J. W. Rearrangements of Vinylcyclopropanesand Related Systems. In Comprehensive Organic Synthesis; Paquette, L. A., Ed.; Pergamon Press:Oxford, U. K., 1991; Vol. 5, pp 899-970.

143. Rigby recently reported a [4+1] heteroannulation based on the addition of dimethoxycarbene to vinylisocyanates; see: Rigby, J. H.; Cavezza, A.; Ahmed, G. J. Am. Chem. Soc. 1996, 118, 12848.

100

substituted cyclopentenones based on the reaction of TAS-vinylketenes with nucleophilic

"carbenoid" reagents such as sulfur ylides and diazo compounds (eq 86).144

40 0R3Si C"0 *LRSW

RI + LR3 RSK? 440 (86)

2 R' R2

(L = leaving group)

[4+11 Annulations with Sulfur Ylides and Diazo Compounds

Our plan for the application of TAS-vinylketenes to the synthesis of five-membered

rings called for their reaction with nucleophilic carbenoid reagents to generate dienolate

species that would then cyclize to the desired cyclopentenones. Several classes of carbanionic

nucleophiles are known to add smoothly to simple silylketenes (including

(trimethylsilyl)ketene itself), 145 and we therefore anticipated that TAS-vinylketenes would

participate in the proposed [4+1] annulation provided that carbenoid reagents of suitable

reactivity could be identified.

Among the several classes of carbenoid reagents we have examined, sulfur ylides146

and diazo compounds have thus far proved most effective for the desired [4+1] annulation.

Concurrently with our studies, Tidwell has found that diazo compounds add to silylated

bisketenes in a related process to produce cyclopentene-1,3-diones and 5-methylene-2(5H)-

furanones.147 As illustrated in eq 87, reaction of bisketene 286 with diazomethane affords a

mixture of dione 287 (22%) and furanone 288 (63%). Reactions of bisketene 286 with

(trimethylsilyl)diazomethane and phenyldiazomethane (PhCHN2) provide only substituted

cyclopentene-1,3-dione products.

144. Loebach, J. L.; Bennett, D. M.; Danheiser, R. L. J. Am. Chem. Soc. 1998, 120, 9690.145. For examples, see refs 31, 52, and 55.146. Reviewed in: Trost, B. M.; Melvin, L. S. Sulfur Ylides; Academic Press: New York, 1975.147. Colomvakos, J. D.; Egle, I.; Ma, J.; Pole, D. L.; Tidwell, T. T.; Warkentin, J. J. Org. Chem. 1996, 61,

9522.

101

Me 3Si C-O

Me3SXC"O%

CH 2N2 _

Et2Ort 2 h

286

0

Me 3Si

Me 3Si 0

28722%

+ M 3Si

Me 3Si CH 2

28863%

Jennifer Loebach of our laboratory initially investigated the new [4+1] annulation

strategy of TAS-vinylketenes with sulfur ylides and diazomethane compounds (Table 6). In

a typical reaction, addition of 1.05 equiv of dimethylsulfonium methylide to ketene 111 in 1:1

THF-DMSO at 0-25 *C for 1.5 h gives the desired cyclopentenone 289 in 75% yield after

chromatographic purification (entry 1). Dimethyloxosulfonium methylide also combines with

111 to produce cyclopentenone 289, though in lower yield (entry 2). Diazomethane reacts

with TAS-vinylketene 111 in a similar fashion to afford the cyclopentenone product in

excellent yield (entry 3). Protodesilylation of annulation product 289 was readily achieved by

exposure to methanesulfonic acid in methanol at 25 *C for 3 h to afford 3,4-

dimethylcyclopentenone in 95% yield (eq 88).

Table 6

(i-Pr)3S j Cr1 0S(P-Pr)3Si

111 289

Carbenoidreagent

G)i G

Me 2S- CH 2

CH 2N2

Conditions

1:1 THF-DMSO0 to 25 *C 1.5 h

1:1 THF-DMSO0 to 25 *C 1.5 h

CH2C3-120 to 25 C 3 h

Yield

75%

43%

96%

102

(87)

Entry

1

2

3

0 20 equiv MsOH 0(i-Pr)3Si MeOH rt 3 h (88)

95%

289 290

Optimizing conditions for the addition of the analogous isopropyl sulfur ylide proved

to be nontrivial. Initially, reaction of diphenylsulfonium isopropylide (291) with ketene 111

was attempted using our standard conditions (0 *C to rt, 1 h) and resulted in formation of the

desired product, but the cyclopentenone 292 could not be isolated in >95% purity. Attempts

to protodesilylate the crude product and then isolate and purify the resulting 3,4,5,5-

tetramethylcyclopentenone also were unsuccessful. Effective removal of the triisopropylsilyl

group from cyclopentenone 292 occurs in the presence of 5 equiv of TBAF (refluxing THF, 3

h) but, unfortunately, the desired product is not stable to silica gel purification and also could

not be isolated in >95% purity. Therefore, we returned to the [4+1] annulation step in an

attempt to obtain pure cyclopentenone 292 directly. It occurred to us that the isopropylide

291 is probably less stable than the methylide derivative previously studied (Table 7, entry 1).

In fact, when ylide 291 was generated at a lower temperature (-20 *C versus 0 *C) smooth

reaction with TAS-vinylketene 111 ensued to afford the desired cyclopentenone 292 in 57%

yield after chromatographic purification (eq 89).

G 0i0 Ph 2S-CMe2 291

THF (i-Pr)3S! (89)-20 *C to rt

57%111 292

Diphenylisopropyl sulfonium fluoroborate (the precursor to ylide 291) was prepared

by the method of Nadeau. Reaction of diphenyl sulfide and isopropyl iodide in the presence

of silver fluoroborate (0 *C to rt, 2.5 h) afforded 18-31% yield (lit. 55%) of the ylide

precursor 295 as a crystalline white solid.148 Treatment of this sulfonium salt with t-BuLi in a

148. Nadeau, R. G.; Hanzlik, R. P. Methods Enzymol. 1969, 15, 347.

103

50% THF-DMSO solution at -20 *C provides the bright yellow ylide 291 which must be used

immediately to avoid decomposition (eq 90).

GBF4/ AgBF 4 P / n-BuLi PG(

Ph2S + I-( - 1 Ph 2 S ~ Ph2S (90)0 Ctort THF -20 *C

293 294 18-31% 295 296

Commercially available (trimethylsilyl)diazomethane1 49 also reacts with TAS-

vinylketene 111 to afford the trans-substituted cyclopentenone 296 (eq 91). However, all

attempts to employ higher diazoalkanes and substituted TMS-diazomethanes in the [4+1]

annulation have been unsuccessful; in each case no significant reaction occurred and the TAS-

vinylketene was recovered unchanged. Fortunately, substituted sulfur ylides are more

nucleophilic and do react with TAS-vinylketenes in the desired fashion, providing access to

highly substituted cyclopentenones in good yield.

00 M03SiCHN2 0(i-Pr)3Si C CH 2C2 25C (i-Pr)3S1 .K SiMe3 (91)

32 h

84-92%111 296

Scheme 26 outlines several alternative pathways to account for the mechanistic course

of our [4+1] annulation. Addition of a carbenoid reagent to the vinylketene should be highly

stereoselective due to the shielding effect of the bulky trialkylsilyl group and should result in

the formation of the (Z)-enolate 297. Cyclization of this intermediate could produce the five-

membered ring product directly, although the planar structure of the dienolate system in 297

may not allow it to achieve an arrangement in which the R electrons are suitably situated for

direct backside displacement of the leaving group. An alternative pathway involves

ionization to produce the 2-oxidopentadienylic cation 298, which can then undergo

149. (a) Seyferth, D.; Dow, A. W.; Menzel, H.; Flood, T. C. J. Am. Chem. Soc. 1968, 90, 1080. (b) Shioiri, T.;Aoyama, T.; Mori, S. Organic Syntheses; Wiley: New York, 1993; Collect. Vol. 8, pp 612-615.

104

conrotatory 4n electrocyclic closure to generate the cyclopentenone product. Pentadienyl

cation electrocyclic ring closures are involved in the mechanism of the Nazarov cyclization,150

and epoxidation of vinylallenes produces cyclopentenones via electrocyclization of 2-

oxidopentadienylic cations analogous to 298.1 A similar mechanism is believed to occur in

the biosynthetic pathway for prostanoid synthesis by marine organisms in which conversion

of an allene oxide intermediate (300) to the cyclopentenone product (302) occurs via a 4n

electrocyclic closure of pentadienyl cation 301 (Scheme 27). A third pathway, proceeding via

the cyclopropanone intermediate 299 (either formed by direct internal displacement or

ionization and subsequent electrocyclic closure), cannot be excluded, particularly in view of

the finding that diazomethane adds to (trimethylsilyl)ketene to generate (trimethylsilyl)-

cyclopropanone in good yield.70

Scheme 26

internaldisplacement

0G

R3Si C-0-7 CH2L -3' I *

C297

ionization

0

299

ectrocyclic)ening

0

R3Si

298

[1,31sigmatropic

rearrangement

4;relectrocyclic

closure

direct internal displacement

150. For a review, see ref 137b.151. Kim, S. J.; Cha, J. K. Tetrahedron Lett. 1988, 29, 5613 and references therein.

105

R3SI

Scheme 27

OP0 CO 2H CO 2H

N.n-C 5H11 . - n-C5H11

300 301

0_

302 - C02Hpreclavuone A -

n-C5 H11

A notable feature of annulations involving substituted carbenoid reagents is the

exclusive formation of trans-4,5-substituted cyclopentenones. The stereochemical outcome of

the reactions of substituted sulfur ylides and diazo compounds is consistent with a mechanism

involving stereospecific conrotatory electrocyclic ring closure if one assumes that ionization

of the initial dienolate intermediate occurs to generate a 2-oxidopentadienylic cation 298 with

the C-1 substitutent cis to the oxygen atom to minimize nonbonded interactions. If a

mechanism involving concerted electrocyclization is indeed operative, then [4+1] annulations

beginning with TAS-vinylketenes with Z-substituted alkenyl groups should afford cis-4,5-

substituted cyclopentenones. Studies are underway in our laboratory to test this hypothesis.

