UvA-DARE (Digital Academic Repository)
Palladium mediated synthesis of N-heterocycles by iminoannulation of allenes.
Diederen, J.J.H.
Publication date
2001
Document Version
Final published version
Link to publication
Citation for published version (APA):
Diederen, J. J. H. (2001). Palladium mediated synthesis of N-heterocycles by iminoannulation
of allenes.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
• wart ':•
m
i •,;..
U S i
wmm
sût l l l i
ffi^mSÊmÊ-••V •'!'•' ' K&ttSK
—
"••''•"••>'•'.'•
m»ll
ilü
ffi
A01
113
2e ex
ladium Mediated Synthesis of
erocycles by Iminoannulation of
Aliènes
Palladium Mediated Synthesis of
N-Heterocycles by Iminoannulation of
Aliènes
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam
op gezag van de Rector Magnificus Prof. Dr. J.J.M. Franse
ten overstaan van een door het college voor promoties ingestelde commissie in het openbaar te verdedigen in de Aula der Universiteit
op vrijdag 11 mei 2001 te 12.00 uur
door
Jeroen Jan Hubert Diederen
Promotiecommissie:
Promotoren: Prof. Dr. H. Hiemstra Prof. Dr. K. Vrieze
Overige leden: Prof. Dr. C.J. Elsevier Dr. H-.W. Frühauf Dr. M. Pfeffer Prof. Dr. J. Reedijk
Prof. Dr. P.W.N.M. van Leeuwen
Faculteit der Natuurwetenschappen, Wiskunde en Informatica
Universiteit van Amsterdam
Dankwoord
Ruim vier jaar academisch onderzoek staat in dit boekje beschreven. Uiteraard heb ik dit resultaat niet geheel aan mezelf te danken gehad en daarom zou ik een aantal mensen hier willen noemen.
Allereerst wil ik mijn promotor Kees Vrieze bedanken voor de mogelijkheid die hij me heeft gegeven om te beginnen met academisch onderzoek waarbij hij me later volledig vrij heeft gelaten in de invulling daarvan. Tijdens onze bijeenkomsten waren we het niet altijd met elkaar eens, echter, jouw gevoel van humor heb ik altijd gewaardeerd. Kees, ik bedank je ook voor het nakijken van het manuscript. Mijn promotor Henk Hiemstra ben ik zeer dankbaar voor de begeleiding die hij mij gaf in de tweede helft van mijn promotieonderzoek. Henk, jouw gedrevenheid en onze leuke bijeenkomsten samen met Willem-Jan in jouw kamer zal ik niet snel vergeten. Ik ben je ook zeer dankbaar voor het tot in de puntjes nakijken van het manuscript. Hans-Werner, ik bedank jou voor een prettige samenwerking tijdens onze bijeenkomsten samen met Boke en het nakijken van mijn manuscript.
Michel Pfeffer m'a donné la possibilité de travailler pour trois mois au Laboratoire de Synthèses Métallo-Induites, à l'Université Louis Pasteur, Strasbourg. Michel, je te remercie pour toute ton aide avec la chimie et la cyclopalladation en particulière.
Hoofdvakstudent Eric Zijp en Hbo-studenten René Sinkeldam en Wouter van der Meulen ben ik erg dankbaar voor al het werk dat zij voor mij gedaan hebben. Eric, jouw inzet, ook al zat het heel vaak tegen, heeft tot leuke resultaten geleid die zijn vermeld in Hoofdstuk 2 en 4. René, jouw opgeruimde zuurkast en karakter straalde uit naar je chemie, wat leidde tot een hele hoop zuivere producten die ik heb gebruikt in Hoofdstuk 3. Wouter, ook jouw inzet heeft tot leuke resultaten geleid in Hoofdstuk 3 en bood een opstapje voor het katalytische werk. Helemaal in het begin heeft Bas Eeltink, een Mbo-student, mij bijgestaan met reacties van keteengas en verschillende palladium-koolstof complexen. Helaas heeft dit niet geleid tot iets bruikbaars. Bas, hartelijk bedankt voor jouw niet te weerhouden inzet.
Willem-Jan Karstens wil ik bedanken voor de prettige samenwerking op het "organische" vlak. Jouw discussies met Henk over chemie tijdens onze bijeenkomsten waren erg aanstekelijk.
Van mijn vier jaar die ik heb doorgebracht op de anorganische afdeling heb ik, zeker de laatste twee jaar, erg genoten. In het begin waren Jos Delis en Hans Groen mijn voorgangers en ik wil hun bedanken voor de geboden hulp.
Later werd het echt enorm gezellig toen Anouk, Boke, Marcel (Cor), Martijn, Sander en Wim in het team kwamen werken. Aan onze Roetertoeter-avonden heb ik leuke herinneringen.
Ook Marcel Duin, Jens Barkemeijer, Sander Gaemers, Hans Donkervoort, Nicolette Rot, Maarten Bakker, Peter Groen, Joris van Slageren en Frank Vergeer, de overige leden van de club, wil ik bedanken voor de fijne samenwerking. Maarten, ik bedank jou voor een mooie zeilherinnering. Frank, de Beaujolais, de erwtensoep (de spoelbak rules) en onze etentjes zullen me altijd bijblijven.
Jan-Meine en Sander ben ik dank verschuldigd voor de hulp bij het gebruik van de NMR-apparatuur en het verkrijgen van vele mooie 2D-platen. Taasje, bedankt voor jouw hulp en jouw nuchtere kijk op de zaken. De mensen van de instrumentmakerij, glasblazerij en electronische dienst ben ik ook dank verschuldigd. Henk en Tjerk de "Neus", jullie stonden altijd gelijk klaar als ik iets moest hebben. Tjerk, drinkebroeder op vrijdagavond, ik heb erg veel gelachen daar aan de bar. Cees Bergfeld, jouw inzet in 't magazijn mag ook niet ongenoemd blijven.
Jan Fraanje, Kees Goubitz, Ton Spek en Wilbert Smeets ben ik erkentelijk voor de kristallografische bepalingen.
Ontzettend opgelucht was ik toen mijn iMac™ mijn proefschrift weer op de juiste manier weergaf na een aantal updates van Rob Balk.
Winfred wil ik bedanken voor het ontspanwerk aan de UvA. Wij hebben heel wat mooie avonden doorgebracht in de Roeter en in Amsterdam.
Mijn ouders ben ik zeer dankbaar voor de morele en financiële steun die ze me de afgelopen jaren tijdens mijn studie en mijn promotie hebben gegeven.
Tenslotte wil ik jou, lieve Catherine, enorm bedanken voor je steun tijdens die soms zware tijd. Naast een moeilijke start aan de UvA had ik ook erg veel problemen met mijn toenmalige huisbazin aan de Ringdijk. Jij hebt me door die hel heengeloodst. Gelukkig is er meer dan het asociale onderzoek.
Jeroen Leiden, februari 2001
CONTENTS
CHAPTER 1 PALLADIUM MEDIATED SYNTHESIS OF CARBO- AND
HETEROCYCLES
1.1 Synthesis of Heterocycles by Palladation of Alkenes 11
1.2 7t-Allylpalladium Chemistry 14 1.3 Synthesis of Carbo- and Heterocycles via 7t-Allylpalladium Chemistry
involving Aliènes 17
1.4 Purpose of the Investigation 24
1.5 Outline of this Thesis 25
1.6 References 27
CHAPTER 2 SYNTHESIS OF N-HETEROCYCLES BY IMINOANNULATION OF
CYCLOPALLADATED IMINES
2.1 Introduction 30 2.2 Results and Discussion 32
2.2.1 Synthesis of Cyclopalladated Complexes 32 2.2.2.1 Reaction of Cyclopalladated Complexes with 1,1-Dimethylallene 36
2.2.2.2 Characterization of Iminium Salts 4a-d,f,h 37
2.2.2.3 Regioselectivity 37
2.2.2.4 Intermediates 39
2.2.2.5 Kinetics 40
2.2.3 Reaction of Cyclopalladated Complexes with Vinylidenecyclohexane 41
2.2.4 Synthesis of neutral N-Heterocycles from Iminium Salts 42
2.3 Conclusions 43 2.4 Experimental Section 45
2.5 Acknowledgement 65 2.6 References 66
CHAPTER 3 SYNTHESIS OF ISOQUINOLINES AND PYRIDINES FROM O-HALIDE
SUBSTITITED ARYL AND VINYL IMINES VIA PALLADIUM-CATALYZED IMINOANNULATION
3.1 Introduction 70
3.1.2 Allenes 72
3.2 Results 75
3.2.1 Stoichiometric Reactions 75
3.2.1.1 Synthesis of Bromo bridged Palladium Dimers of
o-Bromo-Acetophenone Ketimines 75 3.2.1.2 Synthesis of Isoquinolinium Salts 4 76
3.2.2 Catalytic Reactions 78
3.3 Product Selectivities with different Aliène Substitution Patterns 85
3.3.1 1,1-Disubstituted Aliène 85 3.3.2 Mono-substituted Aliènes 88 3.3.3 1,3-Disubstituted Aliènes 89 3.4 Conclusions 90 3.5 Experimental Section 92 3.6 Acknowledgement 115 3.7 References 116
CHAPTER 4 SYNTHESIS OF HETEROCYCLES BY INTRAMOLECULAR OXIDATIVE
IMINATION OF ALLENES AND ALKENES, AND CYCLIZATION OF O-BUTADIENYLBENZYL ALCOHOLS
4.1 Introduction 120
4.1.1 Oxidative Amination of Alkenes mediated by Palladium 120 4.1.2 Oxidative Amination and Hydroamination of Aliènes 122
4.2 Results 124
4.2.1 Oxidative intramolecular Imination Reaction on o-Allylbenzaldimines 124 4.2.2 Intramolecular oxidative Annulation Reactions of
o-AUenyl(methyl)-Benzaldimines 126 4.2.3 Intramolecular oxidative Imination Reactions of
o-Allenyl(methyl)-Benzaldimines 129 4.3 Discussion 131 4.3.1 Oxidative intramolecular Imination Reaction on o-Allylbenzaldimines 131
4.3.2 Intramolecular oxidative Imination Reactions of
o-AUenyl(methyl)-Benzaldimines 134
4.4 Experimental Section 135
4.5 Acknowledgement 148 4.6 References 148
SUMMARY 1 5 1
Chapter 1
Palladium mediated Synthesis of Carbo- and Heterocycles >
1.1 Synthesis of heterocycles by palladation of alkenes
Palladium has proven to be an excellent auxiliary element in the synthesis of a wide variety of heterocycles and carbocycles.[l] The tolerance of palladium reagents to many functional groups such as carbonyl and hydroxyl groups has greatly contributed to the utility of this transition metal. Pd-catalyzed reactions can be carried out without protection of these functional groups. Divalent palladium complexes are able to coordinate to alkenes with formation of jt-complexes. As a result of the electrophilic character of the palladium center, the electron density of the coordinated alkene is decreased and it may react with various nucleophiles.[l] The attack of nucleophiles with formation of a Pd-C bond is called the palladation reaction (Scheme 1). These palladium-alkene complexes may react with a variety of nucleophiles to give either nucleophilic substitution or nucleophilic addition (Scheme 1), depending on the reactants and conditions. Typical nucleophiles are water, alcohols, carboxylic acids, ammonia, amines, enamines and active methylene compounds.[2] The intramolecular version of this reaction is a useful way to synthesize heterocyclic compounds.
