Thesis Reference - Archive ouverte UNIGE

Transcription

Thesis
Asymmetric Synthesis of Configurationally Stable Tröger Bases
SHARMA, Ankit
Abstract
La base de Tröger 1 a été découverte en 1887. Depuis, ces composés de type 1 ont fait
l’objet d’une grande attention en raison de leur préparation aisée et de leur structure
tricyclique rigide dans laquelle les groupes aromatiques sont parfaitement perpendiculaires
les uns aux autres. Cette élégante géométrie permet toute une variété d’applications :
reconnaissance moléculaire, sondes interagissant avec l’ADN, systèmes biomimétiques et
structures auto-assemblées.
Reference
SHARMA, Ankit. Asymmetric Synthesis of Configurationally Stable Tröger Bases. Thèse
de doctorat : Univ. Genève, 2011, no. Sc. 4409
URN : urn:nbn:ch:unige-206683
Available at:
http://archive-ouverte.unige.ch/unige:20668
Disclaimer: layout of this document may differ from the published version.
[ Downloaded 19/10/2016 at 18:55:00 ]
UNIVERSITÉ DE GENÈVE
FACULTÉ DE SCIENCES
Section de chimie et biochimie
Département de chimie organique
Professeur Jérôme Lacour
Asymmetric Synthesis of Configurationally
Stable Tröger Bases
THÈSE
présentée à la Faculté des sciences de l'Université de Genève
pour obtenir le grade de Docteur ès sciences, mention chimie
par
Ankit Sharma
de
Mathura (India)
Thèse N° 4409
GENÈVE
Atelier d'impression ReproMail
2011
Publications
Publications:
1. "Stereoselective synthesis of configurationally stable functionalized ethano-bridged
Tröger bases" Michon, C.; Sharma, A.; Bernardinelli, G.; Francotte, E.; Lacour, J. Chem.
Commun. 2010, 46, 2206.
2. "One-step catalytic asymmetric synthesis of configurationally stable Tröger bases"
Sharma, A.; Guénée, L.; Naubron, J.-V.; Lacour, J. Angew. Chem. Int. Ed. 2011, 50,
3677.
3. “Asymmetric
synthesis
of
ethano-Tröger
bases
using
CuTC-catalyzed
diazo
decomposition reactions” Sharma, A.; Besnard, C; Guénée, L.; Lacour, J. Org. Biomol.
Chem. 2012, 10, 966-969.
Remerciements
Remerciements
The results reported in this manuscript were obtained as part of a thesis done in the laboratory
of Prof. Jérôme Lacour, in the department of organic chemistry at the University of Geneva.
First I would like to express my gratitude to Prof. Jérôme Lacour for giving me the
opportunity to work in his laboratory and also for his continuous support and patience thought
out the four years.
I would also like to thank Prof. Christopher J. Moody (The University of Nottingham) and
Prof. E. Peter Kündig (University of Geneva) for accepting and evaluating my work written in
this thesis.
I also express my gratitude to the teams providing analytical facilities: NMR (André Pinto,
Rupali and Dr. Damien Vitorge Jeannerat) and specially SMS (Prof. Gerard Hopfgartner,
Sophie Michalet, Natalie Oudry, Eliane Sandmeier and Philippe Perrottet) for their essential
contributions. I would specially like to thank Stéphane Grass, Sonya Torche and Mireille
Heimendinger, for all their help thoughout the four years.
I would like to warmly thank my past and present colleagues in the laboratory and from the
department for bearing me all these years and for providing pleasant and enjoyable
environment in the laboratory. I would specially like to thank Martina, Roman, Walid, Rafa,
Diane, Joyram, Mahesh, Sandip, Cecilia, petit Jerome, Nareddy, Santanu, Rupali, Ali, André,
Luca, Marco, Andrea, Javier, Rajesh, Federico, Ravi Kumar, Fedor, Audrey, JB, Chandan,
Dips, Bruno, Fedor (in no particular order) for their support and time inside and outside the
Department.
I would like to thank all the Indian friends for being my stress burster for making life much
more enjoyable in Geneva especially with excellent Indian food  Joyramda, Mahesh,
Pradeep (anna), Sachin, Krishan, Dhaval, Akash, Yogesh, Bankim, Sahana, Santanu, Rupali, Viral.
I would like to express my special gratitude for the cricket team in Geneva for all the fun time
on weekends.
I remain thankful to David Linder, Fedor (master student), Thierry, Diane, Johann and Steven
for their useful suggestions and for correcting different parts of this thesis.
Finally, I want to thank my family and and specially my wife Parul for her understanding and
support in past four years and all who were always by my side (and there are many).
Abbreviations, Symbols and Units
Abbreviations
: wave length
TB: Tröger base
J: coupling constant
b(s): broad (singlet)
tR: retention time
s: singlet
T: temperature
d: doublet
Units
dd: doublet of doublet
°C: degree Celsius
t: triplet
K: Kelvin
dt: doublet of triplet
g: gram
dq: doublet of quartet
mg: miligram
m: multiplet
l: microliter
TLC: Thin layer chromatography
mL: milliliter
Cat.: catalyst
mmol: millimole
equiv.: equivalent
mol: micromole
conv.: conversion
M: molarity
M.P.: melting point
s: second
rac: racemic
min: minute
ee: enantiomeric excess
h: hour
cee: conservation of enantiomeric excess that is
Hz: Hertz
[product ee/reactant ee] X 100.