Loebach found that both a-halocarbanions and "nitrenoid" equivalents (such as mesyl

azide) were ineffective as reagents for [4+1] annulations with TAS-vinylketenes. The

following sections describe other potential carbenoid reagents we explored to expand the

utility of this interesting annulation process.

[4+1] Annulation with Acetylides

In 1994, Jacobi reported the synthesis of methylenecyclopentenones in three steps

from enones via cyclization of an intermediate enynone with catalysis by c-tocopherol

106

(vitamin E).15 2 For example, enynone 304 (prepared from 303 as outlined in Scheme 28)

underwent cyclization to methylenecyclopentenone 306 upon heating at 125 *C in the

presence of a catalytic amount of Vitamin E. The enynones do not undergo cyclization in the

absence of a catalyst and the mechanism of the reaction is believed to proceed by a reductive

radical cyclization.

Scheme 28

200 *C 12 hvitamin E

55%

306

methylenomycin

OH

305

0

We hoped that acetylides might function as carbenoid reagents in [4+1] annulations

with TAS-vinylketenes to afford methylenecyclopentenones (309) in a single step.

Nucleophilic addition of an acetylide to the ketene carbonyl could afford intermediate 308

which might then undergo cyclization to the cyclopentenone 309 (eq 92).

R3Si , "

307

Li

R 3Si

R308

0

R3Si

309

(92)

Thus, treatment of ketene 111 with phenylacetylide 310 at -78 *C followed by slow

warming to -5 *C, addition of 1 equiv of BHT, and then heating at 50 'C for 35 min affords

152. (a) Jacobi, P. A.; Briehnann, H. L.; Cann, R. 0. J. Org. Chem. 1994, 59, 5305. (b) Jacobi, P. A.;Armacost, L. M.; Brielmann, H. L.; Cann, R. 0.; Kravitz, J. I.; Martinelli, M. J. J. Org. Chem. 1994 59,5292.

107

0

H

303

1) Li --

2) MnO 2

40%

0

304I

MeH OI M O

Me~ 0Me 3

vitamin E

the cyclopentenone 311 in 69-75% yield after chromatographic purification (eq 93).

Unfortunately, without addition of BHT the reaction gives only poor yields of the desired

cycloadduct. Therefore, the reaction probably proceeds via a radical mechanism (as described

by Jacobi) rather than by a carbanionic sequence of addition/cyclization in which the acetylide

functions as a carbenoid reagent. Other acetylides (such as trimethylsilylacetylide and -

C=CH) add to TAS-vinylketene 111, but only uncyclized addition products were observed to

form in these reactions by 1H NMR spectroscopy.

(i-Pr)3Si C"0 -78 *C to rt 0

+ ( Ph 2 h a (i-Pr)3 Si -- Ph (93)BHT

69-75%111 310 311

Samarium Iodide-Mediated [4+1] Annulations

Previous work by Loebach established that a-halocarbanions were ineffective as

carbenoid reagents in the [4+1] annulation. Recently, Inanaga reported that coupling of

carbonyl compounds with diiodomethane aided by SmI2 affords iodohydrins. For example,

reaction of benzylacetone (312) with diiodomethane in the presence of SmI 2 (room

temperature, 3 min) affords the iodo alcohol 313 in 96% yield (eq 94).

0 SM12 OH

+ CH212 S Phd (94)THF rt CH 213 min

312 96% 313

We hoped to employ a similar procedure for a [4+1] annulation with TAS-

vinylketenes to afford cyclopentenones (eq 95). Unfortunately, however, no observable

reaction occurred upon treatment of a THF solution of ketene 111 and diiodomethane with

SmI2.

108

0 -0(i-PrhSi Cj CH2I (i-Pr)3 Se H

SM12 N 'H \ipr~ (95)

111 314 289

Synthetic Elaboration of Cyclopentenones

As mentioned previously, silylcyclopentenone 289 can be protodesilylated in the

presence of methanesulfonic acid. We were interested in investigating other possible

synthetic transformations of the silyl group such as conversion to a halide which would

provide access to useful synthetic intermediates for cross-coupling and other reactions.

Negishi previously reported that trimethylsilyl-substituted cyclopentenones can be

converted to the x-bromo enones by treatment with N-bromosuccinimide.15 For example,

treatment of bicycle 315 with 2.5 equiv of NBS in DMF affords the bromo-substituted

compound 316 in 75% yield (eq 96).

SIM03 2.5 equiv NBS Br

o DMF c o (96)

315 75% 316

We anticipated that this transformation would occur at an extremely slow rate for

triisopropylsilyl-substituted cyclopentenones such as 289 as a result of steric hindrance.

However, in preliminary studies we found that the less-hindered triethylsilyl-substituted

cyclopentenone 317 does undergo conversion to the bromo-substituted compound when

treated with 3.5 equiv of N-bromosuccinimide in DMF (room temperature, 54 h), affording

the cyclopentenone 318 in 51% yield (eq 97). Not surprisingly, reaction with this triethylsilyl

compound occurs at a significantly slower rate than with the less hindered triisopropylsilyl-

substituted cyclopentenone 289. Other conditions for more efficient conversion of our

annulation products to vinyl bromides and iodides are under investigation in our laboratory.

153. Negishi, E.; Holmes, S. J.; Tour, J. M.; Miller, J. A.; Cederbaum, F. E.; Swanson, D. R.; Takahashi, T. J.Am. Chem. Soc. 1989, 111, 3336.

109

Et 3Si

317

3.5 equiv NBSDMF

rt 40 h

55%

Confirmation of the structural assignment of bromo cyclopentenone 318 was made by

analogy to previously reported spectroscopic data. As expected, the triethylsilyl resonances

of 317 do not appear in the 1H NMR spectrum of 318. Also, the 13C NMR spectrum of bromo

cyclopentenone 318 exhibits similar resonances as reported for bromo cyclopentenone 316

(Scheme 29).

Scheme 29

136.0198.4

1ISiMe3

0

214.6315

136.6

190.6SIEt3

5=0

213.2

317

116.9

187.1

i Br

202.6

316

122.9177.3

Br

200.4

318

Vollhardt reported that iodine monochloride (IC) 15 4 effectively cleaves carbon-silicon

bonds to produce the corresponding iodo compound. For example, treatment of

trimethylsilyl-substituted 319 with 2 equiv of iodine monochloride in carbon tetrachloride

yields the diiodo derivative 320 in 94% yield (eq 98).155 Presumably, iodination of the

154. Brisbois, R. G.; Wanke, R. A. In Encyclopedia ofReagentsfor Organic Synthesis; Paquette, L. A., Ed.;Wiley: New York, 1995; pp 2 8 1 1-2 8 12 .

155. Hillard, R. L.; Vollhardt, P. C. J. Am. Chem. Soc. 1977, 99,4058.

110

0

Br

318

(97)

silicon-substituted double bond affords a P-silicon stabilized cation. Subsequent cleavage of

the silicon group by attack with chloride generates the iodo alkene 320.

2 equiv ICI

1 ,C 0 E (98)E ) SiM03 94%-

319 320

Preliminary small-scale studies for the iodination of triethylsilyl-substituted

cyclopentenones utilizing similar conditions gave promising results. Treatment of

cyclopentenone 317 with 2 equiv of iodine monochloride (room temperature, 40 h) affords the

iodocyclopentenone 321 in -55% yield (not optimized, eq 99).

0 2 equiv ICI 0Et3S 1CH 2C12 I

rt 40 h

317 321

We envision that TAS-cyclopentenones could also undergo another interesting

transformation, the Tamao/Fleming cleavage of an alkoxy-substituted silyl compound to

afford a hydroxyl group.156 Organosilyl groups bearing at least one electronegative

substitutent have been widely used as synthetic equivalents of the hydroxy group. However,

successful conversion of a vinylsilane to the enol occurs only with specifically substituted

silicon groups. Currently, we are working to develop synthetic routes to TAS-vinylketenes

that contain the proper type of silyl group for this transformation. This methodology could

potentially be applied to the total synthesis of the natural product terpestacin (323) which

exhibits potential anti HIV activity (eq 100).157

156. For reviews on the use of silyl groups as synthetic equivalents of hydroxy groups, see: (a) Jones, G. R.;Landais, Y. Tetrahedron 1996,52, 7599. (b) Fleming, I. Chemtracts - Organic Chem. 1996, 9, 1.