PdCI 2 XH . / YH : Nurleonhilir addition . / R R CIPd \ r / \ palladation - H C l CIPd X -HCl;-Pd(0) Y X - Pd(0) -HCl 1
Nucleophilic substitution via ß-hydrogen éliminât
_ /R XH, YH = nucleophiles, H20 ,
^ ROH, RC02H, RNH2, CH2E2
A
Palladium mediated Synthesis ofCarbo- and Heterocycles
The oxidation of ethylene by PdCl2 in water to give acetaldehyde has been known for
100 years, and is known as the Wacker Process.[3, 4] This reaction is a nice example of oxypalladation of ethylene, followed by ß-elimination to give a vinyl alcohol A, which stays bound to palladium. Rearrangement of the complex leads to B, which forms the aldehyde by ß-elimination (Scheme 2).
OH OH O—H O
PdC,2 - H C l C I P d ^ H ld H C | PdC, - M O )
A B
Scheme 2
Oxidation of ethylene in alcohol with PdCl, in the presence of base gives an acetal and vinyl ether (eq 1).[5, 6]
R 2R'OH R R
>r
+
J=
(1)
T -
2HC1
-Pd(0) Ü H R'O ^R. R'O PdCI2 majorThis method was applied to the synthesis of brevicomin in an intramolecular reaction (eq2).[7-9]
,0 cat. PdCl2, CuCl2
(2)
OH DME, 45% V O
Besides alkyl alcohols, phenolic oxygen reacts similarly in an intramolecular oxypalladation. 2-Allylphenol reacts to give the exo cyclized benzofuran with Pd(OAc)2 (eq 3),
but in the case of PdCl2, the alcohol cyclizes in an endo fashion to give the benzopyran (eq 4).[10]
Pd(OAc),
Clmpter 1
PdCl,
(4)
^
O'
Unsaturated carboxylic acids afford lactones in a ring closure reaction, as shown by Larock et al. [11] The reaction is carried out in the presence of a stoichiometric amount of PdCl2(MeCN)2 to produce the isocoumarin (eq 5). If, however, the reaction is done with a
catalytic amount of Pd(OAc)2 in the presence of molecular oxygen and DMSO, a system that
was used by Hiemstra and coworkers,[12, 13] the Z-phthalide is obtained. Apparently, in this case a Tt-allylpalladium complex is involved (eq 6).[11]
CO2H 1 equiv. PdCl2(MeCN)2 *-Na2CQ3 (5) ^ ^ C 0 2 H 10% Pd(OAc)2 NaOAc 02, DMSO (6)
Nitrogen nucleophiles are also able to react similarly with alkenes. An intermolecular reaction of aliphatic amines and alkenes in the presence of palladium salts is a difficult process because of the strong coordinating capacities of aliphatic amines towards divalent palladium. In a few exceptional cases, alkenes react in a so-called aminopalladation reaction. Treatment of 1-decene with methylamine in the presence of a stoichiometric amount of PdCl2, produces
aziridines. Initial complexation of the palladium to the double bond of 1-decene and subsequent nucleophilic attack of methylamine leads to a o-Pd complex. The secondary amine reacts subsequently intramolecularly to eliminate Pd(0) and the hydrochloride salt of methylamine (eq 7).[14] C«H 17 + MeNH2 PdClp C8H17 PdCI NHMe CsH-17 HCl Pd(0) ^ 7 N I Me (7)
Palladium mediated Synthesis ofCarbo- and Heterocycles
Unlike intermolecular reactions, intramolecular aminopalladation reactions proceed with greater ease. As an example, allylaniline cyclizes in an exo fashion to produce 2-methylindole which shows again that 5-membered ring formation is easier than 6-membered ring formation (eq 8).[15]
^ Y ~ ^ PdCl
2(MeCN)
2^ r ^ V A _
(8)^ ^ NH2 p-benzoquinone M
As was seen in the latter example of aminopalladation, nucleophilic attack takes place at the electronically favored position. Endo cyclization of 2-(3,3-dimethylallyl)aniline leads to dimethyl-l,2-dihydroquinoline, which shows again that regioselectivity is governed by electronic factors (eq 9).[16]
PdCl2(MeCN)2
p-benzoquinone ^ - ^ N' H
(9)
1.2 7t-Allylpalladium chemistry
In many cases, n-allylpalladium chemistry is involved in the formation of heterocycles and carbocycles, as is demonstrated by the great amount of publications.[1] 7t-AUylpalladium intermediates can be generated from alkenes, 1,2-dienes (aliènes) and 1,3-dienes starting from either Pd(II) complexes or Pd(0) complexes. 7t-Allylpalladium complex formation from alkenes takes place by the displacement of an allylic hydrogen of the alkene with Pd(II). The complex may then further react with a nucleophile (carbanion, heteroatom, organotin) with formation of Pd(0) (Scheme 3).
Chapter 1 PdCl2 R fc--HCl PdCI/2 NuH
»•
-HCl - Pd(0) PdCl/ : Nu NuH -HClf^] - Pd(0)
R PdCI/2 Nu Scheme 3Reaction of a 1,2-diene (aliène) with PdCl2 may produce thermodynamic product B,
formed by insertion of aliène into the Pd-Cl bond (Scheme 4).[17] The kinetic product A, formed by halide attack at the terminal carbon of the coordinated aliène, may be trapped by insertion of excess aliène into the palladium-vinyl bond to form a dimeric complex which reacts with chloride to give a dimeric structure. If B reacts with chloride, a monomeric structure is found.
1 CI" *- % / ^ n i = * = ,
r
ci
CI' - ^Y^n
PdCl2 * PdCl/ 2 A kinetic adductr
ci
^KxCi
* PdCl/ 2 A kinetic adduct PdCl/ 2 CI PdCl/ 2or
- Pd(0) CI B th( ad ?rmodynamic duct Scheme 4Reaction of 1,3-dienes with Pd(II)-complexes in the presence of a nucleophile leads to the introduction of the nucleophile at the terminal carbon atom of the conjugated diene (eq 10). When butadiene is treated with PdCl2, the 1-chloromethyl-it-allylpalladium complex is formed
by chloropalladation. Nucleophiles react at the unsubstituted end of the rc-allyl-palladium complex to produce allylic esters, halides or ethers.
Palladium mediated Synthesis ofCarbo- and Heterocycle,
^
^
PdCl2 /X-PdCl/ PdCl/ ; • Pd(0) Nu = Cl, OR, OAc
'X (10)
1,4-Difunctionalization with nucleophiles has wide synthetic applications. [18-20] The oxidative diacetoxylation of butadiene with Pd(OAc)2 affords l,4-diacetoxy-2-butene (A in eq
11) and l,2-diacetoxy-3-butene (B in eq 11). The latter may be isomerized to the former. An industrial process has been developed based on this reaction. 1,4-Butanediol and THF are produced commercially from l,4-diacetoxy-2-butene.