ppm: part per million
dr: diastereomeric ratio
rt: room temperature (25 °C)
Ar: aryl
Ph: phenyl
napht: naphthalene
THF: tetrahydrofuran
nPr: n-propyl
iPr: iso-propyl
tBu: tert-butyl
TMS: trimethylsilyl
Ts: Tosyl
Tf: Triflate
EDA: ethyl diazo acetate
BINOL: 1,1’-Bi-2-naphthol
CuTC: Copper(I) Tiophene carboxylate
DMAP: 4-Dimethylaminopyridine
tfacac: trifluoroacetylacetonate
hfacac: hexafluoroacetylacetonate
Pfm: perfluorbutyramide
Symbols
: chemical shift
English Version
Chimie des Ylures d’Azote : Synthèse Asymétrique de bases de Tröger
La base de Tröger 11 a été découverte en 1887.2 Depuis, ces composés de type 1 ont fait
l’objet d’une grande attention en raison de leur préparation aisée et de leur structure
tricyclique rigide dans laquelle les groupes aromatiques sont parfaitement perpendiculaires les
uns aux autres. Cette élégante géométrie permet toute une variété d’applications :
reconnaissance moléculaire,3 sondes interagissant avec l’ADN,4 systèmes biomimétiques5 et
structures auto-assemblées.6
En stéréochimie, les bases de Tröger ont une place unique car ce sont les premiers
composés chiraux comportant des atomes d’azote stéréogènes à avoir été dédoublés.7
Cependant, dans des conditions acides, ces composés racémisent relativement lentement et de
manière non réversible par ouverture du pont méthylène (∆G‡ = 23.9 kcal/mol pour
l’énantiomérisation).8 Les sels d’ammonium quaternaires dérivés racémisent encore plus vite
(∆G‡ = 21.5 kcal/mol). Par conséquent, cette labilité a limité l’emploi des bases de Tröger en
catalyse asymétrique. Pour résoudre ce problème, le pont méthylène des bases méthanoTröger 1 peut être modifié en un pont éthylène (composés 16). Cependant, le manque de voie
de synthèse générale a limité leur utilisation.9
Summary
En conséquence et en raison du potentiel de ces molécules pour diverses applications,
nous avons établi la première voie de synthèse générale des composés 16 qui consiste en (i)
une dérivatisation d’un ammonium quaternaire, (ii) une formation d’un ylure contrôlée par la
présence d’un groupement fortement électro-attracteur sur la chaine latérale introduite et (iii)
un piégeage intramoléculaire de l’intermédiaire iminium avec l’énolate généré in-situ.10 Cette
procédure est simple de mise en œuvre et permet l’obtention de dérivés fonctionnalisés sous
forme d’un seul stéréoisomère avec un contrôle total du nouveau centre stéréogène généré (dr
> 49:1). Les énantiomères de ce composé sont configurationnellement stables en conditions
acides contrairement aux précurseurs méthano. Aucune racémisation n’a été observée, même
à 100 °C pendant 2 jours. De plus, ces adduits sont facilement transformés en 1,2aminoalcools par réduction ou par addition nucléophile ; cette classe de composés étant
importante pour le design de ligands. De très bonnes sélectivités (dr > 49:1) sont là encore
observées, ce qui peut être facilement expliqué par une approche du nucléophile de type
Felkin-Anh. Cette méthodologie nous a permis de synthétiser des dérivés éthano
configurationnellement stables et d’introduire de façon concomitante de nouveaux centres
stéréogènes avec un stéréocontrôle total. Cependant, de faibles rendements sont généralement
obtenus (35-60%), et peu de bases méthano-Tröger peuvent être alkylées de manière propre
pour obtenir les sels d’ammonium quaternaires nécessaires. De plus, ces sels d’ammonium
quaternaires sont toujours obtenus sous forme de racémates. Par conséquent, les bases éthanoTröger sont isolées sous forme de mélange racémique, même lorsque un substrat de type 1
énantiopur est utilisé. Pour éviter toute racémisation, la voie de synthèse devrait donc être
modifiée : un intermédiaire ylure d’ammonium de type 4 devrait être généré directement.
Une nouvelle procédure a donc été développée dans laquelle les bases éthano-Tröger
sont construites en une seule étape par réaction de composés diazo avec des bases de Tröger
en présence d’un catalyseur au Rh(II).11 Cette fois-ci, un carbone stéréogène quaternaire est
introduit (dr jusqu’à 49:1) et les bases de Tröger 1 sont transformées en dérivés 16 avec de
très bonnes énantiospécificités (ee ≥ 97% pour plus de 10 adduits). De tels transferts de
chiralité aussi efficaces sont rares dans la chimie des bases de Tröger. Le procédé est efficace
English Version
et la base de Tröger dérivée de la p-méthoxyaniline peut également être employée (cee 98%).
La configuration relative du diastéréoisomère majoritaire a été établie par analyse
cristallographique (R1 = Ph, R2 = CO2Me). La configuration absolue a été déterminée avec
certitude par dichroïsme circulaire vibrationnel et montre que l’expansion du cycle a lieu
rétention de configuration.
Une base de Tröger 5 substituée de manière non symétrique, et comportant des
substituants électrodonneurs et électroattracteurs sur les deux cycles aromatiques (OMe et
NO2), a également été préparée. Un seul régioisomère est obtenu (71% de rendement, dr >
49:1) après réaction avec l’ester α-diazo β-diéthyle : l’azote électrodéficient (en para du
groupe nitro) est l’atome réagissant. Ce résultat, de premier abord surprenant, peut être
rationalisé en considérant une formation
préférentielle de l’intermédiaire. Dans cet
intermédiaire, l’atome d’azote cationique est stabilisé par le groupe p-OMe tandis que la
densité électronique de l’autre atome d’azote est délocalisée de manière efficace vers le
substituant nitro.
Finalement, des catalyseurs de Cu(I) ont également été testés. Ce sont les plus utilisés
pour décomposer les dérivés diazo et former des ylures d’azote. Différentes sources de cuivre
ont été testées et le CuTC (Copper(I) Thiophène-2-carboxylate), bien que rarement employé
dans cette chimie, a été en général le catalyseur le plus efficace. Une dichotomie a par ailleurs
été observée dans la réactivité des catalyseurs de Cu(I) et de Rh(II). Par exemple, la réaction
de α-diazo β-dicétones (R1 = Me, R2 = COMe) est improductive avec du Rh2(OAc)4 alors que
le CuTC s’est révélé très actif (70% de rendement, ee 64%). De plus, la réaction de différents
α-diazo-β-cétoesters en présence de CuTC a donné de meilleures sélectivités et réactivités.
Dans le cas des α-diazo-β-phénylcétones (R1 = Ar, R2 = COR), le CuTC s’est révélé être plus
efficace que le Rh2(OAc)4 (rendement 80-87%, dr 9:1-12:1, ee 83-93% vs rendement 70%, dr
5:1, ee 10% respectivement).12
Summary
Après avoir développé une méthodologie pour la synthèse de bases éthano-Tröger avec
de très bonnes puretés énantiomériques, l’étape suivante a été le développement
d’applications pour ces dérivés. Nous nous sommes plus particulièrement intéressés aux
domaines de la catalyse asymétrique et des phases stationnaires chirales (PSC) pour la
chromatographie. Les composés de type 16 peuvent être utilisés pour l’aziridination de
dérivés de chalcone. Des rendements et sélectivités modérés ont été obtenus. Cependant, la
stabilité configurationnelle ainsi que la recyclabilité de ces bases éthano-Tröger a été
clairement établie. De plus, le développement d’une nouvelle phase stationnaire chirale a été
étudiée dans laquelle les bases éthano-Tröger constituent la phase stationnaire (TB-PSC).
Bien que les résultats préliminaires soient décourageants puisque la TB-PSC sépare
difficilement les analytes chiraux, les bases de Tröger elles-mêmes ont été apparemment bien
reconnues. Ceci suggère que nous sommes en présence d’une molécule chirale très rigide qui
ne peut que très difficilement changer de conformation pour s’adapter au substrat. Ces
premiers résultats nous permettront de développer des TB-PSC davantage polyvalentes :
l’introduction de divers groupements fonctionnels devrait aider le phénomène de
reconnaissance chirale.