157. Oka, M.; limura, S.; Narita, Y.; Furumai, T.; Konishi, M.; Oki, T. J. Org. Chem. 1993, 58, 1875.

ill

0 0

XR2Si HO

------------ ,,- (100)HO / HO

OH OH

322 323X = OR, F, CI terpestacin

Summary

We have described herein new, efficient routes to substituted TAS-vinylketenes via

the electrocyclic ring opening of cyclobutenones and the photochemical Wolff rearrangement

of a-silyl-a-diazo enones. TAS-Vinylketenes are reactive as four-carbon building blocks in

organic synthesis and behave as dienes in Diels-Alder reactions with reactive dienophiles,

activated carbonyl compounds, and imines. These ketenes also participate in [4+1]

cycloadditions with carbenoid reagents to form cyclopentenones. We are continuing to

explore promising new synthetic applications of TAS-vinylketenes for the total synthesis of

natural products and investigations are ongoing in our laboratory to develop other new,

efficient routes to these useful synthetic intermediates.

112

PART II

Experimental Section

113

General Procedures

All reactions were performed in flame-dried glassware under a positive pressure of

nitrogen or argon. Reactions were stirred magnetically unless otherwise indicated except

sealed tube reactions, which were not stirred. Air- and moisture-sensitive liquids and

solutions were transferred via syringe or cannula into reaction vessels through rubber septa.

Reaction product solutions and chromatography fractions were concentrated by using a Btichi

evaporator at ca. 20 mmHg unless otherwise indicated.

Materials

Commercial grade reagents and solvents were used without further purification except

as indicated below.

(a) Distilled under argon or vacuum from calcium hydride:

acetonitrile, benzene, dichloromethane, DMSO, diisopropylethylamine, hexane,

toluene, and triethylamine

(b) Distilled under argon or vacuum from sodium benzophenone ketyl or dianion:

tetrahydrofuran and diethyl ether

(c) Distilled under argon

trichloroacetyl chloride

(d) Other

N-Bromosuccinimide was recrystallized from water. 157 Alkyllithium reagents were

titrated in tetrahydrofuran or hexane at 0 *C using 1-10-phenanthroline as an indicator.15

Chromatography

(a) Analytical thin-layer chromatography (TLC)

Analytical thin-layer chromatography (TLC) was performed on Merck pre-coated

glass-backed 0.25 mm silica gel 60-F-254 plates. Visualization of spots was effected by one

or more of the following techniques: (a) ultraviolet irradiation, (b) exposure to iodine vapor,

157. Virgil, S. C. In Encyclopedia ofReagentsfor Organic Synthesis; Paquette, L. A.; Wiley: Chichester,1995, Vol. 1; pp 768-773.

158. Inorganic Synthesis 1963, 7, 10.

114

and (c) immersion of the plate in a 10% solution of phosphomolybdic acid in ethanol followed

by heating at ca. 200 'C.

(b) Column chromatography

Column chromatography was performed on ICN silica gel (32-60ptm).

Instrumentation

(a) Melting points

Melting points (mp) were determined with a Fischer-Johns melting point apparatus

and are uncorrected.

(b) Spectrometry

H NMR spectra were measured with Varian XL-300 (300 MHz), Unity-300 (300

MHz), and Unity-500 (500 MHz) instruments. Chemical shifts are expressed in parts per

million (6) downfield from tetramethylsilane (with the CHC13 peak at 7.26 ppm used as a

standard). 13 C NMR spectra were measured with Varian XL-300 (75 MHz), Unity-300 (75

MHz), and Unity-500 (125 MHz) spectometers. Chemical shifts are expressed in parts per

million (d) downfield from tetramethylsilane (with the central peak of CHC13 at 77.0 ppm

used as a standard). Infrared spectra (IR) were obtained using a Perkin Elmer 1320 grating

spectrophotometer.

(c) Elemental analyses

Robertson Laboratory, Inc. of Madison, New Jersey and E+R Microanalytical

Latoratory, Inc. of Parsippany, New Jersey performed elemental anlayses.

115

0

98 N2

98

0N2

SiEt2Ph

105

1-Diazo-1-(diethyl(phenyl)silyl)-3-methyl-3-penten-2-one (105).

A three-necked, 15-mL, round-bottomed flask equipped with a rubber septum, argon

inlet adapter, and glass stopper was charged with diazoketone 98 (0.052 g, 0.419 mmol), 0.8

mL of ether, and 0.8 mL of hexane then cooled at 0 *C. Diisopropylethylamine (0.073 mL,

0.419 mmol) was added followed by diethylphenylsilyl trifluoromethanesulfonate 104 (0.130

g, 0.416 mmol). The resulting solution was stirred for 30 min and then allowed to warm to

room temperature over 30 min. The reaction mixture was filtered through Celite with the aid

of 5 mL of ether and concentrated at reduced pressure to give 0.129 g of a red-orange oil.

Column chromatography on 3 g of silica gel (elution with 0-2.5% EtOAc-hexane) afforded

0.085 g (74%) of diazo ketone 105 as a yellow oil.

IR (film): 2950, 2060, and 1590 cm1

'H NMR (500 MHz, CDC13):

13C NMR (125 MHz, CDC13):

7.55-7.58 (m, 2H), 7.36-7.40 (m, 3H), 6.09 (q, J= 7.0 Hz, 1H), 1.78 (s, 3H), 1.72 (d, J = 7.0 Hz, 3H),1.09 (q, J= 5.8 Hz, 4H), and 1.05 (t, J= 5.8 Hz, 6H)

197.0, 137.4, 135.1, 134.4, 131.5, 130.5, 128.8, 52.3,14.3, 13.6, 7.9, and 4.1

116

wdd 0 Tc* * ~ * P P P i n I I p p9£96

t-

90

0

v 5

0;Y N2

SiEt2Ph

105

PhEt2SI C 0

113

(E)-2-(1-Methyl-1-propenyl)-2-(diethyl(phenyl)silyl)ketene (113).

A solution of diazo ketone 105 (0.194 g, 0.677 mmol) in 6.7 mL of benzene was

placed in a 25-cm vycor tube (15 mm O.D., 13 mm I.D.) fitted with a rubber septum. A

second rubber septum (inverted) was secured with wire to the tube to insure a good seal

and the reaction mixture was degassed (three freeze-pump-thaw cycles at -196 *C, <0.05

mmHg) and then irradiated with 300 nm light for 3 h in a Rayonet reactor. Column

chromatography on 10 g of silica gel (elution with hexane) provided 0.111 g (64%) of

ketene 113 as a yellow oil.

IR (film): 3360, 2440, and 2060 cm 1

'H NMR (500 MHz, CDCl 3): 7.53-7.57 (m, 2H), 7.32-7.39 (m, 3H), 4.96 (q, J= 6.8 Hz, 1H), 1.80 (s, 3H), 1.51 (d, J = 6.8 Hz,3H), and 0.90-1.04 (m, 1OH)

'3C NMR (125 MHz, CDCl3): 184.8, 136.2, 135.1, 130.1,25.0, 19.2, 14.6, 7.9, and 4.8

128.6, 123.8,

118

119.3,

PhEt 2SI C

113

f

LYJ-UI .

9 8 7 6 5 4 p

((

3 2 1 0 ppm4

SiEt3

I I

Ph

0

C+ c

Et3SI 0

Phgc

122 124

4 ,4 -Dichloro-3-phenyl-2-(triethylsilyl)-2-cyclobutenone (124).

A 500-mL, three-necked, round-bottomed flask equipped with a rubber septum,

reflux condenser, and' pressure-equalizing addition funnel was charged with activated

zinc 59 (7.26 g, Ill mmol) and a solution of (triethylsilyl)phenylacetylene' 60 122 (4.01 g,37.0 mmol) in 80 mL of ether. The resulting solution was heated at reflux while a

solution of trichloroacetyl chloride (4.2 mL, 37.0 mmol) in 125 mL of ether was added

dropwise over 3 h. The reaction mixture was heated at reflux for 14 h and was then

allowed to cool to room temperature and filtered. The filtrate was extracted with two

1 00-mL portions of saturated NaHCO 3, two 1 00-mL portions of water, and two 1 00-mL

portions of saturated NaCl, dried over MgSO4, filtered, and concentrated to give 6.67 g of

a red-brown liquid. Column chromatography on 40 g of silica gel (elution with 0-2%

ether-pentane) provided 5.55 g (92%) of cyclobutenone 124 as a yellow oil.

2950 and 1770 cm 1



7.91-7.94 (m, 2H), 7.57-7.60 (m, 3H), and 0.85-1.00 (m, 15H)

183.4, 181.9, 149.8, 133.1, 129.9, 129.7, 129.2,91.0, 7.2, and 3.1.

159.Brady, W. T.; Liddell, H. G.; Vaughn, W. L. J. Org. Chem. 1966, 31, 626.160. Jun, C.-H.; Crabtree, R. H. J. Organomet. Chem. 1993, 447, 177.