< ^ ^ AcOH •*• AcO ^ ^ ^ O A c +
Preparation of 7t-allylpalladium complexes by the oxidative addition of allylic compounds (esters, carbonates etc), and their reactions with nucleophiles are catalytic, because Pd(0) is regenerated after the reaction with the nucleophile, and reacts again with the allylic compounds (Scheme 5). The stereochemistry of the Pd-catalyzed allylation of nucleophiles has been studied extensively. [21, 22] In the first step, the n-allylpalladium complex is formed by
anti-attack of the palladium on the allylic carbon atom to give inversion of configuration. Then
subsequent reaction of soft carbanions, N- and O-nucleophiles, proceeds by inversion to give overall retention of configuration (Scheme 5). Reactions of hard carbanions of organometallic compounds (RMgX, RSnR'.,, RZnX, RB(OH)2) proceed via transmetallation. In this case overall
inversion is observed (Scheme 5).
Ri-OAc + Pd(0) inversion (OAc)/ 2 RM R1v -* Pd(OAc)/ 2 soft NuH -HOAc retention Pd(OAc)/ 2 -HOAc inversion Scheme 5 Nu
Chapter 1
1.3 S y n t h e s i s of carbo- and h e t e r o c y c l e s via 7t-allylpalladium c h e m i s t r y i n v o l v i n g a l i è n e s
The 1,2-diene or aliène moiety is an interesting functionality because of the specific stereochemical features of its two orthogonal cumulated double bonds and the associated reactivity.[23] N u m e r o u s preparative methods for allenes, functionaliz»d or not, are available.[24] Because it is well known that organopalladium compounds will add to aliènes to produce 7t-allylpalladium complexes[25-27] (Chapter 1.2), many publications have appeared about the application of this very useful reaction. One of the pioneering groups who used aliènes as synthetic building blocks was that of Cazes and coworkers. Their papers are about the carbopalladation of aliènes leading to functionalized 1,3-dienic (or styryl) compounds in a three component reaction (Scheme 6).[28] Oxidative addition of a vinyl (or aryl) halide to a Pd(0) complex first generates a vinyl (or aryl) palladium complex I. In the second step, the aliène inserts into the newly formed a-Pd-C bond to generate a Pd-ir-allyl complex II. Finally, attack of the carbon nucleophile leads to the formation of functionalized dienic compounds III a n d / o r IV.
Pd(0)Ln
t
Pd(dba)2 + dppe
Scheme 6
Tsuji's group described a related synthesis of dienic amines, based on the same methodology with pyrolidine as the nucleophile (eq 12).[29] The latter reaction was further extended with other amines and alkyl halides in our group. The major drawback in this type of reactions seems that nucleophilic attack of most amines on alkyl halides is a faster process than oxidative addition of the alkyl halide to Pd(0) resulting in alkylated amines. [30]
Palladium mediated Synthesis ofCarbo- and Heterocycles
O
NH
( ' \ ^^ V
H'—\ 5% Pd(OAc)2 5% dppe MeCN, 100 °C, 3hO
Grigg and coworkers showed that intramolecular carbocyclization of aliènes leads to the formation of carbo- and heterocycles and an excellent review has appeared about this methodology (Scheme 7,8 and 9).[31] Oxidative addition of the aryl halide to a Pd(0) complex, formed after in situ reduction of Pd(OAc)2 by PPh3, forms a Pd-aryl species (Scheme 7).
Cyclization takes place at the central carbon of the aliène to form a Pd-rc-allyl complex which reacts further with secondary amines, under basic conditions.[31] If K2C03 is used as a base
(condition 1), nucleophilic attack takes place at the sterically favored position, whereas in the case of Ag2CG*3 as base (condition 2), the amine reacts at the most substituted position of the
allyl terminus (kinetic product). The regioselectivity with Ag2C03 as the base is ascribed to a
cationic Pd(II) intermediate which reacts fast with the nucleophile to give the electronically favored kinetic product. The xc-allyl species generated by cyclization of aryl halides onto proximal aliènes can react with a wide range of nucleophiles (azides, sulphinates) to give functionalized heterocyclic products.
PdILn HN. J^ ^ O condition 1 NMe .... . „ „ _ or condition 1: K2C03 condition 2: Ag2CÖ3 Scheme 7
Chapter 1
Allenes were used to create allylic amines (Scheme 8). Heck type cyclization of the Pd-aryl species creates a o-Pd-C bond which inserts allene to provide a Pd-rc-allyl species which is trapped by secondary amines. In this way it is possible to create 5- and 6-membered ring structures.
h
Pd(OAc)2 -Y pph3 PdLn = . = PdLn NR,R2 Scheme 8 HNRjRzGrigg and coworkers found in certain reactions that, if aliènes are used and no nucleophile is present to trap the Pd-n-allyl species formed, the reaction ends with a ß-hydrogen elimination to produce a diene. They applied the ß-hydrogen elimination in a 3-component cascade reaction, terminating in a Diels-Alder reaction using N-methylmaleimide as dienophile (Scheme 9). R Pd(OAc)2 / PPh3 0 ;
J
\[f NMe
0 , î ~ » ^ r - ' ' \"i R' ("'' '\l R'"•V -
J^~
i ^ R
slow ß-elimination ^-**x
. A ^ ,R
1\
H
- H P d l - Ï / \fS,
PdlLn Scheme 9Palladium mediated Synthesis ofCarbo- and Heterocycles
Gallagher and coworkers described cyclization reactions on y-allenic amines leading to pyrolidines (eq 13). [32]
R2
J
U3)
NHR' Pd(PPh3)4 ^ N ;
K2C 03 R R
An elegant procedure to cyclize y-hydroxyallenes in the presence of aryl halide, a catalytic amount of Pd(PPh3)4 and base with incorporation of CO was published by Walkup et
al. (eq 14).[33] Cyclizations without in situ carbonylation were described in an earlier paper by
the same group.[34]
10%(PPh3)4
V_x _ ^ £ ^
„
\—V CO (1 bar)
5equiv DMF, 55-60 °C
12-18 hr cis-trans (+/-25:75)
Allene-substituted lactams were cyclized to bicyclic enamides with unprecedented nucleophilic attack of the amide nitrogen atom on the central carbon of the aliène unit (eq 15).[35] 4 Phi 10%Pd(PPh3)4 r K2C03, TBAC1 MeCN NH
Liebeskind and coworkers published a method to synthesize various A'-carbapenems by a cyclization of 4-allenylazetidones in the presence of different activated olefins in excess, mediated by a stoichiometric amount of palladium chloride (Scheme 10).[36] It is presumed that these transformations proceed via palladium(II) induced N-C bond formation to form a palladium-vinyl species which reacts with the olefin in a Heck-type reaction. After ß-hydrogen elimination the A'-carbapenem is produced.
Cliapter 1 TBDMSO | VÇ' TBDMSO Et3N - Et3N.HCl CH2C12
W^r
E
-
Et3N TBDMSO , . I H H /o
E = COOEt, COCH3, CHO, CN TBDMSO PdCI ^ — N ^ -Et3N.HCl O - Pd(0) Scheme 10Many intermolecular reactions of allenes and heteroatom or carbon nucleophiles were described by Larock and coworkers (eq 16 and 17).[37, 38] It was found that the regioselectivity was dependent on the ring size. Formation of 5-membered rings mainly involved cyclization at the more substituted end of the allene (eq 16) whereas 6-membered rings were formed from intramolecular cyclization of the nucleophiles at the less substituted end of the allene (eq 17). This chemistry was later extended to larger ring N-heterocycles (7- and 8-membered).[39] An enantioselective version of this reaction type was performed by choosing bisoxazoline ligands, developed by Pfaltz and others[40-42], to produce enantioselectively different hetero- and carbocycles with ee's up to 82%. [43]
XH 1 (or Br) R N a2C 03 TBAC1 or LiCl cat. Pd(0) DMF, A (16)
Palladium mediated Synthesis ofCarbo- and Heterocycles XH I (or Br)
==ƒ
N a2C 03 TBAC1 or LiCl cat. Pd(0) DMF, A X = O, NR, COO, C(C02Et) (17)Other groups used the same methodology for the synthesis of heterocyclic products. Mérour and coworkers published a palladium catalyzed heteroannulation with aliènes and 3-iodo-2-aminopyridines yielding 3-methylene-pyrrolo[2,3-b]pyridine derivatives (eq 18).[44]
,R2 ~ ^ T v ' -NHR N a2C° 3 A N H H 1 5 mol% Pd(OAc)2 5 mol% PPh3 DMF X = N, CH VX N Ri (18)
The group of Alper showed that Larock's reactions of o-iodophenols could be extended with an extra carbonylation step producing l-benzopyran-4-ones (eq 19).[45]
^ _ / ,
20 H Na2C03or(i-Pr)2NEt
5% Pd(OAc)2 ;5% dppb
benzene
(19)
Benzoquinolizinium derivatives were synthesized in the group of Pfeffer by reacting cyclopalladated pyridines with 1,1-dimethylallene (eq 20).[46]
Î ^ <
^ N PdCI/2 CH2C12 ; A - Pd°
Clmpter 1
Bäckvall's group reported on the 1,2-oxidation of aliènes by LiBr in the presence of a catalytic system based on Pd(OAc)2 and p-benzoquinone in HOAc leading to 1,2-dibromides
(Scheme 11).[17] In the mechanism, the aliène first coordinates to the electrophilic palladium center. Then nucleophilic attack of the bromide on the central carbon of the aliène produces a n-allylpalladium complex with p-benzoquinone as a stabilizing ligand. Nucleophilic attack of a second bromide on the n-allylpalladium complex produces the dibromide together with reduced palladium complex. The palladium is subsequently reoxidized by p-benzoquinone in the presence of HOAc to the catalytically active Pd(II)-species.