En conclusion, nous avons développé une nouvelle procédure qui permet (i) une
transformation directe de bases méthano-Tröger en éthano-Tröger avec de bons rendements,
(ii) une introduction hautement diastéréosélective de centres stéréogènes tertiaires ou
quaternaires, (iii) la synthèse d’une variété d’adduits énantioenrichis qui peuvent être
directement utilisés ou facilement transformés en d’autres bases de Tröger fonctionnalisées,
(iv) la distinction chimique possible entre les deux atomes d’azote stéréogènes. Enfin,
English Version
quelques applications de ces molécules ont été étudiées et qui souligne leur potentiel pour
remplacer les bases méthano-Tröger.
Références :
1. J. Tröger, J. Prakt. Chem. 1887, 36, 225-245.
2. a) S. Sergeyev, Helv. Chim. Acta 2009, 92, 415-444; b) B. Dolensky, J. Elguero, V. Kral, C. Pardo and M.
Valik, Adv. Heterocycl. Chem. 2007, 93, 1-56; c) M. Demeunynck and A. Tatibouet, Prog. Heterocycl.
Chem. 1999, 11, 1-20.
3. a) S. Satishkumar and M. Periasamy, Tetrahedron: Asymmetry 2009, 20, 2257-2262; b) M. Valik, J. Cejka,
M. Havlik, V. Kral and B. Dolensky, Chem. Commun. 2007, 3835-3837; c) E.-i. Kim, S. Paliwal and C. S.
Wilcox, J. Am. Chem. Soc. 1998, 120, 11192-11193; d) M. J. Crossley, L. G. Mackay and A. C. Try, J.
Chem. Soc., Chem. Commun. 1995, 1925-1927; e) E. Weber, U. Müller, D. Worsch, F. Vögtle, G. Will and
A. Kirfel, J. Chem. Soc., Chem. Commun. 1985, 1578-1580.
4. a) E. B. Veale, D. O. Frimannsson, M. Lawler and T. Gunnlaugsson, Org. Lett. 2009, 11, 4040-4043; b) A.
Tatibouët, M. Demeunynck, C. Andraud, A. Collet and J. Lhomme, Chem. Commun. 1999, 161-162.
5. C. S. Wilcox and M. D. Cowart, Tetrahedron Lett. 1986, 27, 5563-5566.
6. a) T. Weilandt, U. Kiehne, G. Schnakenburg and A. Lutzen, Chem. Commun. 2009, 2320; b) Y. M. Jeon, G.
S. Armatas, D. Kim, M. G. Kanatzidis and C. A. Mirkin, Small 2009, 5, 46-50; c) U. Kiehne, T. Bruhn, G.
Schnakenburg, R. Frohlich, G. Bringmann and A. Lützen, Chem. Eur. J. 2008, 14, 4246-4255; d) M. S.
Khoshbin, M. V. Ovchinnikov, C. A. Mirkin, J. A. Golen and A. L. Rheingold, Inorg. Chem. 2006, 45, 26032609.
7. V. Prelog and P. Wieland, Helv. Chim. Acta 1944, 27, 1127-1134.
8. a) D. A. Lenev, K. A. Lyssenko, D. G. Golovanov, V. Buss and R. G. Kostyanovsky, Chem. Eur. J. 2006, 12,
6412-6418; b) O. Trapp, G. Trapp, J. W. Kong, U. Hahn, F. Vögtle and V. Schurig, Chem. Eur. J. 2002, 8,
3629.
9. a) Y. Hamada and S. Mukai, Tetrahedron: Asymmetry 1996, 7, 2671-2674; b) D. A. Lenev, D. G.
Golovanov, K. A. Lyssenko and R. G. Kostyanovsky, Tetrahedron: Asymmetry 2006, 17, 2191-2194.
10. C. Michon, A. Sharma, G. Bernardinelli and E. Francotte, J. Lacour Chem. Commun. 2010, 46, 2206-2208.
11. A. Sharma, L. Guenee and J. V. Naubron, J. Lacour Angew. Chem. Int. Ed. 2011, 50, 3677-3680.
12. A. Sharma, C. Besnard, L Guénée; J. Lacour, J. Org. Biomol. Chem. 2012, 10, 966-969.
English Version
Novel Nitrogen Ylide Chemistry: Asymmetric Synthesis of Tröger bases
Tröger base (TB, 1)1 was first discovered in 1887.2 Since then, compounds of type 1 have
attracted considerable attention because of their facile preparation and rigid tricyclic core in
which the two aromatic moieties are perfectly perpendicular to each other. This elegant
geometry imparts a variety of applications such as molecular recognition,3 DNA-interacting
probes,4 biomimetic systems,5 and self-assembled structures.6
In stereochemistry, Tröger bases are unique being the first chiral compounds with stereogenic
nitrogen atoms to be resolved.7 However, these compounds undergo relatively slow yet
definite racemization reactions under acidic conditions through the ring opening of the
methylene bridge (∆G‡ = 23.9 kcal/mol for the enantiomerization).8 Furthermore, derived
quaternary ammonium salts racemize even faster (∆G‡ = 21.5 kcal/mol). This lability has thus
limited use of TB in asymmetric synthesis and catalysis. Transforming methano-Tröger bases
1 into ethano-bridged derivatives 16 is a solution to this problem though it has been rarely
used due to the lack of a general synthetic route to these ring-expanded compounds.9
In view of this knowledge and the potential of these molecules for different applications, we
search and established a first route to compounds 16 involving (i) a quaternary ammonium
derivatization, (ii) selective ylide formation controlled by the presence of a strong electronwithdrawing group on the introduced side chain and (iii) an effective rearrangement.10
Interestingly, this simple-to-run protocol affords functionalized derivatives as single
stereoisomers with complete control at the newly generated C-stereo center (dr > 49:1).
Importantly, the enantiomers of these novel compounds are indeed quite more
configurationally stable under acidic conditions than their methano-precursors. No
racemization was observed even at 100 oC for 2 days.
Summary
Further, these adducts are easily transformed into 1,2-aminoalcohols by reduction or
nucleophilic addition reactions, this class of compounds being important for ligand design.
Very high selectivity (dr > 49:1) was again observed which can be readily explained by a
“Felkin-Anh” transition state for the nucleophilic attack. Although this methodology allowed
the making of configurationally stable derivatives with concomitant introduction of two novel
stereogenic centers with complete stereocontrol, low yields of ethano-Tröger bases were
generally afforded (35-60% overall), and few methano-Tröger bases could be alkylated
cleanly to afford the necessary quaternary ammonium salts. More unfortunate was the fact
that racemic ammonium salts were always obtained from enantiopure methano-Tröger base
precursors and hence racemic ethano-Tröger bases were obtained. So, to avoid this
unfortunate racemization, it was decided to optimize the synthetic strategy and try to generate
directly the ylide intermediate necessary for the rearrangement to occur.