120

IR (film):

Et 3SI 0

Ph Cl

124

t~J

2

rJL )

[ I19 3 7

2 1 S ppm

5 49 i 7 ppo

Me 3SI 0 M03SI

Ph I PhX

123 125

3 -Phenyl-2-(trimethylsilyl)-2-cyclobutenone (125).

A 50-mL, three-necked, round-bottomed flask equipped with a rubber septum,

25-mL pressure-equalizing addition funnel, and argon inlet adapter was charged with zinc

dust (0.626 g, 9.53 mmol), TMEDA (1.44 mL, 9.53 mmol), and 8 mL of ethanol and then

cooled at 0 "C using an ice-water bath. Acetic acid (0.55 mL, 9.53 mmol) was added

over 2 min, and then a solution of cyclobutenone 123 (0.463 g, 1.64 mmol) in 5 mL of

ethanol was added dropwise via the addition funnel over 12 min. The reaction mixture

was stirred for 15 min, and then the ice-bath was removed and the reaction mixture was

stirred for 3 h at 25 *C. The resulting mixture was filtered through Celite with the aid of

80 mL of 1:1 Et2O:pentane. The filtrate was extracted with 100 mL of 1 M aqueous HCl

solution, 100 mL of water, 80 mL of saturated NaHCO 3, and 80 mL of saturated NaCl

solution, dried over MgSO 4, filtered, and concentrated to give 0.409 g of a yellow liquid.

Column chromatography on 20 g of silica gel (gradient elution with 0-4% EtOAc-

hexane) provided 0.257 g (74%) of 125 as a yellow oil.

IR (film): 2950 and 1785 cm'

'H NMR (300 MHz, CDCl3): 7.59-7.61 (m, 2H), 7.50-7.51 (m, 3H), 3.71 (s, 2H),and 0.32 (s, 9H)

"C NMR (75 MHz, CDC13): 191.0, 176.7, 131.4, 129.2, 128.7, 128.6, 126.4,54.6, and -1.2

HRMS: Calcd For C13HI6OSi: 216.0970Found: 216.0970

122

M03SI 0

Ph

126

3.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 PPM

Et 3Si 0 Et3SI 0

CIPh)I Ph

124 126

3-Phenyl-2-(triethylsilyl)-2-cyclobutenone (126).

A 100-mL, 3-necked, round-bottomed flask equipped with a rubber septum, 50-

mL pressure-equalizing addition funnel, and argon inlet adapter was charged with zinc

dust (1.14 g, 17.5 mmol), TMEDA (2.6 mL, 17.5 mmol), and 13 mL of ethanol and then

cooled at 0 *C using an ice-water bath. Acetic acid (1.00 mL, 17.5 mmol) was added

over 8 min, and then a solution of cyclobutenone 124 (0.99 g, 3.0 mmol) in 16 mL of

ethanol was added dropwise via the addition funnel over 1 h. The reaction mixture was

stirred for 15 min, and then the ice-bath was removed and the reaction mixture was

stirred for 4.5 h at room temperature. The resulting mixture was filtered through Celite

with the aid of 150 mL of 1:1 Et20:pentane. The filtrate was extracted with 50 mL of 1

M aqueous HCl solution, 50 mL of water, 40 mL of saturated NaHCO3, and 40 mL of

saturated NaCl solution, dried over MgSO 4, filtered, and concentrated to give 0.848 g of a

yellow liquid. Column chromatography on 25 g of silica gel (gradient elution with 0-

10% ether-pentane) provided 0.428 g (55%) of 126 as a yellow oil.

IR (film): 2070 and 1740 cm 1

'H NMR (300 MHz, CDCl 3): 7.63-7.66 (m, 2H), 7.51-7.54 (m, 3H), 3.75 (s, 2H),0.99 (t, J =6.8 Hz, 9H), and 0.86 (q, J= 6.8 Hz,6H)

'3C NMR (75 MHz, CDCl3 ): 191.5, 178.1, 147.2, 133.6, 131.4, 129.0, 128.7,52.1, 7.4, and 3.4

HRMS: Calcd For C16H22 OSi: 258.1440Found: 258.1441

124

id~d t £9

9I

0

1

Me 3SI 0

PhX

125

M03SI C40

Ph:

127

2-(1-Phenylethenyl)-2-(trimethylsilyl)ketene (127).

A 25-mL, two-necked, round-bottomed flask equipped with a glass stopper and

reflux condenser was charged with cyclobutenone 125 (0.106 g, 0.497 mmol) in 18 mL

of benzene and then heated at 60 *C for 4.5 h. The resulting mixture was allowed to cool

to room temperature and then concentrated to give 0.113 g of a yellow oil. Column

chromatography (twice) on 2.0 g Florisil (elution with hexane) provided 0.039 g (37%) of

ketene 127 as a pale yellow oil.

2080 and 1740 cm'

'H NMR (300 MHz, CDCl 3):


HRMS:

7.41-7.43 (m, 2H), 7.30-7.40 (m, 3H), 4.98 (s, 2H),and 0.28 (s, 9H)

180.9, 142.7, 139.4, 128.3,26.0, and -0.6

Calcd For C13H16OSi:Found:

128.0, 126.8, 110.9,

216.0970216.0970

126

IR (film):

HddO 0 9 L V

IZ

) 4ldI

SOL-

Me 3Si 0

Ph)ft125

3-Phenyl-2-trimethylsilyl butenolide (129).

A 50-mL, one-necked, round-bottomed flask equipped with a reflux condenser

was charged with cyclobutenone 125 (0.153 g, 0.703 mmol) in 30 mL of toluene. Air

was bubbled through the solution at a rate of about 5 bubbles per second. The reaction

mixture was heated at reflux for 18 h then cooled to room temperature and concentrated

to give 0.213 g of a yellow oil. Column chromatography on 10 g of silica gel (elution

with 0-10% EtOAc-hexane) afforded 0.069 g (42%) of butenolide 129 as a yellow oil.

IR (CDCl3): 2950 and 1740 cm"


'3 C NMR (75 MHz, CDC13):

7.45-7.46 (m, 3H), 7.29-7.44 (m, 2H), 4.93 (s, 2H),and 0.149 (s, 9H)

176.9, 174.0, 133.4, 130.0, 128.5, 127.9, 127.4,73.7, and -0.97

128

0

Me 3SI

Ph

129

0

M03SI 0

Ph129

K-

:97.0 6.0 5.0 4.0 3.0 2.0 1.0 0. 0 Ppm

O.-A

Me3SI

PhX

OH

Me 3SI CO 2Me

Ph) CO 2MO

125 171

Dimethyl 3-Hydroxy-5-phenyl-4-(trimethylsilyl)phthalate (171).


reflux condenser was charged with a solution of cyclobutenone 125 (0.109 g, 0.511

mmol), DMAD (0.063 mL, 0.511 mmol), and 7 mL of toluene. The resulting solution

was heated at reflux for 42 h and then allowed to cool to room temperature.

Concentration of the resulting mixture afforded 0.260 g of an orange oil. Column

chromatography on 26 g of silica gel (elution with 20% EtOAc-hexane) provided 0.116 g

(63%) of phenol 171 as a white solid, mp 51-52 *C.

1740 and 1660 cmn~

'H NMR (300 MHz, CDCl3): 11.21 (s, lH), 7.33-7.35 (m, 3H), 7.22-7.24 (m,2H), 6.79 (s, lH), 3.91 (s, 3H), 3.85 (s, 3H), and-0.04 (s, 9H)

13C NMR (75 MHz, CDCl3 ): 170.0, 169.4, 166.5, 156.0,128.7, 127.9, 127.8, 121.1,0.8

143.0, 135.1, 129.3,106.9, 52.8, 52.5, and

HRMS: Calcd For C19H220 5Si:Found:

130

IR (CC14):

358.1236358.1235

OH

M03SI C0 2Me

Ph): C0 2MO

171

-a

6 4

K

10 B 2 0oPPIV

Et 3Si 0

PhX

OH

Et 3SI CO 2MO

Ph CO 2Me

126 172

Dimethyl 3-Hydroxy-5-phenyl-4-(triethylsilyl)phthalate (172).


reflux condenser was charged with a solution of cyclobutenone 126 (0.111 g, 0.429

mmol), DMAD (0.053 mL, 0.429 mmol), and 7 mL of toluene. The resulting solution

was heated at reflux for 55 h and then allowed to cool to room temperature.

Concentration of the resulting mixture afforded 0.167 g of an orange oil. Column

chromatography on 10 g of silica gel (gradient elution with 0-70% benzene-hexane)

afforded 0.095 g (55%) of phenol 172 as a white solid, mp 61-61.5 'C.