Pd(ll)
Scheme 11
Extension of the latter reaction to an intramolecular version was also reported by the group of Bäckvall. y-Allenic acids cyclized under similar reaction conditions to y-lactones with mainly Z stereochemistry (eq 21).[47, 48]
n OH cat. Pd(OAc)2 LiBr, LiOAc p-benzoquinone HOAc, 40 °C (21)
Palladium mediated Synthesis ofCarbo- and Heterocycles
Recently, azetidines and tetrahydropyridines were synthesized via a palladium catalyzed cyclization of enantiopure allenes (eq 22). [49]
^ ^ .*R R!X Pd(PPh3)4 R1 -|R or H f \ k '/ N -N- vR K2CO3 I I M e C N P (22)
1.4 Purpose of the investigation
Many reactions of amines, alcohols or carboxylic acids as nucleophiles with K-allylpalladium complexes or activated alkenes are known. The main goal of the investigation described in this thesis is to probe, whether imines can be used as nucleophiles in an intramolecular annulation reaction to produce N-heterocycles, which may be further functionalized.
Chapter 1
1.5 Outline of this thesis
In Chapter 2 the synthesis of several iminium salts from cyclopalladated a-tetralone ketimines and 1,2-dimethylallene or vinylidene cyclohexane will be described. Iminium salts 4, synthesized from cyclopalladated a-tetralone ketimines and 1,2-dimethylallene with high regioselectivities, reacted with KOH to enamines 8, or with NaBH4 to amines 9 (Scheme 12).
F
PF6-,+
9?
Scheme 12
In Chapter 3 , a new iminoannulation catalyzed by palladium will be described. Different aryl- or vinylimines were reacted with mono-substituted aliènes in the presence of a palladium catalyst and base to produce various pyridines and isoquinolines (eq 23). In the case of N -benzyl substituted imines, the N-heterocyclic products were deprotected by hydrogenolysis of the benzyl group. Tert-butyl groups could be removed under basic conditions and high temperatures via a ß-elimination step. However, the latter reaction was found to be
Palladium mediated Synthesis ofCarbo- and Heterocycles
difficult to perform, therefore we introduced a propane nitrile group on the imine, which enormously facilitated the deprotection step.
.R Pd cat. X = Br, I R = Bn, i-Bu, CH2CH2CN R' = n-Bu, f-Bu, Ph, Cy (23)
In Chapter 4 reactions of o-allyl-substituted arylimines with palladium salts will be described. Extension of this reaction towards vinyl imines was impossible due to our inability to introduce an allyl moiety at the vinylic position. Furthermore, intramolecular annulation reactions of o-allenylmethylarylimines or o-allenylmethylbenzylalcohol in the presence of an electrophile (RX) catalyzed by palladium will be described leading to tetrahydroisoquinolinium salts or tetrahydropyrans. The iminium salts reacted with a variety of Grignard reagents and NaBH4 to produce different tetrahydroisoquinolines (Scheme 13).
RX Pd cat. M+ X" Rj = H, Me RX Pd cat. base Scheme 13
Chapter 1
1.6 R e f e r e n c e s
[I] J. Tsuji, Palladium Reagents and Catalysts - Innovations in Organic Synthesis, Wiley, Chichester 1995.
[2] J. Tsuji, Ace. Chem. Res. 2 (1969) 144.
[3] W. Hafner, R. Jira, J. Sedlmeier, J. Smidt, Chem. Ber. 95 (1962) 1575. [4] R. Jira, J. Sedlmeier, J. Smidt, Liebigs Ann. Chem. 693 (1966) 99.
[5] 1.1. Moiseev, M. N. Vargaftik, Y. K. Syrtik, Dokl. Akad. Nauk SSSR 133 (1960) 377. [6] E. W. Stern, M. L. Spector, Proc. Chem. Soc. (1961) 370.
[7] N. T. Byrom, R. Grigg, B. Kongkathip, Chem. Commun. (1976) 216.
[8] N. T. Byrom, R. Grigg, B. Kongkathip, J. Chem. Soc. Perkin Trans. (1984) 1643. [9] K. Mori, Y. B. Seu, Tetrahedron 41 (1985) 3429.
[10] T. Hosokawa, S. Miyagi, S. Murahashi, A. Sonoda, J. Org. Chem. 43 (1978) 2752. [II] R. C. Larock, T. R. Hightower, ƒ. Org. Chem. 58 (1993) 5298.
[12] R. A. T. M. Benthem van, H. Hiemstra, W. N. Speckamp, ƒ. Org. Chem. 57 (1992) 6083. [13] R. A. T. M. Benthem van, H. Hiemstra, J. J. Michels, W. N. Speckamp, ƒ. Chem. Soc. Chem.
Commun. (1994) 357.
[14] L. S. Hegedus, K. Siirala-Hansen, J. Am. Chem. Soc. 97 (1975) 1184.
[15] L. S. Hegedus, Comprehensive Organic Synthesis, Vol. 4, Pergamon Press, Oxford 1991. [16] L. S. Hegedus, J. M. McKearin, J. Am. Chem. Soc. 104 (1982) 2444-51.
[17] J. E. Bäckvall, C. Jonasson, Tetrahedron Lett. 38 (1997) 291-94. [18] J. E. Bäckvall, Ace. Chem. Res. 16 (1983) 335.
[19] J. E. Bäckvall, Pure Appl. Chem. 55 (1983) 1669. [20] J. E. Bäckvall, Neiv J. Chem. 14 (1990) 447.
[21] B. Âkermark, A. Jutand, ƒ. Organomet. Chem. 217 (1981) C41. [22] B. M. Trost, L. Weber, ]. Am. Chem. Soc. 97 (1975) 1611.
[23] S. Landor, The chemistry of aliènes, Academic Press, London 1982.
[24] L. Brandsma, H. D. Verkruijsse, Studies in Organic Chemistry 8, Synthesis of Acetylenes,
Aliènes and Cumulenes, Elsevier, Amsterdam 1981.
[25] D. Medema, R. Helden van, C. F. Kohll, lnorg. Chim. Acta 3 (1969) 255. [26] R. Medema, R. Helden van, Reel. Trav. Chim. Pays-Bas 90 (1971) 304.
[27] R. Helden van, C. F. Kohll, G. Medema, T. Verberg, T. Jonkhoff, Reel. Trav. Chim.
Pays-Bas 87 (1968) 961.
[28] B. Cazes, Pure & Appl. Chem. 62 (1990) 1867-78. [29] I. Shimizu, J. Tsuji, Chem. Lett. (1984) 233-236.
Palladium mediated Synthesis o/Carbo- and Heterocycles
[31] R. Grigg, V. Sridharan, ƒ. Organomet. Chem. 576 (1999) 65-87. [32] I. W. Davies, D. I. C. Scopes, T. Gallagher, Synlett (1993) 85.
[33] R. D. Walkup, L. Guan, Y. S. Kim, S. W. Kim, Tetrahedron Lett. 36 (1995) 3805-8. [34] R. D. Walkup, M. D. Guan, S. W. Mosher, S. W. Kim, Y. S. Kim, Synlett (1993) 88-90. [35] W. F. J. Karstens, F. P. J. T. Rutjes, H. Hiemstra, Tetrahedron Lett. 38 (1997) 6275-78. [36] J. S. Prasad, L. S. Liebeskind, Tetrahedron Lett. 29 (1988) 4257-60.
[37] R. C. Larock, Y. He, W. W. Leong, X. Han, M. D. Refvik, J. M. Zenner, J. Org. Chem. 63 (1998) 2154-60.
[38] R. C. Larock, N. G. Berrios-Pena, ƒ. Org. Chem. 56 (1991) 2615-17. [39] R. C. Larock, C. Tu, P. Pace, ƒ. Org. Chem. 63 (1998) 6859-66.
[40] R. E. Lowenthal, A. Abiko, S. Masamune, Tetrahedron Lett. 31 (1990) 6005-8. [41] D. Muller, G. Umbricht, B. Weber, A. Pfaltz, Helv. Chim. Acta 74 (1991) 232-40. [42] G. Helmchen, A. Krotz, K. T. Ganz, D. Hansen, Synlett (1991) 257-8.
[43] R. C. Larock, J. M. Zenner, ƒ. Org. Chem. 60 (1995) 482-83. [44] E. Desarbre, J.-Y. Mérour, Tetrahedron Lett. 37 (1996) 43-46. [45] K. Okuro, H. Alper, ƒ. Org. Chem. 62 (1997) 1566-67.