A new procedure was thus developed in which, in a single step, ethano-Tröger bases are
constructed by the reaction of diazo-compounds with methano-Tröger bases under Rh(II)
catalysis.11 This time, a quaternary carbon stereogenic center is introduced (dr up to 49:1) and
Tröger bases 1 can be transformed into derivatives 73 with high level of enantiospecificity (ee
≥ 97% for more than 10 adducts); Such an efficient transfer of chirality is a rarity in both
Tröger and [1,2]-Stevens chemistry. The process is efficient and the Tröger base derived from
English Version
p-methoxyaniline can also be used effectively (cee 98%). The relative configuration of the
major diastereomer was ascertained by X-ray crystallographic analysis (R1 = Ph, R2 =
CO2Me). The absolute configuration was established with certainty first by vibrational
circular dichroism and then later through a Flack parameter analysis of the X-ray structure of
compound (−)-73oh (R1 = p-ClPh, R2 = CO2Me). It clearly demonstrates that the ringexpansion reaction occurs with retention of configuration.
An unsymmetrically-substituted Tröger base 79 was also prepared carrying electron-donating
OMe and electron-withdrawing NO2 substituents on the two aromatic rings. Interestingly, a
single regioisomer was obtained (71% yield, dr > 49:1) from its reaction with α-diazo βdiethyl ester, which arises from the reaction of the electron-poor nitrogen atom only (para to
the nitro group). This result, at first glance surprising, can be rationalized considering a
preferred formation of intermediate 84. In this intermediate, the cationic nitrogen atom is
stabilized by the p-OMe group while the electron-density on the other nitrogen atom is
efficiently delocalized towards the nitro substituent.
Finally, copper salts were tested as possible catalysts. They are most commonly used in diazodecomposition chemistry to generate nitrogen ylides. After trying different copper sources, it
Summary
was shown that CuTC (Copper(I) Thiophene-2-carboxylate), although rarely used in this kind
of chemistry, is generally the most effective catalyst. For this reaction, we also observed a
dichotomy in the reactivity of Cu(I) and Rh(II) catalysts. For instance, the reaction of α-diazo
β-diketones (R1 = Me, R2 = COMe) was unproductive with Rh2(OAc)4 but high yielding with
CuTC (70% yield, ee 64%). Also, treatment with different α-diazo-β-ketoesters with CuTC
gave better selectivity and reactivity. In the case of α-diazo-β-phenylketones (R1 = Ar, R2 =
COR), CuTC is specifically better than Rh2(OAc)4 (80-87% yield, dr 9:1-12:1, ee 83-93% as
compared to 70% yield, dr 5:1, ee 10% respectively).12
The next step was the development of applications for these derivatives. We looked more
closely in the fields of asymmetric catalysis and chiral stationary phase (CSP)chromatography. We have showed that these compounds of type 16, can be used for the
aziridination of chalcone derivatives. Moderate selectivity and yields were provided.
However, importantly, it was shown that these TBs are configurationally stable during the
reaction unlike their methano-congeners and can be recovered in good yield (> 90%) at the
end. Also, a new chiral stationary phase (TB-CSP) was developed for analytical purpose using
a designed ethano-Tröger base as chiral selector. Although results were a bit discouraging as
this TB-CSP hardly separated any chiral analyte, it was shown later that it is specific for the
chiral recognition of other TB molecules. These results suggest that compound 16 are very
stiff chiral molecules which hardly undergo an “induced fit” phenomenon upon interactions
with other entities.
In conclusion, we have developed novel protocols which allow (i) a direct (high yielding)
transformation of methano-TBs into ethano-TBs, (ii) highly diastereoselective introductions
of a tertiary or quaternary C-stereogenic centers, (iii) and, most of the times the synthesis of a
variety of highly enantio-enriched adducts which can be used directly or easily transformed
into other functionalized Tröger bases. We have also shown an interesting example of
regioselective ring expansion reaction, which led to a clean chemical distinction of the two
stereogenic nitrogen atoms. With these results in hand, a mechanistic rationale involving
iminium/enolate species as key intermediates has been proposed. This proposition was also in
perfect agreement with the other observations made during the course of the study.
Finally, two applications of these configurationally stable scaffolds were shown which
highlighted their candidature as surrogate of methano-Tröger bases in many of its
applications.
English Version
Reference:
1. J. Tröger, J. Prakt. Chem. 1887, 36, 225-245.
2. a) S. Sergeyev, Helv. Chim. Acta 2009, 92, 415-444; b) B. Dolensky, J. Elguero, V. Kral, C. Pardo and M.
Valik, Adv. Heterocycl. Chem. 2007, 93, 1-56; c) M. Demeunynck and A. Tatibouet, Prog. Heterocycl.
Chem. 1999, 11, 1-20.
3. a) S. Satishkumar and M. Periasamy, Tetrahedron: Asymmetry 2009, 20, 2257-2262; b) M. Valik, J. Cejka,
M. Havlik, V. Kral and B. Dolensky, Chem. Commun. 2007, 3835-3837; c) E.-i. Kim, S. Paliwal and C. S.
Wilcox, J. Am. Chem. Soc. 1998, 120, 11192-11193; d) M. J. Crossley, L. G. Mackay and A. C. Try, J.
Chem. Soc., Chem. Commun. 1995, 1925-1927; e) E. Weber, U. Müller, D. Worsch, F. Vögtle, G. Will and
A. Kirfel, J. Chem. Soc., Chem. Commun. 1985, 1578-1580.
4. a) E. B. Veale, D. O. Frimannsson, M. Lawler and T. Gunnlaugsson, Org. Lett. 2009, 11, 4040-4043; b) A.
Tatibouët, M. Demeunynck, C. Andraud, A. Collet and J. Lhomme, Chem. Commun. 1999, 161-162.
5. C. S. Wilcox and M. D. Cowart, Tetrahedron Lett. 1986, 27, 5563-5566.
6. a) T. Weilandt, U. Kiehne, G. Schnakenburg and A. Lutzen, Chem. Commun. 2009, 2320; b) Y. M. Jeon, G.
S. Armatas, D. Kim, M. G. Kanatzidis and C. A. Mirkin, Small 2009, 5, 46-50; c) U. Kiehne, T. Bruhn, G.
Schnakenburg, R. Frohlich, G. Bringmann and A. Lützen, Chem. Eur. J. 2008, 14, 4246-4255; d) M. S.
Khoshbin, M. V. Ovchinnikov, C. A. Mirkin, J. A. Golen and A. L. Rheingold, Inorg. Chem. 2006, 45, 26032609.
7. V. Prelog and P. Wieland, Helv. Chim. Acta 1944, 27, 1127-1134.
8. a) D. A. Lenev, K. A. Lyssenko, D. G. Golovanov, V. Buss and R. G. Kostyanovsky, Chem. Eur. J. 2006, 12,
6412-6418; b) O. Trapp, G. Trapp, J. W. Kong, U. Hahn, F. Vögtle and V. Schurig, Chem. Eur. J. 2002, 8,
3629.
9. a) Y. Hamada and S. Mukai, Tetrahedron: Asymmetry 1996, 7, 2671-2674; b) D. A. Lenev, D. G.