1730 and 1660 cm~


'3C NMR (75 MHz, CDCl3):

HRMS:

11.28 (s, 1H), 7.32-7.36 (m, 5H), 6.79 (s, 1H), 3.93(s, 3H), 3.87 (s, 3H), 0.78 (t, J= 7.8 Hz, 9H), and0.48 (q, J= 7.8 Hz, 6H)

170.1, 169.5, 166.8, 157.2, 143.1, 135.0, 128.7,127.8, 127.7, 127.0, 121.3, 106.6, 52.8, 52.5, 7.8and 4.2

Calcd For C22 H28O5Si:Found:

400.1706400.1707

132

IR (CCL4):

Ndd 0 P 9 9 al

9

HO

v II

I

I

j

o(i-Pr)3S1 C'

111

EtO2C CO 2Et

0(I-Pr)3SI 0

CO 2EtCO 2Et

196 202

3,4-Dimethyl-6-oxo-5-(triisopropylsilyl)-3,6-dihydropyran-2,2-dicarboxylic

diethyl ester (202).

acid

A 10-mL, two-necked, pear-shaped flask equipped with a glass stopper and a

reflux condenser fitted with a rubber septum and an argon inlet needle was charged with

ketene 111 (0.121 g, 0.479 mmol), diethyl ketomalonate (0.110 mL, 0.719 mmol), and

0.43 mL of acetonitrile. The reaction mixture was heated at reflux for 15 min and then

cooled and concentrated at reduced pressure to give 0.280 g of a yellow liquid. Column

chromatography on 14 g of silica gel (elution with 5% EtOAc-hexane) provided 0.191 g

(94%) of lactone 202 as a white solid, mp 69-70 *C.

IR (CH 2C 2): 2930, 2860, and 1715 cm1


'3C NMR (75 MHz, CDCl3 ):

4.21-4.36 (m, 3H), 4.04-4.36 (m, 1H), 3.21 (q, J=7.2 Hz, IH), 2.12 (s, 3H), 1.47 (sept, J= 7.5 Hz,

3H), 1.32 (t, J= 7.2 Hz, 3H), 1.24 (t, J= 7.2 Hz,3H), 1.13 (d, J= 7.0 Hz, 3H), and 1.04 (d, J= 7.5Hz, 18H)

168.9, 166.6, 164.4, 163.7, 122.0, 85.0, 62.7, 62.5,41.5, 23.3, 19.0, 14.0, 13.7, 12.9, and 12.4

Elemental Analysis: Calcd for C22H 38O6Si:Found:

C, 61.93; H, 8.98C, 62.12; H, 9.05

134

wdd 0I I I I a I I

£ V 9L

I -fl

U

a'b NO 0--- Wova,

0

mw

V/-4

(I-Pr) 3Si C4 0

6110

+

EtO2C CO 2 Et

196

0

(I-Pr)3Si 0

CO2EtCO 2Et

203

3-Oxo-3,5,6,7,8,8a-hexahydro-isochromene-1,1-dicarboxylic acid diethyl ester (203).

A flame-dried, threaded Pyrex tube (13 mm O.D., 10 mm I.D.) was charged with

ketene 110 (0.208 g, 0.679 mmol), diethyl ketomalonate (0.155 mL, 1.02 mmol), and

0.68 mL of acetonitrile. The tube was tightly sealed with a teflon cap and then heated at

reflux for 2 h. The reaction mixture was cooled to room temperature and concentrated at

reduced pressure to give 0.461 g of an yellow-orange oil. Column chromatography on

10 g of silica gel (elution with 5% EtOAc-hexane) provided 0.250 g (77 %) of lactone

203 as a white solid, mp 79-80 *C.

IR (CDCl3):



2930,2850, 1740 and 1710 cm-1

4.24-4.38 (m, 3H), 4.05-4.14 (m, 1H), 3.26 (dd, J=12.0, 3.3 Hz, IH), 2.76 (m, 1H), 2.35 (dt, J= 12.3,4.6 Hz, 1H), 2.10 (m, 1H), 1.82-1.91 (m, 2H), 1.54-1.67 (m, 3H), 1.46 (sept, J = 7.3 Hz, 3H), 1.32 (t, J= 7.3 Hz, 3H), 1.25 (t, J= 7.3 Hz, 3H), and 1.05 (d,J= 7.3 Hz, 18H)

174.1, 167.9, 165.7, 164.6, 122.0, 85.0, 63.9, 63.4,46.8, 37.4, 31.0, 30.5, 26.2, 19.8, 14.8, 14.5, and13.0

136

0

(I.Pr)3S1

C0 2Et

C O02Et

203

7 6 5 A 3 0 1 PPM

+

EtO2C CO2 Et

196

0

Et3Si

Ph CO 2EtCO 2Et

205

6-Oxo-4-phenyl-5-(triethylsilyl)-3,6-dihydropyran-2,2-dicarboxylic acid diethyl ester

(205).


cyclobutenone 126 (0.115 g, 0.445 mmol), diethyl ketomalonate (196, 0.100 mL, 0.667

mmol), and 0.40 mL of acetonitrile. The tube was tightly sealed with a teflon cap and

heated at reflux for 15 h. The reaction mixture was cooled and concentrated at reduced

pressure to afford 0.250 g of a yellow oil. Column chromatography on 12 g of silica gel

(elution with 10% EtOAc-hexane) provided 0.176 g (92%) of the lactone 205 as a white

solid, mp 63 *C.

IR (film): 2960, 2880, 1735, and 1575 cm 1



7.31-7.33 (m, 3H), 7.18-7.21 (m, 2H), 4.24 (q, J=7.3 Hz, 4H), 3.21 (s, 2H), 1.26 (t, J= 7.3 Hz, 6H),0.71 (t, J= 7.9 Hz, 9H), and 0.34 (q, J= 7.8 Hz,6H)

166.3, 164.6, 163.4, 139.9, 129.9, 129.5, 128.2,127.7, 82.8, 63.1, 38.1, 13.9, 7.4 and 4.2

Elemental Analysis: Calcd for C23H320 6Si:Found:

C, 63.86; H, 7.46C, 63.78; H, 7.31

138

Et3SI

PhX

126

0

Et3SI

10COEPhA CO 2Et

205

b-"

3 76 54 32 10Im Ir-l

PPM

0

Et 3SI 0

Ph CO 2EtCO 2Et

205

0

10Ph CO 2Et

CO 2Et

206

6-Oxo-4-phenyl-3,6-dihydropyran-2,2-dicarboxylic acid diethyl ester (206).


a solution of silyl lactone 205 (0.065 g, 0.151 mmol) in 0.75 mL of CH2Cl 2.

Methanesulfonic acid (0.049 mL, 0.756 mmol) was added rapidly dropwise. The tube

was tightly sealed with a teflon cap and heated at reflux for 15 h. The reaction mixture

was cooled, diluted with 10 mL of CH 2C 2, and washed with 10 mL of saturated NaHCO 3

solution, 10 mL of water, 10 mL of 10% aqueous HCl solution, and 10 mL of saturated

NaCl solution, dried over MgSO 4, filtered, and concentrated at reduced pressure to give

0.060 g of a cloudy brown oil. Column chromatography on 14 g of silica gel (elution

with 20-50% EtOAc-hexane) provided 0.048 g (100%) of the lactone 206 as a white

solid, mp 68.5 *C.

IR (film): 2950,2900, 1705, 1435, and 1360 cm7

'H NMR (500 MHz, CDCl3 ):


7.45-7.55 (in, 5H), 6.33 (s, 1H), 4.31 (q, J= 7.1 Hz,4H), 3.48 (s, 2H), and 1.29 (t, J= 7.2 Hz, 6H)

165.9, 161.9, 152.3, 135.2, 131.0, 129.0, 126.1,114.4, 82.9, 63.2, 31.1, and 13.8

Elemental Analysis: Calcd for C17Hi80 6:Found:

C, 64.14; H, 5.70C, 64.25; H, 5.88

140

0

0 CO2Et

Ph CO2Et

206

p

9 8 7 6 5

cf

K

14 3 .2

p.

1 0 ppm

"-A

3 24

MeH , NP

H Ph

0

(I-Pr)3Si NH

Ph

215 233

cis-4,5-Dimethyl-6-phenyl-3-(triisopropylsilyl)- 5,6-dihydro-1H-pyridin-2-one (233).



ketene 111 (0.134 g, 0.531 mmol), imine 215 (0.147 g, 0.829 mmol), and 0.47 mL of

acetonitrile. The reaction mixture was heated at reflux for 1.5 h then cooled to room

temperature and concentrated at reduced pressure to give 0.272 g of a yellow oil.

Column chromatography on 27 g of silica gel (elution with 5% EtOAc-hexane) provided

0.151 g (79%) of lactam 233 as a white solid, mp 197-199 *C.

IR (CDCl3): 3400, 2930, 2850, and 1625 cm~1


"C NMR (75 MHz, CDCl3):

7.28-7.42 (m, 5H), 5.51 (br s, 1H), 4.77 (d, J= 3.6Hz, IH), 2.21 (m, 1H), 2.10 (s, 3H), 1.53 (sept,J= 7.3 Hz, 3H), 1.12 (d, J= 5.7 Hz, 18H), and 0.82(d, J= 7.1 Hz, 3H)

170.0, 165.9, 138.4, 128.5, 127.6, 126.6, 126.4,58.1,45.0, 23.0, 19.3, 13.0, and 11.0

Elemental Analysis: Calcd for C22H35NOSi:

Found:

C, 73.89; H, 9.86;N, 3.92C, 73.80; H, 9.88;N, 3.84

142

( )'0

+

111

wdd 06 p p I

cez

ISC(Jd-,)

0

-I

J

I t I I a I

Et3Si C'

+

112

Me3SiN

H Ph

0Et 3Si NH

Ph

215 236

cis-4,5-Dimethyl-6-phenyl-3-(triethylsilyl)- 5,6-dihydro-1H-pyridin-2-one (236).