[46] J. Chengebroyen, C. Sirlin, M. Pfeffer, Tetrahedron Letters 37 (1996) 7263-66. [47] C. Jonasson, J.E. Bäckvall, Tetrahedron Lett. 39 (1998) 3601-4.
[48] C. Jonasson, A. Horvâth, J.E. Bäckvall, J. Am. Chem. Soc. 122 (2000) 9600-9.
[49] F. P. T. J. Rutjes, K. C. M. F. Tjen, L. B. Wolf, W. F. J. Karstens, H. E. Schoemaker, H. Hiemstra, Org. Lett. 1 (1999) 717-20.
[50] J. J. H. Diederen, M. Pfeffer, H.-W. Frühauf, H. Hiemstra, K. Vrieze, Tetrahedron Lett. 39 ( 1998)4111-4.
[51] J. J. H. Diederen, R. W. Sinkeidam, H.-W. Frühauf, H. Hiemstra, K. Vrieze, Tetrahedron
Chapter 2
Synthesis of N-Heterocycles by Iminoannulation of
Cyclopalladated Imines[i]
Abstract
Several imines 1 (R=Ph, p-Tol, p-An, p-Cl-C6H4, p-Br-C6H4, p-I-QH4/ p-N02-C6H4, CH2Ph, i-Pr)
were synthesized from a-tetralone by a condensation reaction with different amines. Cyclopalladation of imines 1 with Pd(OAc)2 and subsequent substitution of the acetate bridge
by chloride gave chloro-bridged palladium complexes 3. Reaction of 3 with 1,1-dimethylallene and vinylidenecyclohexane produced iminium salts 4 and 6, respectively, with high regioselectivities and in good yields. Iminium salts 4 were converted into enamines 8 by deprotonation with KOH or into amines 9 by reduction with NaBH4.
Synthesis of'N-Heterocycles by Iminoannulation of Cyclopalladated Imines
2.1 Introduction
Palladium complexes have found numerous applications in organic synthesis.[2] Most importantly, palladium offers many possibilities of carbon-carbon and carbon-heteroatom bond formation. One important property of palladium complexes is the ease of formation of a-Pd-C bonds either by oxidative addition, insertion or transmetallation.
Insertion of a Pd(II)-species into a C-H bond with a stabilizing intramolecularly coordinating heteroatom, has often led to the formation of aa-Pd-C bond (eq 1). This so-called cyclopalladation reaction (or orf/zo-palladation in the specific case of reactions involving the orf/ïo-position in the aromatic ring of the ligand), has been carried out with a large number of substrates and excellent reviews about this subject are available.[3-5]
PdX, A = Heteroatom X = CI, Br, OAc .PcT
V
(1)In our laboratory it has been shown that aliènes insert into a a-Pd-C bond in such a way that migration of the nucleophilic a-bonded carbon substituent on palladium exclusively occurs to the central aliène carbon atom leading to a Pd-allyl species (eq 2).[6, 7] These Pd-allyl complexes are interesting intermediates for the formation of carbon-carbon and carbon-nitrogen bonds by reacting them with a wide range of nucleophiles (hard and soft carbanions, heteroatoms, organotins, etc).[2]
X
CH, CI" \
P d — ^ - C H3
ci-
(2)An interesting example of C-N bond formation was published by the group of Pfeffer, who synthesized various pyridinium salts via insertion of 1,1-dimethylallene into the a-Pd-C bond of cyclopalladated pyridines and subsequent intramolecular attack of the pyridine nitrogen. It was found that this latter nucleophilic attack of the sp2 nitrogen atom of the pyridine
occurred exclusively at the sterically less hindered carbon atom of the allylic unit. In the early stages of the reaction electronically favored products were produced (i.e. nucleophilic attack at
Chapter 2
the most substituted position). These products isomerized during the reaction towards the thermodynamically favored products (Scheme 1).[8]
PdCI/.
CH2C12 ; A
-Pd°
PdCI/2
Scheme 1
In the Vrieze research group the stoichiometric synthesis of 4H-isoquinolinium salts by reaction of cyclopalladated N,N-dimethylbenzylamine (dmba) with various aliènes has been investigated. It was shown that, in the case of 1,1-dimethylallene, nucleophilic displacement of Pd exclusively takes place at the least substituted position of the Pd-it-allyl intermediate, leading to products with an exocyclic double bond (eq 3).[9]
rPd-
-NMe2 R, * CHC13 or MeCN RT NMe2R
2I Ri
PdCI/2 NMe2 (3) a) Rj = R2 = H b) Ri = H, R2 = Me c) Ri = Me, R2 = MeRecently it was reported that the carbonylation of o-iodophenol in the presence of an aliène, a base and a Pd(0)-catalyst affords 2,3-dihydro-4H-l-benzopyran-4-ones (eq 4). [10]
OH
V
R? ,R3 CO / Pd(OAc)2 / dppb (;-Pr)2NEt (4)These stoichiometric heteroannulation reactions of cyclopalladated complexes with aliènes prompted us to investigate the possibility of employing imines in this type of reactions.
Synthesis of N-Heterocycles by Iminoanmtlation of Cyclopalladated Imines
It is known that alcohols [10-20], carboxylic acids [16, 17, 19, 20], amines [8, 16-18, 21-32], (sulfon)amides [17, 33-37] and oximes [38, 39] are sufficiently nucleophilic to attack Pd-7i-allyl species leading to N-heterocycles. Cyclopalladated imines, bearing an sp^-nitrogen atom, are closely related to cyclopalladated pyridines and a large number of these complexes have been reported in the literature.[3-5] We have selected cyclopalladated a-tetralone ketimines[40] (Fig. 1) as substrates to investigate the nucleophilic character of imines and the regioselectivity of these iminoannulations.
Figure 1
2.2 Results and d i s c u s s i o n
2.2.1 Synthesis of cyclopalladated complexes
Starting from a-tetralone, the corresponding a-tetralone ketimine can be obtained via a condensation reaction with the appropriate amine in toluene at reflux temperature with a catalytic amount of p-TsOH (Scheme 2, step (i)). Nine different amines were used and the corresponding imines la-i are presented in Scheme 2, of which l h was already published in the literature.[41, 42] Most imines were recrystallized from pentane to give air-stable solids. Benzylimine l h could not be recrystallized and was used as an oil without further purification. These imines did not show any tautomerism towards the corresponding enamines as shown by 'H-NMR. It is known for imines that cis-trans isomerism can occur,[43] although generally imines exist only in the more stable trans configuration. Also, the imines of a-tetralone most likely exist as the trans isomers.
Chapter 2
(i)R-NH2 f Y (ii)Pd(OAc)2
p-TsOH cat. NaOAc; toluene, A ^ ^ CH2C12 - H20 l a. i a) R = Ph b)R = p-Tol
(iii)5equivLiCl I V , / X J) R = p-Q-C
6H
4 g)R = p-N02-C6H4 3a-h h) R = CH2Ph i) R = j-Pr Scheme 2The cyclopalladations of a-tetralone ketimines la-h by using Pd(OAc)2 (1 equiv) in the
presence of NaOAc (1 equiv) in dichloromethane at room temperature were straightforward, leading only to the desired ort/io-palladated products bearing an acetato-bridge between the Pd nuclei (Scheme 2, step (ii)).[40] Reaction of l i with Pd(OAc)2 did not lead to the desired
ortho-palladated product. Adding imine l i to a mixture of Pd(OAc)2 and NaOAc in CH2C12 at room
temperature or at 0 °C, gave rise to a rapid color change from yellow to dark green with extensive formation of palladium black. The only organic material that could be detected was imine l i . The presence of NaOAc is not necessary in these reactions but it facilitates the electrophilic aromatic substitution.
The products 2a-h may be purified by flash chromatography (MeOH / CH2CH2 2:98)
and recrystallized from a dichloromethane pentane mixture to give air-stable yellow solids. 'H-NMR of these complexes only showed the presence of one methyl group of the acetato unit indicating that in these complexes the two cyclopalladated a-tetralone ketimine moieties are in
trans position. Reaction of l h with Pd(OAc)2 led to the formation of 2h in which C-H activation
takes place exclusively at the tetralone ring. In principle this substrate offers two positions where C-H activation may occur, namely in the tetralone ring (position 1, Fig. 2) and in the N-benzyl group (position 2, Fig. 2). The 'H-NMR of 2h demonstrated that H„ and Hi r (see
experimental section) are not equivalent as in the free ligand. It is possible that the benzyl group is restricted in its rotation giving rise to two separate signals for Hn and Hn. Another possibility
Synthesis of N-Heterocycles by Iminoannulation of Cyclopalladated Imines
is that the Pd-center is chiral because of two different coordination modes of the acetate ligands. They show a chemical shift difference of 0.6 ppm and a geminal coupling of 14 Hz.