Golovanov, K. A. Lyssenko and R. G. Kostyanovsky, Tetrahedron: Asymmetry 2006, 17, 2191-2194.
10. C. Michon, A. Sharma, G. Bernardinelli and E. Francotte, J. Lacour Chem. Commun. 2010, 46, 2206-2208.
11. A. Sharma, L. Guenee and J. V. Naubron, J. Lacour Angew. Chem. Int. Ed. 2011, 50, 3677-3680.
12. A. Sharma, C. Besnard, L Guénée; J. Lacour, J. Org. Biomol. Chem. 2012, 10, 966-969.
Table of Contents
Table of Contents
1. Tröger base − A Privileged Molecule
1
1.1
Historical Development
1
1.2
Stereoselective Synthesis of Tröger Bases
6
1.3
Applications
8
1.4
Configurationally Stable Analogues
14
1.5
Ethano-Tröger Bases
15
2. Rearrangement Reactions of Ammonium Ylides
17
2.1
[1,2]-Stevens Rearrangement
17
2.2
Mechanism of the [1,2]-Stevens Rearrangement
18
2.3
Methods to generate the Ammonium Ylides
22
2.3.1 Base-induced ylide generation.
22
2.3.2 Ylide formation via desilylation
23
2.3.3 Ylide formation under Mitsunobu conditions
24
2.3.4 Ylide formation via carbenes
25
2.4
Ylides via metallocarbenes and subsequent rearrangement
25
2.4.1 Diazo-compound influence
26
2.4.2 Metal catalyst influence
27
2.5
Enantiospecific [1,2]-Stevens Rearrangements
32
2.5.1 Base promoted processes
33
2.5.2 Metallocarbenes promoted processes
36
3. Two Step Synthesis of Ethano-Tröger Bases
3.1
Preamble
39
39
Table of Contents
3.2
Base Catalyzed Rearrangement
40
3.3
Acid Catalyzed Rearrangement
47
3.4
“Control” Experiments
48
3.5
Functionalization of Ethano-Tröger Bases
51
3.6
Enantiospecificity/Chirality Transfer
56
4. One-Step Synthesis using Metallocarbenes
59
4.1
General Considerations
59
4.2
Synthesis of Diazo-Compounds
60
4.3
Rhodium-catalyzed Synthesis of Ethano-Tröger Bases
61
4.3.1 Initial discovery
61
4.3.2 Substrate Screening
62
4.3.3 Enantiospecificity / Chirality transfer
65
4.3.4 Scope and limitations of the reaction
70
4.4
Copper-catalyzed Synthesis of Ethano-Tröger Bases
72
4.4.1 Screening of copper sources
72
4.4.2 Dichotomy in copper and rhodium based catalysis
74
4.4.3 Enantiospecificity
76
4.5
Mechanistic Insight
80
4.5.1. Proposition
80
4.5.2. Elements confirming the mechanistic rationale
81
4.5.3. In-situ monitoring
88
4.5.4. Conclusion
90
5. Applications of Ethano-Tröger Bases
91
5.1
Asymmetric Catalysis
91
5.2
Applications in CSP-Chromatography
93
5.2.1 Grafting of ethano-Tröger base
93
5.2.2 Chiral resolution using Tröger base grafted CSP (TB-CSP)
96
Table of Contents
6. Conclusion and Perspectives
101
6.1
Conclusions
101
6.2
Perspectives
101
7. Experimental Part
105
7.1
Generalities
105
7.2
General Procedure for the Alkylation of rac-1
106
7.3
General procedure for the rearrangement of ammonium salts of rac-1110
7.3.1 Basic alumina (Al2O3) assisted rearrangement
110
7.3.2 Hexamine assisted rearrangement
110
7.3.3 Acid/Alcohol assisted rearrangement
111
7.3.4 Reduction of ethano-Tröger bases
115
7.3.5 Nucleophile addition on 64
116
7.3.6 Functionalization of ethano-Tröger base
118
7.4
One-Step Synthesis using Metallocarbenes
121
7.5
Synthesis of Chiral Selector TB-CSP:
143
7.6
Characterization of 80 by NMR Analysis:
145
7.7
Characterization of Salt 85
150
7.8
CSP-HPLC Data
152
7.9
Crystallographic Data
164
7.8.1 Compound rac-64
164
7.8.2 Compound 67
165
7.8.3 Compound 73e
166
7.8.4 Compound 73g
167
7.8.5 Compound 73m
168
7.8.6 Compound 73me
169
7.8.7 Compound 73o
170
7.8.8 Compound 73o
171
Chapter 1
1. Tröger base − A Privileged Molecule
1.1
Historical Development
Tröger base (TB, 1) is a tertiary diamine which was first discovered in 1887 by Julius Tröger,
as an unexpected product from the reaction of p-toluidine with methylal (CH2(OCH3)2).1
Several possible structures were proposed but it remained an unsolved puzzle for the chemists
for almost 50 years. In 1935, Spielman elucidated the correct structure of 1 (Figure 1-1),2
which was confirmed later by X-ray diffraction analysis.3
Figure 1-1 Tröger base 1
After almost six years, in 1941 Wagner et al proposed a sequence of reactions possibly
involved in the formation of compound 1 (Scheme 1-1).4 This sequence involved aminal
formation followed by an electrophilic aromatic substitution reaction. The presence of parasubstituent was crucial to avoid polymerization. In the same paper, the authors also reported
different analogues of type 1 bearing similar geometry but different substituents on the
aromatic moieties. These compounds of type 1 have attracted considerable attention because
of their rigid tricyclic core in which aromatic moieties are perfectly perpendicular to each
other (Scheme 1-1).5
1
J. Tröger, J. Prakt. Chem. 1887, 36, 225-245.
M. A. Spielman, J. Am. Chem. Soc. 1935, 57, 583-585.
3
S. B. Larson, C. S. Wilcox, Acta Crystallogr., Sect. C 1986, 42, 224-227.
4
T. R. Miller, E. C. Wagner, J. Am. Chem. Soc. 1941, 63, 832-836.
5
(a) B. G. Bag, Curr. Sci. 1995, 68, 279-288; (b) M. Demeunynck, A. Tatibouet, Prog. Heterocycl. Chem. 1999,
11, 1-20; (c) B. Dolensky, J. Elguero, V. Kral, C. Pardo, M. Valik, Adv. Heterocycl. Chem. 2007, 93, 1-56; (d) S.
Sergeyev, Helv. Chim. Acta 2009, 92, 415-444.