A 5-mL, pear-shaped flask equipped with a reflux condenser fitted with a rubber

septum and an argon inlet needle was charged with ketene 111 (0.160 g, 0.761 mmol) and

imine 215 (0.202 g, 1.14 mmol) then stirred at room temperature for 2 h. The resulting

yellow oil was purified by column chromatography on 10 g of silica gel (elution with 0-

10% EtOAc-hexane) which provided 0.182 g (76%) of lactam 236 as a white solid, mp

155 *C.

IR (CH 2C 2): 2930, 2850, and 1630 cm'1



7.28-7.41 (m, 5H), 5.56 (br s, 1H), 4.79 (d, J= 4.0Hz, 1H), 2.22 (m, 1H), 2.08 (s, 3H), 0.97 (t, J= 7.8Hz, 9H), and 0.86 (m, 9H)

171.0, 167.5, 139.4, 129.3, 128.5, 127.3, 127.1,58.9, 44.9, 22.8, 11.7, 8.5, and 5.8

144

wddK I I a Ic vyI I I I p I 0 1 1 1 a I I A a I I I I I I

I ~t* r..[vf

It)

99Z

H NlISC13

s9

Me3SSi

H Ph

0

(i-Pr)3S! NH

/Ph

215 237

trans-1-Phenyl-4-(triisopropylsilyl)-1,5,6,7,8,8a-hexahydro-2H-isoquinolin-3-one

(237).



acetonitrile. The tube was tightly sealed with a teflon cap and then heated at reflux for 25

h. The reaction mixture was cooled to room temperature and concentrated at reduced

pressure to give 0.403 g of a yellow oil. Column chromatography on 14 g of silica gel

(elution with 10% EtOAc-hexane) provided 0.256 g (91%) of lactam 237 as a white solid,

mp 211-212 *C.

IR (CDCl3 ): 3400, 2930, 2850, and 1625 cm 1



7.29-7.41 (m, 5H), 5.43 (br s, 1H), 4.79 (d, J= 5.2Hz, IH), 2.82 (d, J= 11.6 Hz, 1H), 2.25 (m, 2H),2.00 (m, 1H), 1.73 (m, 1H), 1.52 (sept, J= 7.4 Hz,3H), 1.25-1.48 (m, 4H), and 1.12 (d, J= 7.5 Hz,18H)

170.5, 170.3, 138.5, 128.5, 127.7, 126.7, 123.7,57.3, 49.2, 36.9, 30.9, 28.2, 25.9, 19.3, and12.6


Found:

C, 75.14; H, 9.72;N, 3.65C, 75.10; H, 10.01;N, 3.66

146

110

(/-Pr)3Si CV.'

IIII it

'if

HNISC(Jd-I)

0

odd *IUCU Sa

I r

I

)

'-4

now

e i

Et 3Si 0

PhX

126

*1-

Me 3SK .

H Ph

0

Et3Si NH

Ph Ph

215 239

4,6-Diphenyl-3-(triethylsilyl)- 5,6-dihydro-1H-pyridin-2-one (239).



cyclobutenone 126 (0.134 g, 0.518 mmol), imine 215 (0.133 g, 0.778 mmol), and 0.47

mL of acetonitrile. The reaction mixture was heated at reflux for 45 min and then cooled

and concentrated at reduced pressure to give 0.259 g of an yellow oil. Column

chromatography on 25 g of silica gel (elution with 5-20% EtOAc-hexane) provided 0.157

g (84%) of lactam 239 as a white solid, mp 130-131 *C.

IR (film): 3400, 2940,2860, 2220, and 1630 cm 1

'H NMR (300 MHz, CDCl3): 7.29-7.41 (m, 8H), 7.12-7.18 (m, 2H), 5.72 (s, 1H),4.74 (dq, J= 5.1, 1.6 Hz, 1H), 2.80 (m, 2H), 0.81(t, J= 7.6 Hz, 9H), and 0.43 (q, J= 7.6 Hz, 6H)


Elemental Analysis:

170.7, 162.7, 142.8, 141.6,128.9, 128.7, 128.5, 127.3,

Calcd for C23H29NOSi:

Found:

134.4, 129.6, 129.1,56.0, 43.7, 8.5, and 5.2

C, 75.98; H, 8.04;N, 3.85C, 75.76; H, 7.99;N, 3.77

148

0

Et3SI

I NH

Ph & Ph

239

87 65 4 32 1 0OPpm

MeSi, N

H

240 Ph

0

(i-Pr)3Si NH

Ph241

cis-4,5-Dimethyl-6-styryl-3-(triisopropylsilyl)-5,6-dihydro-1H-pyridin-2-one (241).

A 10-mL, pear-shaped flask equipped with a rubber septum and argon inlet

needle was charged with ketene 111 (0.165 g, 0.654 mmol) and imine 240 (0.135 g,

0.654 mmol). The reaction mixture was stirred at 25 *C for 15 min and then transferred

to a round-bottomed flask with 5 mL of CH2Cl2 and concentrated at reduced pressure to

give 0.321 g of an orange oil. Column chromatography on 12 g of silica gel (elution with

0-10% EtOAc-hexane) provided 0.196 g (78%) of lactam 241 as pale yellow oily solid.

IR (film): 3160, 3040, 2900, 1625, and 1455 cm~1



7.24-7.40 (m, 5H), 6.62 (d, J= 15.8 Hz, IH), 6.16(dd, J= 16.0, 7.8 Hz, 1H), 5.84 (br s, 1H), 4.26 (dd,J= 7.7, 3.9 Hz, 1H), 2.19 (m, 1H), 2.06 (s, 3H),1.50 (sept, J= 7.5 Hz, 3H), 1.07 (d, J= 7.5 Hz,18H), and 1.07 (d, J= 7.0 Hz, 3H)

170.5, 166.9, 136.9, 133.6, 129.4, 128.7, 127.5,127.2, 126.8, 56.8, 44.1, 23.6, 20.0, 13.5, and 11.7

150

-o(i-r)Si C-

+

111

wdd- 0 T

w

-f

S 9L

LI lID II1'

9 6I I

f

4id

HN I~Ji

0

I I t I I --i I I I -- I --- L- I I I L-- I I I I

c

V"

(I-Pr)3SI C

110

Me 3SI , N+ -

H

Ph240

0

(i-Pr)3SINH

Ph242

trans-i-Styryl-4-(triisopropylsilyl)-1,5,6,7,8,8a-hexahydro-2H-isoquinolin-3-one

(242).


ketene 110 (0.199 g, 0.649 mmol) and imine 240 (0.133 g, 0.649 mmol). The reaction

mixture was stirred at 25 *C for 10 min and then transferred to a round-bottomed flask

with 5 mL of CH2Cl2 and concentrated at reduced pressure to give 0.384 g of a yellow-

orange liquid. Column chromatography on 12 g of silica gel (elution with 0-50% EtOAc-

hexane) provided 0.209 g (73%) of lactam 242 as a white solid, mp 162-163 *C.

IR (film): 3400, 2930, 2860, and 1625 cm-1

'H NMR (300 MHz, CDC13):


7.28-7.38 (m, 5H), 6.57 (d, J= 15.8 Hz, 1H), 6.20(dd, J= 8.1, 15.8 Hz, 1H), 5.37 (br s, 1H), 4.21 (m,1H), 2.79 (m, 1H), 2.35 (m, 1H), 2.20 (m, 1H), 1.98(m, 1H), 1.84 (m, 2H), 1.50 (sept, J= 7.5 Hz, 3H),and 1.09 (d, J= 7.5 Hz, 18H)

169.6, 169.2, 136.2, 132.6, 128.7, 128.0, 126.5,126.4, 124.6, 55.5, 46.8, 36.3, 29.3, 28.0, 25.5,19.4, and 12.7

Elemental Analysis: Calcd for C26H39NO:

Found:

C, 76.23; H, 9.60;N, 3.42C, 75.97; H, 9.84;N, 3.57

152

0

(I-Pr)3SINH

Ph

242

F 43

9 U7 6 4 3 a 1*

(i-Pr)3SI C

111

.SiMe3

Ph Ph

0

(i-Pr)3S NH P

PhPh

219 244

4,5-Dimethyl-6,6-diphenyl-3-(triisopropylsilyl)-5,6-dihydro-1H-pyridin-2-one (244).





pressure to give an orange oil. Column chromatography on 10 g of silica gel (elution

with 0-10% EtOAc-hexane) provided 0.208 g (79%) of lactam 244 as a white solid, mp

182 *C.