Hu1'
y
^
H i. ,.
pd
x
n
/\ X
Hu y Y 2L (not formed) position 1 2h position 2 Figure 2The cyclopalladation reaction of l h is similar to that of Albert et al., who studied cyclopalladation reactions of N-(benzylidene)-benzylamines with Pd(OAc)2 and PdCl2 (eq
5).[44] Our result may be explained in terms of the formation of a 5-membered aromatic ring involving the two conjugated double bonds of the (-C=C-C=N-) system and the filled palladium
d orbitals of appropriate symmetry. [45]
PdX,
AcOH, A
R = H, CI, N 02
X = Cl,OAc
PdX/2 (5)
Cyclopalladated a-tetralone ketimines are the starting materials for intramolecular iminoannulations with aliènes leading to N-heterocycles. Because of the slightly nucleophilic character of the acetate anion which may compete with the imine nucleophile, the acetate bridges in complexes 2a-h were substituted by chloride. Reaction of complexes 2a-h with LiCl in acetone at room temperature led to the formation of chloro-bridged palladium complexes
3a-h (Sc3a-heme 2, step (iii)). In general t3a-he c3a-hloro-bridged palladium complexes were air-stable
Cliapter 2
driving force for the formation of the chloro-bridged complexes is their insolubility. Characterization by NMR was performed on the bridge-opened monomeric deuterated pyridine complexes, produced after adding deuterated pyridine to the chloro-bridged dimers in CDCI3. In the case of 3h it was possible to determine a 'H-COSY of the chloro-bridged dimer without addition of pyridine to investigate whether the cyclopalladation indeed took place at the tetralone ring. In the aromatic region of the spectrum two sets of coupling protons could be seen, i.e. one set of three coupling protons and one set of five protons, respectively. In the case of cyclopalladation at the N-benzyl group (position 2, Fig. 2), two sets of four aromatic protons would be observed. This 'H-COSY spectrum also showed three sets of methylene protons of the a-tetralone moiety. The deuterated pyridine coordinates in a cis position relative to the palladated carbon, as indicated by the high-field shift of ca. 1 ppm of the aromatic proton in the
ortho position relative to the palladated carbon, owing to the shielding effect of the pyridine
ring. [46]
We were able to grow crystals of complex 3a', formed by chloride abstraction from 3a by AgO,SCF3 and subsequent coordination with 2,2'-bipyridine in acetonitrile (eq 6). The crystals
were suitable for an X-ray crystal structure determination. From the crystal structure (Fig. 3) it can be seen that the Pd-atom forms a square planar complex with the bipyridine ligand and the -V-phenyl-a-tetralone imine ligand. Carbon atom 17 showed disorder in the crystal structure. It appeared that the molecule packed into two distinct conformations of the saturated cyclohexane moiety (ratio 71.3 : 28.7). See Table 1 and Table 2 for information about bond distances and bond angles (experimental section).
1) CF3S03Ag 2) 2,2'-Bipy » MeCN -AgCl CF3SO3- (6) 3a 3a'
Synthesis of N-Heterocycles by Iminoannulation of Cyclopalladated Imines
Figure 3
2.2.2.1 Reaction of the cyclopalladated complexes with 1,1-dimethylallene
Reactions of 1,1-dimethylallene (DMA) with cyclopalladated complexes 3a-h in dichloromethane at room temperature led to iminium salts in which the imine nucleophile predominantly attacked the disubstituted position of the aliène leading to 4a-d,f,h and 5a-d,f,h (Scheme 3 and Table 3). Because of a complex product mixture in the crude reaction mixture of the iminium salts with a chloride anion, the latter was exchanged with KPF6, leading to higher
yields of N-heterocyclic products. Reactions of 3d-f with DMA were generally more sluggish than 3a-c and 3g-h. It was also difficult to obtain pure iminium salts from substrates 3d,f. In the case of 3e no ring closed products could be observed. The reaction of 3g with DMA was very slow, requiring one week for complete conversion into a clear purple solution contaminated with some palladium black. The iminium salt was aromatized to product 4g by dehydrogenation of the tetrahydronaphthalene ring by the formed Pd(0) (Scheme 3). The iminium salts were successfully characterized by 'H and "C-NMR based on a detailed NMR investigation of 4b. All resonances of iminium salt 4b could be assigned by means of different 2D-NMR techniques (Fig. 4).
Cltapter 2
'.2.2.2 Characterization of iminium salts 4a-d, f,h
A 'H-COSY experiment provided information about the chemical shifts of the protons in the saturated ring and in the aromatic rings. It was, however, impossible to assign the two .inylic protons around 5 ppm without the help of 'H-NOESY NMR. H5. in 4b (Fig. 4) shows a
arger NOE effect with H, than H5. Furthermore, NOE effects were detected between H9 and H,7/
H7 and H3 and between the protons of the p-tolyl group and H18. It is interesting to note that
jrotons H9 resonate at a higher field than H7 (see experimental section). One would expect on
the basis of the inductive effect of N, that H„ is deshielded. However, H, is above the plane of the p-tolyl group which effectively shields these two protons, resulting in an overall high field shift of PL. 3a-h 1) A (IQ 2) 10 KPF6 MeOH (i) CH2C12 R PF6 a) R = Ph b) R = p-Tol c) R = p-An d) R = p-Cl-C6H4 f) R = p-I-C6H4 g) R = P-NO2-QH4 h) R = Bn N02 4a-f,h 5a-f,h Scheme 3 2.2.2.3 Regioselectivity
The regioselectivity found in these reactions was generally high. Nucleophilic attack of the imine predominantly took place at the sterically most hindered terminal allyl-carbon atom in reaction with DMA (Table 3).
Synthesis of N-Heterocycles by Iminoannulation of Cyclopalladated Imines
Table 3: Regioselectivity in iminium salts 4 and 5' reaction with ratio yield (4 + 5)
DMA 4 / 5 (%) 3a 1 0 0 / 0 60 3b 9 3 / 7 81 3c 1 0 0 / 0 51 3d 7 5 / 2 5 79 3f 1 0 0 / 0 75 3g 8 2 / 1 8 65 3h 9 1 / 9 86 a: ratios of regioisomers were determined by NMR
NMR studies on (di)phosphine palladium-allyl complexes showed that alkylation with anions of dialkyl malonates predominantly takes place at the most stabilized position (i.e. the terminal allyl carbon atom with donating R groups, see eq 7). It was found ("C-NMR) that donating R-groups on an allyl terminus make this position more electrophilic and thus more prone towards nucleophilic attack, whereas the other terminus is made more nucleophilic.[47]
Ph2P PPh2 BF, X = electron donating Y = electron withdrawing Nu" (7)
From a steric point of view it is clear that nucleophilic attack would be more favored at the least hindered position. Recently, a similar annulation with cyclopalladated pyridines was published by Pfeffer and coworkers.[8] They showed that in the reaction of their complexes with DMA, ring closure took place at the sterically most favored position, in contrast to our results. Apparently electronic effects direct the regiochemistry in our reactions. The latter may be explained as follows: the net positive charge on the rc-allyl-palladium species is more located at the alkyl-substituted terminus. [17] Furthermore, electron-rich Pd(0)-ligand complexes favor coordination to the more electron deficient (less-substituted) double bond of the product. [48-50]
CImpter 2
Recently a carbonylative heteroannulation of o-iodophenol with 1,1-dimethylallene was :eported[10], in which nucleophilic attack of the phenolate anion also took place at the methyl
lisubstituted position of the allyl (eq 4).
It is interesting to note that our products 4, in which nucleophilic attack took place at the most hindered position of the allyl terminus, are kinetic products but can be isolated. They do not isomerize without the presence of a suitable Pd"-source. Refluxing solutions of regioisomers n MeOH with palladium metal did not lead to a substantial change in the ratio of regioisomers. The kinetic product 4b was converted into the thermodynamically more favored product 5b (with the iminium nitrogen adjacent to the methylene group) by refluxing 4b in methanol with 'd(PPh3)4 for 24 hours (eq 8). Due to separation problems, this regioisomer could not be isolated
n a pure state. 'H-NMR showed the presence of a syn and an anti methyl signal which are -trongly indicative for the regioisomer, with the methylene group adjacent to the iminium nitrogen atom.[9] cat. Pd(PPh3)4 MeOH (8) 4b 5b 2.2.2.4 Intermediates
It is difficult to observe Pd-n-allyl intermediates during this reaction with 'H-NMR. The activation barrier for the formation of the heterocycles is similar to the activation barrier for the insertion of DMA into the a-Pd-C bond leading to a Pd-it-allyl complex (Scheme 3, step (i)). Reaction of 3a with DMA in CDC13 showed the presence of broad signals around 3.98 ppm and
3.53 ppm, indicative for the syn- and anti protons of a Pd-7t-allyl complex.[6, 7] The reaction probably proceeds via a rc-allyl intermediate which is attacked intramolecularly by the nucleophilic imine nitrogen atom.
During the reaction of cyclopalladated complexes 3 with DMA, besides the iminium products, enamines could be observed in the crude reaction mixture, with 'H-NMR spectra identical to the ones described for enamines 8 (eq 9). This shows that the iminium salt, which possesses acidic protons (H9) next to the iminium group, is in equilibrium with the
corresponding enamine in CDC13 (eq 9). Probably owing to this equilibrium it was difficult to
Synthesis of N-Heterocycles by Iminoannulation of Cyclopalladated Imines
enamine and iminium salt is shifted towards the iminium salt because of the higher pKa value
for HPF6 as compared with HCl.