2
-1-
Tröger Base – A Priviliged Molecule
This peculiar geometry imparts a variety of applications particularly in the field of
supramolecular chemistry such as molecular recognition,6 DNA-interacting probes,7
biomimetic systems,8 self-assembled structures9 and designing materials with optical and
optoelectronic properties.10
Scheme 1-1 Wagner’s mechanism for the synthesis of methano-Tröger base
Tröger base 1 however came to limelight in 1944 when future Nobel-Prize winner Vladimir
Prelog recognized the chiral nature of 1 and also that it could exist as stable enantiomers at
room temperature (Figure 1-1).11 This stability towards inversion is due to the rigidity
6
(a) E. Weber, U. Müller, D. Worsch, F. Vögtle, G. Will, A. Kirfel, J. Chem. Soc., Chem. Commun. 1985, 15781580; (b) M. J. Crossley, L. G. Mackay, A. C. Try, J. Chem. Soc., Chem. Commun. 1995, 1925-1927; (c) E.
Kim, S. Paliwal, C. S. Wilcox, J. Am. Chem. Soc. 1998, 120, 11192-11193; (d) M. Valik, J. Cejka, M. Havlik, V.
Kral, B. Dolensky, Chem. Commun. 2007, 3835-3837; (e) S. Satishkumar, M. Periasamy, Tetrahedron:
Asymmetry 2009, 20, 2257-2262.
7
(a) A. Tatibouët, M. Demeunynck, C. Andraud, A. Collet, J. Lhomme, Chem. Commun. 1999, 161-162; (b) E.
B. Veale, D. O. Frimannsson, M. Lawler, T. Gunnlaugsson, Org. Lett. 2009, 11, 4040-4043; (c) E. B. Veale, T.
Gunnlaugsson, J. Org. Chem. 2010, 75, 5513-5525.
8
C. S. Wilcox, M. D. Cowart, Tetrahedron Lett. 1986, 27, 5563-5566.
9
(a) M. S. Khoshbin, M. V. Ovchinnikov, C. A. Mirkin, J. A. Golen, A. L. Rheingold, Inorg. Chem. 2006, 45,
2603-2609; (b) T. Weilandt, U. Kiehne, G. Schnakenburg, A. Lutzen, Chem. Commun. 2009, 2320-2322; (c) T.
Weilandt, U. Kiehne, J. Bunzen, G. Schnakenburg, A. Lutzen, Chem. Eur. J. 2010, 16, 2418-2426; (d) Y. M.
Jeon, G. S. Armatas, D. Kim, M. G. Kanatzidis, C. A. Mirkin, Small 2009, 5, 46-50.
10
(a) X. T. Tao, C. X. Yuan, Y. Ren, Y. Li, J. X. Yang, W. T. Yu, L. Wang, M. H. Jiang, J. Phys. Chem. C 2007,
111, 12811-12816; (b) X. T. Tao, C. X. Yuan, Q. A. Xin, H. J. Liu, L. Wang, M. H. Jiang, Sci. China. Chem.
2011, 54, 587-595.
11
V. Prelog, P. Wieland, Helv. Chim. Acta 1944, 27, 1127-1134.
-2-
Chapter 1
(conformational strain for pyramidal racemization) provided by the methano-bridge present
between the two nitrogen atoms. In the same article, Prelog et al also successfully resolved
rac-1 by chromatography on a chiral stationary phase, using D-lactose.11 This was the first
example of a resolution of an amine wherein the chirality is solely due to stereogenic Natom(s) and also the first example of preparatively useful application of chromatography on a
chiral stationary phase. Since then, it became the popular model analyte for the evaluation of
chiral chromatography techniques.5
Scheme 1-2 Racemization pathway of 1 via iminium intermediate 2
Yet, Prelog et al also reported that TB-1 is configurationally labile and can undergo acid
catalyzed racemization. It was postulated later that it proceeds via the reversible formation of
an iminium intermediate 2 (Scheme 1-2).12 This pro-chiral iminium intermediate (2) can close
from both re and si faces, to give both enantiomers and hence lead to the racemization
(Scheme 1-2). Uncorrected data for barrier of racemization was also proposed (barrier of
inversion 79.1-94.6 kJ/mol) by comparison with eight member ring 5,11-diacetyl-5,6,11,12tetrahydrodibenzo [b,f][1,5]diazocine, a quite reasonable model for 2.
Greenberg et al investigated the mechanistic proposal. However, the authors did not find any
spectroscopic evidence to support this assumption presumably due to the undetectable amount
12
(a) O. Trapp, V. Schurig, J. Am. Chem. Soc. 2000, 122, 1424-1430; (b) O. Trapp, G. Trapp, J. W. Kong, U.
Hahn, F. Vögtle, V. Schurig, Chem. Eur. J. 2002, 8, 3629-3634; (c) D. A. Lenev, D. G. Golovanov, K. A.
Lyssenko, R. G. Kostyanovsky, Tetrahedron: Asymmetry 2006, 17, 2191-2194.
-3-
of 2 involved in racemization process.13 Their studies also implied that this racemization
occurs more readily in dilute rather than concentrated acid. They suggested that, this
observation is due to the protonation of both nitrogen atoms of 1 in concentrated acid
conditions which then avoids formation of the postulated iminium intermediate.
Later in 2000, Schurig et al revisited the racemization process of 1 in the gas and liquid
phases.12a They showed that methano-Tröger base-114 also undergoes racemization without
the assistance of an acid or a catalyst and also suggested that the enantiomerization in gas
medium proceeds via a degenerated retro-hetero-Diels-Alder ring opening yielding compound
3 (which undergo subsequent Diels−Alder cyclisation) rather than the formation of the
zwitterionic species 4 (Scheme 1-3). Later, the same group also studied the enantiomerization
of 1 under acidic conditions using an enantioselective dynamic electrokinetic chromatography
technique.12b These studies have shown that 1 undergo relatively facile racemization under
acidic conditions (barrier of inversion ca. 100.9 kJ mol-1, t1/2 = 886 min at 25 °C) and
introduction of permanent positive charge on 1 (by alkylating one of the nitrogen with benzyl
bromide) again significantly lower the barrier (barrier of inversion ca. 90.2 kJ mol -1, t1/2 = 12
min at 25 °C).12b
Scheme 1-3 Racemization pathway of 1 via retro Diels−Alder reaction
13
A. Greenberg, N. Molinaro, M. Lang, J. Org. Chem. 1984, 49, 1127-1130.
In this thesis, Tröger base 1 with methylene bridge in between the two nitrogen atoms is also referred as
methano-Tröger base.
14
-4-
Chapter 1
Recently, Schröder et al also studied the epimerization of bis-Tröger bases in the gas phase
using ion-mobility mass spectroscopy (IM-MS) (Scheme 1-4).15 They suggested that this
epimerization proceeds via formation of zwitterionic species rather than through the formation
of a degenerated retro-hetero-Diels−Alder ring opening compound instead.