IR (CDCl 3): 2930, 2850, and 1620 cm 1



7.24 (m, 11 H), 3.09 (q, J= 6.9 Hz, 1H), 2.13 (s,3H), 1.36 (sept, J= 7.5 Hz, 3H), 0.94 (d, J= 6.9 Hz,3H), 0.85 (d, J= 7.3 Hz, 9H), and 0.82 (d, J= 7.5Hz, 9H)

170.0, 165.5, 146.6, 142.6, 128.4, 128.0, 127.4,126.9, 126.5, 126.4, 125.8, 65.0, 46.5, 23.7, 19.0,14.0, and 12.5


Found:

C, 77.54; H, 9.06;N, 3.23C, 77.87; H, 9.17;N, 3.32

154

ITS 9 S -

W/)

4d

4d

HN

y IOt(Jd-,)0

odd -a2tL

) r

S SMe3

+ Ph kPh

(i-Pr)3Si C'! '

6

110

0

(i-Pr)3Si NH

Ph

Ph

245

1,1-Diphenyl-4-(triisopropylsilyl)-1,5,6,7,8,8a-hexahydro-2H-isoquinolin-3-one

(245).





pressure to give 0.342 g of a yellow-orange oil. Column chromatography on 10 g of

silica gel (elution with 10% EtOAc-hexane) provided 0.149 g (66%) of lactam 245 as a

white solid, mp 237-237.5 *C.

IR (film): 3040, 2930, 2850, and 1630 cm-1

'H NMR (500 MHz, CDCl3 ): 7.14-7.30 (m, 10 H), 7.03 (br s, 1H), 2.99 (dd, J=2.4, 9.1 Hz, 1H), 2.84 (dd, J = 3.1, 11.6 Hz, IH),2.39 (m, 1H), 2.07 (m, 1H), 1.76 (m, 1H), 1.49 (m,3H), 1.33 (sept, J = 7.3 Hz, 3H), 1.13 (d, J = 12.8Hz, IH), 0.87 (d, J= 7.3 Hz, 9H), and 0.80 (dd, J=7.3 Hz, 9H)


Elemental Analysis:

170.4, 170.3, 146.9, 143.2,126.9, 126.6, 126.4, 125.3,31.2, 26.5, 19.0, and 12.5

Calcd for C3oH 41NOSi:

Found:

128.3, 127.9, 127.3,64.3, 51.1, 37.3, 31.5,

C, 78.38; H, 9.01;N, 3.05C, 78.44; H, 9.41;N, 3.03

156

219

udd g 1 3

S3 La 6 I I ..I ..I .I

TI

4d

4d !HN

ISC(Jd-I)

0

odd

rl-

(i-Pr)3Si Cl0

6 +

110

Me 3Si.. N

H t-Bu

251

0

(i-Pr)3Si NH

)< t-Bu

252

trans-1-(1,1-Dimethylethyl)-4-(triisopropylsilyl)-1,5,6,7,8,8a-hexahydro-2H-

isoquinolin-3-one (252).



acetonitrile. The tube was tightly sealed with a teflon cap and heated at reflux for 4 h and

then at 110-110 *C for 65 h. The reaction mixture was cooled to room temperature and

additional imine (0.093 g, 0.53 mmol) was added then heated at 110-130 *C for 23 h.

The reaction mixture was cooled to room temperature and concentrated to give 0.307 g of

a yellow-gray liquid. The reaction mixture was filtered through Celite, rinsing with the

aid of CH2Cl 2 and concentrated at reduced pressure to give 0.187 g of a pale yellow solid.

The solid was washed with heptane to afford 0.115 g (56%) of lactam 252 as a white

solid, mp 235-237 *C.

IR (CH2Cl 2): 2950,2850, and 1630 cm~1



5.19 (br s, 1H), 3.21 (d, J= 4.1 Hz, 1H), 2.72 (m,1H), 2.25 (m, 2H), 2.00 (m, 2H), 1.80 (m, 1H), 1.47(sept, J= 7.5 Hz, 3H), 1.47 (m, 3H), 1.07 (d, J= 7.1Hz, 18H), and 1.02 (s, 9H)

171.1, 170.4, 124.0, 60.7, 47.2, 37.2, 32.5, 30.8,28.6, 27.3, 26.0, 19.3, and 12.6

Elemental Analysis: Calcd for C22H4iNOSi:

Found:

C, 72.66; H, 11.36;N, 3.85C, 72.59; H, 11.65;N, 3.78

158

0

(I.Pr)3S1NH

252

I-AA

3 7 654 3 21 0OPPM

(i-Pr)3Sj~ - H3C

H Ph

0 0

(i-Pr)3S1 CH3 (i-Pr)3S N CH3

Ph Ph

253 254 255

trans-6-Phenyl-3-(triisopropylsilyl)-1,4,5-trimethyl-5,6-dihydro-1H-pyridin-2-one

(254) and cis-6-Phenyl-3-(triisopropylsilyl)-1,4,5-trimethyl-5,6-dihydro-1H-pyridin-

2-one (255).


a solution of ketene 111 (0.069 g, 0.273 mmol), 0.27 mL of acetonitrile, and imine 253

(0.050 g, 0.410 mmol). The tube was tightly sealed with a threaded teflon cap and heated

at 120 C for 42 h. The reaction mixture was cooled and concentrated at reduced pressure

to give 0.120 g of an orange oil. Column chromatography on 12 g of silica gel (elution

with 5% EtOAc-hexane) afforded 0.072 g (71%) of a 3:1 mixture of lactams 254 and 255

as a pale yellow oil.

Both isomers:

IR (neat): 2940,2850, and 1615 cm 1

Major isomer (254):


'3C NMR (75 MHz, CDCl 3):

Minor isomer (255):


7.19-7.29 (m, 5H), 4.32 (d, J= 6.1 Hz, IH), 2.90 (q,obscured by the singlet at 2.89 ppm, IH), 2.89 (s,3H), 1.92 (s, 3H), 1.54 (m, 3H), 1.06-1.13 (d, J=7.4 Hz, 18H), and 0.92 (d, J= 7.3 Hz, 3H)

168.9, 160.4, 137.2, 129.0, 128.0, 127.8, 126.1,66.2,40.6, 33.5, 21.7, 19.4, 19.3, and 13.1

7.19-7.29 (m, 5H), 4.19 (s, 1H), 3.03 (s, 3H), 2.35(q, J= 7.0 Hz, 1H), 1.79 (s, 3H), 1.54 (sept, J= 7.4Hz, 3H), 1.30 (d, J= 7.0 Hz, 3H), and 1.00 (d, J=

160

7.4 Hz, 18H)

3C NMR (75 MHz, CDC 3): 167.8, 160.9, 140.3, 128.6, 128.2, 127.4, 127.2,66.7, 45.4, 35.1, 23.5, 19.2, 14.0, and 12.9

161

0 0

(/-Pr)3S N .*CH3 (NPr)3SI N CH3

"Ph Ph

254 255

5 5

F

4 32

v.

6 pp.

t'j

B a 7

.J- -- w , ___j

0

C

(i-Pr)3Si H

281

Me3Si N

H Ph

215

trans-N-(Trimethylsilyl)-4-phenyl-3-(triisopropylsilyl)-azetidin-2-one (283).


(triisopropylsilyl)ketene (0.063 g, 0.318 mmol) and imine 215 (0.057 g, 0.321 mmol).

The tube was tightly sealed with a threaded teflon cap and then heated at 70 *C for 20 h

and then at 140 *C for 3 h. The reaction mixture was cooled to give 0.117 g of an orange

oil. Column chromatography on 10 g of silica gel (elution with 0-20% EtOAc-hexane)

afforded 0.085 g (71%) of lactam 283 as a pale yellow oil.

IR (film): 2930, 2850, 2230, 1710, and 1450 cm 1



7.29-7.40 (m, 5H), 4.47 (d, J= 3.1 Hz, 1H), 3.00 (d,J= 3.1 Hz, 1H), 1.26 (sept, J= 7.0 Hz, 3H), 1.09 (d,J = 7.3 Hz, 9H), and 1.05 (d, J = 7.6 Hz, 9H), and0.10 (s, 9H)

176.3, 143.0, 129.3, 128.7, 127.2, 54.6, 52.1, 19.4,11.1, and -0.4

163

23SiMeN

(i-Pr)3Si P" Ph

283

) NlAeIM3

(I-Pr)3S1 Ph

283

b."

f

Al *1~I 9uu u i i ~ p 5 1 1 ,t

3 2

jfj

1 0 -ppm

I

9

0

Et3Si NH

Ph Ph

239

0

NH

Ph Ph

284

4,6-Diphenyl-5,6-dihydro-1H-pyridin-2-one (284)


a solution of silyl lactam 239 (0.091 g, 0.250 mmol) in 1.01 mL of CH2Cl 2.

Methanesulfonic acid (0.081 mL, 1.25 mmol) was added. The tube was tightly sealed

with a threaded teflon cap and the reaction mixture was heated at reflux for 4 h. The

reaction mixture was cooled and then diluted with 10 mL of CH2Cl 2 . The organic

portions were separated and washed with 10 mL of saturated NaHCO3 solution, 10 mL of

water, 10 mL of saturated NaCl solution, dried over MgSO4, filtered, and concentrated at

reduced pressure to give 0.095 g of a pale yellow solid. Column chromatography on 14 g

of silica gel (elution with 50-100% EtOAc-hexane) afforded 0.034g (55%) of the lactam

284 as a white solid, mp 182 *C.