-HCl + HC1
(9)
Reaction of 3g with 1,1-dimethylallene led to the dehydrogenated product 4g (Scheme 3). This is due to the fact that during the reaction predominantly the enamine was formed as indicated by 'H-NMR. This can be explained by the electron-withdrawing capacity of the p-N02
group attached to the imine nitrogen. The latter enamine was then easily dehydrogenated by the formed palladium black.
2.2.2.5 Kinetics
Reactions of complexes 3 (Scheme 3) with electron-releasing groups in the para position of the aniline with DMA were much faster than the corresponding reactions with electron-withdrawing groups. Very fast reactions were observed with the benzyl substituted imine 3h. Qualitatively, the reactions proceeded faster in the following order : R = p-N02-Ph < Ph < p-Tol
< p-An < Bn. The overall reaction rate is determined by the slowest of the following steps: the bridge breaking reaction by coordination of the aliène, insertion of the aliène into the a-Pd-C bond, and nucleophilic attack of the imine nitrogen on the formed Pd-n-allyl species. Precoordination of the aliène is faster if the palladium metal center is electron rich by electron donating substituents on the imine and therefore the bridge breaking reaction proceeds faster with these electron-rich imines. The rate of migration of the aryl group onto the aliène is enhanced by electron withdrawing groups on the imine, which makes the palladium more electrophilic. Nucleophilic attack of the imine on the allyl is enhanced by more electron releasing groups on the imine. Only in one case a Pd-ji-allyl species could be detected in the reaction mixture, which indicates that nucleophilic attack of the imine nitrogen cannot be the rate determining step in the formation of iminium salts. Therefore, taking into account that electron releasing groups on the imine accelerate the formation of iminium salts, it can be concluded that the bridge breaking reaction by the aliène probably is the rate determining step. Ryabov studied the mechanism of alkyne insertions in cyclopalladated complexes and showed, that there too the halide bridge breaking step by the alkyne was the rate determining step.[51]
Chapter 2
2.2.3 Reaction of cyclopalladated complexes with vinylidenecyclohexane
Reaction of complexes 3a-c with vinylidenecyclohexane (VCH) led to iminoannulation at the least substituted end of the allyl. Only low yields for iminium salts 6a-c were obtained. Electronically it would be favorable that nucleophilic attack takes place at the most substituted position (leading to 7a-c), but the bulky cyclohexane ring prevents this (Scheme 4 and Table 4).
Table 4: Regioselectivity in iminium salts 6 and 7'
reaction with ratio yield (6 + 7)
VCH (%)
3a 9 7 / 3 < 5
3b 9 9 / 1 < 5
3c 9 6 / 4 < 5 a: Ratios of regioisomers were determined by 'H-NMR
MeCN 3a-c 1)A 2) 10 KPF6 MeCN 6a-c 7a-c a) R = Ph b) R = p-Tol c) R = p-An Scheme 4
Reactions of cyclopalladated complexes 3 with phenylallene did not lead to any N-heterocycle. We do not have a clear explanation for this result.
Synthesis ofN-Heteroq/cles by Iminoannulation of Cydopalkdated Imines
2.2.4 Synthesis of neutral IV-heterocycles from iminium salts
Conversion of the iminium salts into neutral N-heterocycles was carried out either by reaction with a suitable base (eq 10) or by hydrogenolysis of the iminium functional group by NaBH4 (eq 12). A strong base like KOH was needed as reactions with Na2C03 and NaOAc were
not successful to convert iminium salts 4 into enamines 8 (eq 10). Unfortunately these enamines could not be purified by flash chromatography because of the instability of these products under wet acidic conditions. Purification by extraction with pentane was suitable as demonstrated by NMR. However, yellow solutions of these enamines in CDCL, also showed decomposition to dark green products, probably due to the acidic character of CDC13 and the
presence of traces of water. Perhaps this is an acid catalyzed hydrolysis or air oxidation. The decomposition products could not be characterized.
4a-d,h
KOH MeOH
(10)
8a-d,h
Enamine 8g was prepared by reaction of 3g with DMA in the presence of pyridine as base (eq 11).
>
-pyridine pyridine »HCl CDC13 ( H ) 3g 8gThe reaction between 3g and DMA was followed by NMR in the absence of base. It turned out that the equilibrium (as shown in eq 9) lay towards the enamine as the p-N02QH4
Chapter 2
Hydrogénation of the iminium group in 4a-d,h with NaBHi in MeOH at room temperature easily gave amines 9a-d,h (eq 12). Purification of the amines proceeded without any decomposition. Reduction only took place at the iminium functional group. The exocyclic vinylic group was not attacked by NaBH4. Only in the case of 4h was the exocyclic double bond
hydrogenated, probably due to Pd impurities which may catalyze the hydrogénation of the exocyclic double bond. Besides the hydrogénation of the exocyclic double bond, a dehydrogenation of the tetrahydronaphthalene ring had occurred, similar to the formation of 4g from 3g by Pd(0) impurities (see Scheme 3).
4a-d,h
NaBH4
MeOH
(12)
2.3 C o n c l u s i o n s
Reactions of complexes 3a-h with aliènes show successful regioselective iminoannulations towards new N-heterocycles. In the case of 1,1-dimethylallene, nucleophilic attack predominantly takes place at the disubstituted end of the Pd-n-allyl terminus, leading to products 4a-h. Apparently the regiochemistry in these reactions is controlled by electronic effects, as two donating methyl groups on the most substituted terminus of the Pd-re-allyl moiety can stabilize a positive charge. Reactions of cyclopalladated complexes 2 with vinylidenecyclohexane lead to products in which nucleophilic attack takes place at the least substituted position. Electronically products 7a-c would be favored, but the bulky cyclohexyl group prevents this, resulting in the low yield of 6a-c.
Qualitatively, the reactions of complexes 2 with DMA proceed faster in the order p-N02
-Ph < -Ph < p-Tol < p-An < Bn. The rate determining step in these reactions is very likely the breaking of the chloride bridge upon coordination of the aliène to form a mononuclear species which then undergoes migration of the aryl unit to the central carbon atom of the aliène and subsequent intramolecular nucleophilic attack of the imine nitrogen to one of the terminal carbon atoms of the allyl unit.
Unfortunately these iminium salts cannot be fully purified by e.g. column chromatography. Iminium salts 4a-d,h were therefore successfully converted into neutral
N-Synthesis of N-Heterocycles by Iminoannulation of Cyclopalladated Imines
heterocycles either by reaction with KOH, leading to enamines 8a-d,h, or by reduction of the iminium group by NaBH4 to amines 9a-d,h. Pd impurities may be responsible for the
concomitant hydrogénation of an exocyclic double bond (9h) or the dehydrogenation of oc-tetralone ketiminium salts towards aromatized products (4g and 9h).
Cliapter 2
2.4 Experimental Section
General remarks
All manipulations were carried out in an atmosphere of purified, dry nitrogen by using standard Schlenk techniques. Solvents were dried according to literature procedures[52] and stored under nitrogen. a-Tetralone, p-toluidine, p-chloroaniline, p-bromoaniline, p-iodoaniline, p-nitroaniline, acetic acid, isopropylamine, KPF6 and KOH, were purchased from Acros
Chimica. Lithium chloride and sodium borohydride were obtained from Aldrich, p-anisidine and 1,1-dimethylallene from Fluka and p-toluene sulphonic acid monohydrate and sodium acetate from Merck. All of these compounds were used without further purification. Aniline and benzylamine were dried over calcium hydride and distilled. Propargyl alcohol was distilled from K2C03. Xylene (mixture of isomers) and toluene were distilled from sodium. Elemental
analyses were performed on a Vario EL in the Inorganic Laboratory of this university or by Kolbe Microanalytisches Laboratorium, Mülheim an der Ruhr, Germany. 'H and 13C{'H}-APT
NMR spectra were recorded on a Bruker AMX 300 spectrometer. 'H-COSY, HETCOR fH-13C)
and 'H-NOESY NMR were recorded on a Bruker DRX 300 spectrometer using standard COSY-45 (Gradient Selected), HETCOR [J(C-H) = 140 Hz] and 2D NOESY pulse sequences. C-H-HETCOR spectrum was used to calculate 'JC-H- AU NMR spectra were recorded in CDCL, (unless indicated otherwise). 'H-NMR spectra of complexes 3 were recorded in CDC13 with a drop of
pyridine-d5, except 3h which was measured in CDC13. Mass spectra and accurate mass
determinations were performed on a JEOL JMS SX/SX102A four-sector mass spectrometer, coupled to a JEOL MS-MP7000 data system.