Scheme 1-4 Syn-anti epimerization of bis-Tröger base
Mysteries continued with this molecule, as determination of the absolute configuration of
Tröger base and its derivatives was still an issue to be solved. The first tentative assignment
was based on the comparison of chiroptical properties of 1 with argemonine alkaloid which
lead to the assignment of compound (5S,11S)-1 as dextrorotatory enantiomer.16 Few months
later in 1967, Mason et al assigned opposite configuration i.e. (+)-(5R,11R) using circular
dichroism (CD) for the analysis and an exciton coupling method, in particular.17 Then, in
1991 Wilen et al unambiguously established the absolute configuration of (+)-1 by X-ray
diffraction (XRD) analysis of the salt of (+)-1 with anion (−)-5 (Figure 1-2) to be (+)-(5S,11S)
which was opposite to that assigned by Mason.18
Figure 1-2 Chiral acids used for resolution of Tröger base
This paper from Wilen also presented for the first time, the isolation of enantioenriched
Tröger base 1 (ee 98%) through the formation of diastereomeric salts with chiral acids (−)-5
(Figure 1-2). This process, initially believed to be impossible due to the on-going
15
D. Schroder, A. Revesz, T. A. Rokob, M. Havlik, B. Dolensky, Angew. Chem. Int. Ed. 2011, 50, 2401-2404.
O. Cervinka, A. Fabryova, V. Novak, Tetrahedron Lett. 1966, 5375.
17
S. F. Mason, G. W. Vane, Schofiel.K, R. J. Wells, J. S. Whitehurst, J. Chem. Soc. B 1967, 553-556.
18
S. H. Wilen, J. Z. Qi, P. G. Williard, J. Org. Chem. 1991, 56, 485-487.
16
-5-
racemization under acidic conditions. It was also reported that, this acid (−)-5 not only leads
to the precipitation of one of the diastereomeric ion-pair but would also induce an
enantiomerization and hence the resolution is coupled to a crystallization-induced asymmetric
transformation (CIAT) to give after basification, (+)-1 in 93% yield.
This assignment of the (+)-(5S,11S) absolute configuration was later reconfirmed by
Periasamy and co-workers by XRD analysis of the salt of (+)-1 with 6 (Figure 1-2).19 The
cliché details the bond lengths of the carboxylic acid groups and OH─N distances clearly
indicate that the aggregate of 1 and acid 6 is not an actual salt, but rather an H-bonded
aggregate.
Finally, the calculated and measured vibrational circular dichroism (VCD) spectra of 1 were
determined and they are also in accord with this assignment.20 This assignment of absolute
configuration was then extended to other analogues of Tröger base by comparison of the sign
of the lowest-energy cotton effect in CD spectra.21 Though recently, Lützen et al have clearly
shown that the direct comparison of CD spectra is sometimes unreliable as the sign and the
magnitude of different cotton effect may change dramatically between various Tröger base
analogues differing only in substitution patterns of the aromatic ring.22
1.2
Stereoselective Synthesis of Tröger Bases
As mentioned, Tröger base 1 is classically synthesized by condensation of anilines and
formaldehyde under strong acidic conditions.5 Many other procedures have been reported
using different reagents which can act as formaldehyde equivalent like paraformaldehyde,23
hexamethylenetetramine24 or dimethoxymethane,25 including procedures involving pummerer
fragmentation of DMSO;26 These alternatives being used to make derivatives of methanoTröger bases containing heterocyclic rings, in particular.27 Studies from different research
19
S. Satishkumar, M. Periasamy, Tetrahedron: Asymmetry 2006, 17, 1116-1119.
A. Aamouche, F. J. Devlin, P. J. Stephens, J. Am. Chem. Soc. 2000, 122, 2346-2354.
21
(a) S. Sergeyev, F. Diederich, Angew. Chem. Int. Ed. 2004, 43, 1738-1740; (b) D. A. Lenev, K. A. Lyssenko,
D. G. Golovanov, O. R. Malyshev, P. A. Levkin, R. G. Kostyanovsky, Tetrahedron Lett. 2006, 47, 319-321.
22
U. Kiehne, T. Bruhn, G. Schnakenburg, R. Frohlich, G. Bringmann, A. Lützen, Chem. Eur. J. 2008, 14, 42464255.
23
I. Sucholeiki, V. Lynch, L. Phan, C. S. Wilcox, J. Org. Chem. 1988, 53, 98-104.
24
T. H. Webb, H. Suh, C. S. Wilcox, J. Am. Chem. Soc. 1991, 113, 8554-8555.
25
B. G. Bag, U. Maitra, Synth. Commun. 1995, 25, 1849-1856.
26
D. P. Becker, P. M. Finnegan, P. W. Collins, Tetrahedron Lett. 1993, 34, 1889-1892.
27
B. Cekavicus, E. Liepinsh, B. Vigante, A. Sobolevs, J. Ozols, G. Duburs, Tetrahedron Lett. 2001, 42, 42394241.
20
-6-
Chapter 1
groups have suggested that a strong acid is indeed necessary to obtain 1. These studies also
suggested that electronic and steric effects of the substituents on the aromatic parts have the
detrimental influence on the yield of Tröger bases.23,28
A step-wise procedure was developed by Wilcox et al to make unsymmetrically substituted
methano-Tröger bases.29 Following this protocol many interesting scaffolds have been made.
This procedure, although quite efficient, requires again strong acidic conditions which make
the prospects of asymmetric synthesis unlikely because of the configurational lability of these
compounds of type 1 under acidic conditions.
In search for a general stereoselective process for the synthesis of Tröger base, Metlesics et al
reported in 1966 a diastereoselective formation of phenyl substituted Tröger bases starting
with cis or trans tetrahydro-diphenyl[1,5]diazocine 7 and formaldehyde (Equation 1-1).30
They showed that the trans- and cis-endo isomers 8, quantitatively isomerize to the cis-exo
isomer 9 by treatment with sodium hydride.
Equation 1-1 Diastereoselective synthesis of Tröger base analog 9
In 1991, Wilcox et al reported an elegant synthesis via the diastereoselective cyclization of a
chiral precursor.24 Following this report, Maitra et al reported for the first time an induction of
asymmetry using 7-deoxycholic acid steroid as chiral template (Equation 1-2). The authors
were able to obtain 77% yield of the major diastereomer (de 70%).5,25,31 They also showed
that the diastereoselectivity of the cyclization depends on the spacers linking the two aniline
fragments to the steroid and the best result was obtained with m = 1, n = 2 (Equation 1-2).
28
(a) L. I. Smith, W. M. Schubert, J. Am. Chem. Soc. 1948, 70, 2656-2661; (b) M. J. Crossley, A. C. Try, R.
Walton, Tetrahedron Lett. 1996, 37, 6807-6810; (c) J. Cudero, C. Pardo, M. Ramos, E. Gutierrez-Puebla, A.
Monge, J. Elguero, Tetrahedron 1997, 53, 2233-2240; (d) M. Demeunynck, C. Fontaine, J. Lhomme, Magn.
Reson. Chem. 1999, 37, 73-76; (e) W. W. K. R. Mederski, M. Baumgarth, M. Germann, D. Kux, T. Weitzel,
Tetrahedron Lett. 2003, 44, 2133-2136.