IR (CH2 Cl2 ): 3380, 3020,2960, and 1640 cm~



7.35-7.51 (m, 10H), 6.37 (s, 1H), 5.65 (br s, 1H),4.86 (dd, J= 5.5, 11.6 Hz, 2H), and 2.96 (m, 1H)

167.5, 149.9, 141.1, 137.3, 129.6, 128.9, 128.7,128.3, 126.4, 125.8, 118.9, 55.7, and 35.7

165

wdd 0 T p S 9 1 6

I I I I~ II * *V

t1Z2)4id H V i d

HN

0

)

0

Et 3Si NH

Ph

236

0

NH

Ph

285

4,5-Dimethyl-6-phenyl-5,6-dihydro-1H-pyridin-2-one (285).

A flame-dried, 25-mL, round-bottomed flask equipped with a reflux condenser

fitted with a rubber septum and argon inlet needle was charged with a solution of silyl

lactam 236 (0.075 g, 0.238 mmol) in 1.2 mL of CH 2C 2. Methanesulfonic acid (0.077

mL, 1.19 mmol) was added and the reaction mixture was heated at reflux for 6 h. The

reaction mixture was cooled, diluted with 10 mL of CH2Cl2, and washed with 10 mL of

saturated NaHCO 3 solution, 10 mL of water, 10 mL of 10% aqueous HCl solution, 10 mL

of saturated NaCl solution, dried over MgSO 4, filtered, and concentrated at reduced

pressure to give 0.087 g of an oily yellow-brown solid. Column chromatography on 8 g

of silica gel (elution with 20-100% EtOAc-hexane) afforded 0.040g (83%) of the lactam

285 as a white solid, mp 170.5 *C.

IR (CH 2Cl2): 3980, 2950, 1655, and 1615 cm 1

'H NMR (500 MHz, CDCl3): 7.29-7.42 (m, 5H), 5.76 (s, 1H), 5.45 (br s, 1H),4.90 (d, J= 4.6 Hz, 1H), 2.29 (m, 1H), 2.00 (s, 3H),and 0.80 (d, J= 7.0 Hz, 3H)

13C NMR (125 MHz, CDCl3 ): 168.6, 159.4, 139.2, 129.5,59.9, 41.4, 22.5, and 12.2

128.8, 127.2, 119.6,

167

T Z £ I U aIFs9£ 9 6 I I I I a I I I I I I I I I I I I I a I I a

00

4td

HN

0

wdd 0

IfI

.p+ - 0

(-Pr) 3Si C Ph 2S- CMe 2 (i-Pr)3Si

111 292

3,4,5,5-Tetramnethyl-2-(triisopropylsilyl)cyclopent-2-en-1-one (292).

A 25-mL, two-necked, round-bottomed flask equipped with an argon inlet

adapter and rubber septum was charged with a solution of ketene 111 (0.140 g, 0.554

mmol) in 5.0 mL of 50:50 THF-DMSO and cooled at -20 'C. A 15-mL, two-necked,

round-bottomed flask equipped with an argon inlet adapter and rubber septum was

charged with diphenylisopropylsulfonium fluoroborate (0.193 g, 0.610 mmol) in 2.5 mL

of THF and cooled at -78 *C with an acetone-dry ice bath. t-BuLi (1.55 M in pentane,

0.38 mL, 0.582 mmol) was added dropwise down the side of the flask over 10 sec. The

resulting bright yellow ylide solution was immediately transferred via cannula over 5 min

to the ketene solution, and the resulting solution was stirred for 30 min at -20 *C. The

cooling bath was removed and the reaction mixture was allowed to warm to room

temperature over 3 h and then diluted with 20 mL of water. The aqueous layer was

separated and extracted with three 20-mL portions of diethyl ether, and the combined

organic phases were washed with 40 mL of saturated NaCl solution, dried over MgSO4,

filtered, and concentrated to give 0.275 g of a yellow oil. Column chromatography on 10

g of silica gel (gradient elution with 0-2.5% EtOAc-hexane) afforded 0.093 g (57%) of

cyclopentenone 292 as a colorless oil.

IR (CDCl3): 2960, 2940, 2860, and 1680 cm-'

'H NMR (500 MHz, CDCl3): 2.45 (q, J= 7.3 Hz, IH), 2.15 (s, 3H), 1.51 (s, 3H),1.51 (sept, J= 7.6 Hz, 3H), 1.07 (d, J= 7.3 Hz, 3H),1.05 (s, 3H), 1.04 (d, J= 7.6 Hz, 18H), and 0.94 (s,3H)

13C NMR (125 MHz, CDCl 3): 217.8, 187.6, 132.2, 53.2, 47.7, 26.4, 20.5, 19.0,

169

18.8, 14.7, and 11.7

HRMS: Calcd For Cl 8H34OSi: 294.2379Found: 294.2378

170

Wdd 0 9

-~~~~~~~~ IT- It II If iriirii

ISC(Jd-I)

0

0(i-Pr)3S! C'.O

+ Li --=-- Ph O (i-Pr)hSi Ph

111 310 311

5-(Phenylmethylene)-3,4-dimethyl-2-(triisopropylsilyl)-2-cyclopentenone (311)

A 25-mL, two-necked, round bottomed flask equipped with a glass stopper and

reflux condenser was charged with phenyl acetylene (0.065 mL, 0.592 mmol) and 3 mL

of THF then cooled at -78 *C. n-BuLi (2.6 M in hexane, 0.23 mL, 0.592 mmol) was

added and the reaction mixture was stirred for 15 min. A solution of ketene 111 (0.136 g,

0.539 mmol) in 1 mL of THF was added dropwise via cannula to the -78 *C solution

over 3 min (1 mL THF rinse). The reaction mixture was stirred at -78 *C for 1.5 h then

warmed to -5 *C over 40 min. BHT (0.133 g, 0.592 mmol) was added and the solution

was allowed to warm to room temperature over 15 min then heated at 50 *C for 35 min.

The reaction mixture was cooled to room temperature and quenched by the addition of 10

mL of saturated NaCl solution. The organic portion was separated and the aqueous layer

was extracted with two 30-mL portions of CH 2Cl2. The organic portions were combined,

dried over Na2SO 4, filtered, and concentrated to give 0.379 g of an orange oil. Column

chromatography on 25 g of silica gel (elution with 0-1% ether-pentane) afforded 0.143 g

(75%) of cyclopentenone 311 as a pale yellow solid, mp 117 *C.

IR (film): 2940,2860, and 1705 cm~

'H NMR (300 MHz, CDCl3 ): 7.54 (d, J= 7.3 Hz, 2H), 7.26-7.43 (m, 4H), 3.73 (q,J= 7.0 Hz, lH), 2.29 (s, 3H), 1.61 (sept, J= 7.5 Hz,3H), 1.24 (d, J= 7.0 Hz, 3H), and 1.09 (d, J= 7.5Hz, 18H)

"C NMR (75 MHz, CDCl3 ): 201.5, 186.3, 139.7, 136.2, 135.2, 130.3, 129.2,128.7, 128.6, 44.7, 18.8, 16.7, 12.0, and 11.8

172

0

(I-Pr)3S1 P

311

-.4

7 6 5 mr~mrlnrFI I II I I

2 1 0

6

98 I I I T

PPM

0 0Et3SI)

Br)4

317 318

2-Bromo-3,4-dimethyl-1-cyclopentenone (318)

A 25-mL, pear-shaped flask equipped with a rubber septum and argon inlet needle

was charged with silylcyclopentenone 317 (0.120 g, 0.535 mmol), N-bromosuccinimide

(0.238 g, 1.34 mmol), and 1.8 mL of DMF and the solution was stirred at room

temperature for 48 h. An additional portion of NBS was added (0.093 g, 0.522 mmol)

and the reaction mixture was stirred for 16 h then quenched by the addition of 10 mL of a

10% aqueous HCl solution. The organic portion was separated and the aqueous portion

was extracted with 3 20-mL portions of ether. The organic portions were combined and

washed with 15 mL of water, 15 mL of saturated NaHCO 3 solution, 15 mL of water, and

15 mL of saturated NaCI solution, then dried over MgSO4, filtered, and concentrated to

afford 0.197 g of a red-orange liquid. Column chromatography on 10 g of silica gel

(elution with 0-10% EtOAc-hexane) afforded 0.051 g (5 1%) of cyclopentenone 318 as a

pale yellow oil.

IR (film): 2930, 1705, and 1610 cmf

'H NMR (500 MHz, CDC 3 ): 2.88 (m, 1H), 2.77 (dd, J= 18.6, 6.4 Hz, 1H), 2.15(s, 3H), 2.14 (dd, J= 18.6, 2.2 Hz, 1H), and 1.25 (d,J= 7.0 Hz, 3H)

'3C NMR (125 MHz, CDC3 ): 200.4, 177.3, 122.9, 41.7, 38.3, 18.8, and 16.8

174

wdd 0 I £ '7 S 9 I II 9 6

I I I I

)J

SKC

Jg

0

Etp9 wdd -0

Documents

SYNTHESIS AND CHEMISTRY OF (TRIALKYLSILYL)VINYLKETENES