Preparation of l a as general procedure A for the synthesis of a-tetralone ketimines (la-i)
3,4-Dihydro-2H-naphthalen-l-ylidene)-phenyl-amine (la) See step 1 in Scheme 2. In a 250 mL 3-necked flask containing 4Â molecular sieves, a-tetralone (10.96 g, 75 mmol) was dissolved in toluene (50 mL) together with freshly distilled aniline (6.98 g, 75 mmol) and p-toluenesulphonic acid monohydrate (0.30 g, 1.58 mmol). The mixture was heated to reflux for 1 day. After the mixture was cooled to room temperature the brown solution was filtered to remove the molsieves. The molsieves were washed with diethyl ether (2 x 25 mL). The volatile solvents in the combined extracts were removed in vacuo to give a brown oil. After recrystallization from pentane, l a (10.58 g, 48 mmol, 64%) was obtained as a yellow solid. 'H-NMR (300 MHz): 8.35 (d, 1 H, J = 7.7 Hz), 7.37 (m, 4 H),
Synthesis of N-Heterocycles by Iminoannulation of Cyclopalladated Intines
7.23 (d, 1 H, J = 7.2 Hz), 7.10 (t, 1 H, J = 7.4 Hz), 6.83 (d, 2 H, J = 7.2 Hz), 2.93 (t, 2 H, J = 6.1 Hz), 2.55 (t, 2 H, J = 6.4 Hz), 1.94 (p, 2 H, J = 6.3 Hz). "C-NMR (75 MHz): 165.4, 151. 5, 141.1, 133.7, 130.5, 128.8, 128.5, 126.3, 126.2, 122.8, 119.3, 117.4, 29.8, 29.7, 22.8. Anal. Found (calc. for C16H15N): C = 86.81 (86.84); H = 6.83 (6.84); N = 6.23 (6.33).
(3,4-Dihydro-2H-naphthalen-l-ylidene)-p-tolyl-amine (lb) According to
general procedure A, starting from a-tetralone (10.96 g, 75 mmol), p-toluidine (8.11 g, 75 mmol) and p-TsOH (catalytic amount) in xylenes (200 mL) with molecular sieves using a reaction time of 2 days at reflux, l b (10.72 g, 45.6 mmol, 61%) was obtained as a yellow solid. 'H-NMR (300 MHz): 8.33 (d, 1 H, J = 7.8 Hz), 7.28 (m, 3 H), 7.16 (d, 2 H, J = 8.1 Hz), 6.72 (d, 2 H, J = 8.1 Hz), 2.91 (t, 2 H, J = 6.1 Hz), 2.55 (t, 2 H, J = 6.4 Hz), 2.36 (s, 3 H, PhCH3), 1.92 (p, 2 H, J = 6.3 Hz).
"C-NMR (75 MHz): 165.4, 148.8, 141.0, 133.8, 132.2, 130.3, 129.3, 128.5, 126.2, 126.1, 119.3, 29.8, 29.6, 22.8, 20.7 (PhCH3). Anal. Found (calc. for C17H17N): C = 86.31 (86.76); H =
7.19 (7.29); N = 5.85 (5.95).
(3,4-Dihydro-2H-naphthalen-l-ylidene)-(4-methoxy-phenyl)-amine (lc)
According to general procedure A, starting from a-tetralone (10.96 g, 75 mmol), p-anisidine (9.24 g, 75 mmol) andp-TsOH (catalytic amount) in xylenes (200 mL) with molecular sieves using a reaction time of 2 days at reflux, l c (13.39 g, 53.3 mmol, 71%) was obtained as a yellow solid. !H-NMR (300 MHz): 8.36 (d, 1 H, J = 7.4 Hz), 7.30 (m, 2 H), 7.20 (d, 1 H, J = 7.4 Hz), 6.91 (d, 2 H, J = 8.8 Hz), 6.79 (d, 2 H, J = 8.8 Hz), 3.82 (s, 3 H, OCH3), 2.91 (t, 2 H, J = 6.1 Hz), 2.58 (t, 2 H, J = 6.4
Hz), 1.92 (p, 2 H, J = 6.2 Hz). ^C-NMR (75 MHz): 165.8,155.7,144.4,141.0,133.8, 130.4, 128.5, 126.2, 126.1, 120.7, 114.0, 55.3 (OCH3), 29.8, 29.7, 22.9. Anal. Found (calc. for C17H17NO): C = 79.78 (81.24); H = 6.61 (6.82); N = 5.37 (5.57).
(4-Chloro-phenyl)-(3,4-dihydro-2H-naphthalen-l-ylidene)-amine (Id)
According to general procedure A, starting from a-tetralone (10.96 g, 75 mmol), p-chloroaniline (8.59 g, 75 mmol) and p-TsOH (catalytic amount) in xylenes (200 mL) with molecular sieves using a reaction time of 2 days at reflux, Id (16.6 g, 54.8 mmol, 73%) was obtained as a yellow solid. XH-NMR (300 Mhz): 8.28 (d, 1
H, J = 7.4 Hz), 7.38 (t, 1 H, J = 7.4 Hz), 7.28 (m, 3 H), 7.20 (d, 1 H, J = 7.4 Hz), 6.73 (d, 2 H, J = 8.6 Hz), 2.90 (t, 2 H, J = 6.1 Hz), 2.50 (t, 2 H, J = 6.8 Hz), 1.92 (p, 2 H, J = 6.5 Hz). 13C-NMR (75 MHz): 166.1,150.0, 141.2,133.4,130.7,128.8,128.6,128.1,
Chapter 2
26.3, 126.2, 120.7, 29.7, 22.7. Anal, found (calc. for C16H14NCI): C = 74.69 (75.14); H = 5.45 5.52); N = 5.42 (5.48).
(4-Bromo-phenyl)-(3,4-dihydro-2H-naphthalen-l-ylidene)-amine ( l e )
According to general procedure A, starting from oc-tetralone (10.96 g, 75 mmol), p-bromoaniline (12.15 g, 75 mmol) and p-TsOH (catalytic amount) in xylenes (200 mL) with molecular sieves using a reaction time of 2 days at reflux, l e (14.6 g, 48.8 mmol, 65%) was obtained as a yellow solid. aH-NMR (300 MHz): 8.28 (d,
1 H, J = 7.4 Hz), 7.45 (d, 2 H, J = 8.6 Hz), 7.39 (t, 1 H, J = 7.4 Hz), 7.30 (t, 1 H, J = 7.4 Hz), 7.20 (d, 1 H, J = 7.4 Hz), 6.69 (d, 2 H, J = 8.6 Hz), 2.91 (t, 2 H, J = 6.1 Hz), 2.50 (t, 2 H, J = 6.8 Hz), 1.93 (p, 2 H, J = 6.4 Hz). 13C-NMR (75 MHz): 166.1, 150.5,
141.2, 133.4, 131.7, 130.7, 28.6, 126.3, 126.2, 121.2, 115.7, 29.72, 29.68, 22.7. Anal. Found (calc. for Ci6Hi4NBr): C = 63.81 (64.01); H = 4.67 (4.70); N = 4.64 (4.67).
(3,4-Dihydro-2H-naphthalen-l-ylidene)-(4-iodo-phenyl)-amine (If) According
to general procedure A, starting from a-tetralone (10.96 g, 75 mmol), p-iodoaniline (16.4 g, 75 mmol) and p-TsOH (catalytic amount) in xylenes (200 mL) with molecular sieves using a reaction time of 2 days at reflux, If (13.5 g, 39.0 mmol, 52%) was obtained as a yellow solid. JH-NMR (300 MHz): 8.27 (d, 1 H, J
= 7.7 Hz), 7.64 (d, 2 H, J = 8.4 Hz), 7.39 (t, 1 H, J = 7.7 Hz), 7.30 (t, 1 H, J = 7.7 Hz), 7.21 (d, 1 H, J = 7.7 Hz), 6.58 (d, 2 H, J = 8.4 Hz), 2.91 (t, 2 H, J = 6.1 Hz), 2.50 (t, 2 H, J = 6.5 Hz), 1.93 (p, 2 H, J = 6.2 Hz). 13C-NMR (75 MHz): 166.0, 151.2, 141.3,
'37.7, 133.4, 130.7, 128.6, 126.3, 126.2,121.6, 29.73, 29.69, 22.7. Anal. Found (calc. for Ci6Hi4NI):
= 51.56 (53.35); H = 4.06 (4.07); N = 3.77 (4.03).
(3,4-Dihydro-2H-naphthalen-l-ylidene)-(4-nitro-phenyl)-amine (lg) According
to general procedure A, starting from a-tetralone (10.96 g, 75 mmol), p-nitroaniline (10.36 g, 75 mmol) and p-TsOH (catalytic amount) in xylenes (200 mL) with molecular sieves using a reaction time of 2 days at reflux, l g (6.38 g, 24
mmol, 35%) was obtained as a yellow solid. !H-NMR (300 MHz): 8.26 (d, 2 H, J
= 9.8 Hz), 8.23 (d, 1 H, J = 8.9 Hz), 7.42 (t, 1 H, J = 7.6 Hz), 7.32 (t, 1 H, J = 7.6 Hz), 7.24 (d, 1 H, J = 7.6 Hz), 6.89 (d, 2 H, J = 8.9 Hz), 2.93 (t, 2 H, J = 6.0 Hz), 2.49 (t, 2 H, J = 6.4 Hz), 1.97 (p, 2 H, J = 6.2 Hz). 13C-NMR (75 MHz): 166.1, 157.9, 143.3,
141.6,132.7,131.3,128.7, 126.4,124.9,119.5, 30.2, 29.6, 22.6. Anal. Found (calc. for C16H14N2O2): C = 72.25 (72.16); H = 5.24 (5.30); N = 10.37 (10.52).