29
T. H. Webb, C. S. Wilcox, J. Org. Chem. 1990, 55, 363-365.
30
W. Metlesics, R. Tavares, L. H. Sternbach, J. Org. Chem. 1966, 31, 3356-3362.
31
U. Maitra, B. G. Bag, J. Org. Chem. 1992, 57, 6979-6981.
-7-
Equation 1-2 Enantioselective synthesis of methano-Tröger base
Recently, Král et al have also attempted a synthesis by chiral induction using (R or S)-1phenylethanol as chiral auxiliary for the preparation of four pyrrolo fused Tröger bases, 32 but
they observe no chiral induction in the product. Interestingly, though the final mixture of
diastereomers can undergo CIAT to give only one diastereomer on treatment with dilute acid
in methanol (Equation 1-3).
Equation 1-3
Although, these methodologies shown above were efficient in producing the enantioenriched
Tröger base derivatives, but it is specific to particular Tröger base skeleton only. And hence
there is need to develop general asymmetric processes to make enantiomerically pure TB.
1.3
Applications
Since its discovery, Tröger base has been used as a standard probe and to test the efficiency of
newly designed chiral stationary phases (CSP) and also for various chromatographic
processes.
As mentioned earlier, Tröger base chemistry is also widely explored in supramolecular
chemistry whereas asymmetric transformations using Tröger bases as ligand or auxiliary itself
32
M. Valik, B. Dolensky, E. Herdtweck, V. Kral, Tetrahedron: Asymmetry 2005, 16, 1969-1974.
-8-
Chapter 1
remained relatively unexplored.5 In the following paragraphs, some examples of applications
have been selected which clearly demonstrate the potential of TB.
In 1994,33 Wilcox and co-workers developed a molecular design using Tröger base as a
structural anchor, called as ‘Molecular torsion balance (MTB)’ (Scheme 1-5) to measure weak
intra-molecular interactions like aromatic edge-to-face interactions, CH– interactions. These
results were extrapolated later to intermolecular situations.
Scheme 1-5 Wilcox’s molecular torsion balance to measure weak supramolecular interactions
This MTB approach quantifies these forces at play by measuring the small difference in free
energy between folded and unfolded conformers using NMR experiments. Further, improved
design for the pendulum MTB was made; to offer solubility in water and hence avoids
corrections for the change of dipole moment between folded and unfolded conformers.34
This original design from Wilcox and co-workers was further used by Diederich to develop a
new balance comprising an indole fragment to provide evidence for the existence of attractive
orthogonal dipolar interactions between a Csp2−F bond and an amide CO group.35 These weak
attractions between F-atoms and an amide group are of primary importance for the
understanding of protein folding, for the rational design of enzyme inhibitors and also in
medicinal-chemistry research.36
Furthermore, different self-assembly of metallohelicates were developed using Tröger base as
a backbone structure.37 Interestingly, all these self-assemblies have shown a great preference
to form a single stereoisomer.38 These double- and triple-stranded helicates form a relatively
large cavity that makes them attractive ‘molecular container’.
33
S. Paliwal, S. Geib, C. S. Wilcox, J. Am. Chem. Soc. 1994, 116, 4497-4498.
C. S. Wilcox, B. Bhayana, Angew. Chem. Int. Ed. 2007, 46, 6833-6836.
35
F. Diederich, F. Hof, D. M. Scofield, W. B. Schweizer, Angew Chem Int Ed. 2004, 43, 5056-5059.
36
S. Purser, P. R. Moore, S. Swallow, V. Gouverneur, Chem. Soc. Rev. 2008, 37, 320-330.
37
U. Kiehne, A. Luetzen, Eur. J. Org. Chem. 2007, 5703-5711.
38
U. Kiehne, T. Weilandt, A. Lützen, Org. Lett. 2007, 9, 1283-1286
34
-9-
Figure 1-3 Molecular cavities using Tröger base as basic motif
Continuous efforts have also been made to make use of chirality of Tröger bases. Wilen et al
have used enantiomerically pure Tröger base (+)-(S,S)-1 as a chiral solvating agent and
observed enantiomer discrimination in the
1
H-NMR spectra of racemic alcohols.18
Demeunynck et al also reported a remarkable enantioselective interaction of an acridine
analogue of Tröger base (Figure 1-4) with calf-thymus DNA, presumably, via minor-groove
binding of the V-shaped motif rather than by the intercalation of the planar acridine moiety.7a
Diederich and co-workers have also shown that Tröger base analog 10 can also be used to
selectively functionalize C60 to give tethered compound with excellent diastereoselectivity
(Equation 1-4) and thus afforded enantiomerically pure fullerene derivatives.21a This tethering
further directs the double Bingel cyclopropanation of C60. Saigo et al also studied
independently a similar strategy that allowed the regioselective preparation of bis- and
tetrakis-cyclopropanated adducts of C70 using tether-directed double Bingel cyclopropanation
of C60 with various racemic Tröger base analogues.39
10
11
Equation 1-4
Enantiomerically pure Tröger base has been also used to make molecularly imprinted
polymers (MIP) by copolymerization of methacrylic acid (2-methylprop-2-enoic acid) and
39
Y. Ishida, H. Ito, D. Mori, K. Saigo, Tetrahedron Lett. 2005, 46, 109-112.
-10-
Chapter 1
ethylene glycol dimethacrylate followed by removal of template. The resulting MIP showed
high enantioselectivity in the separation of Tröger base by HPLC40 or capillary liquid
chromatography/electrochromatography.41 Chiral recognition of an analyte by MIP takes
place in chiral cavities created during the polymerization process and featuring the shape of
the template used for the imprinting. Since the general shape of the molecule remains similar
for different Tröger base analogues, one may expect considerable cross-selectivity of soprepared MIP towards other Tröger base analogues. This can be a viable alternative to HPLC
on chiral stationary phases for the separation of synthetically valuable Tröger base
derivatives, and possibly of other chiral molecules.5d
Despite all these efforts, relatively little has been reported to exploit the potential of Tröger
bases in asymmetric transformations, as organo-catalyst or as enantiopure ligand. In most of
the reported cases, unfunctionalized methano-Tröger base 1 has been used just to test the
catalytic efficiency rather than making an attempt to develop a rational design of 1 for
particular applications.
Xu et al used (−)-1 as an additive in the 1,4-addition reactions of aryllithium reagents to ,unsaturated t-butyl esters afforded the corresponding 1,4-addition product 12 in 57% ee
(Equation 1-5) .42
Equation 1-5
Baiker et al have also studied enantioselective heterogeneous hydrogenation of ethyl pyruvate
over alumina- and carbon-supported Pt-metal catalysts modified by Tröger base 1 resulting in
corresponding hydrogenated product in 65% ee (Equation 1-6).43
40
(a) K. Adbo, H. S. Andersson, J. Ankarloo, J. G. Karlsson, M. C. Norell, L. Olofsson, J. Svenson, U. Ortegren,
I. A. Nicholls, Bioorg. Chem. 1999, 27, 363-371; (b) K. Adbo, I. A. Nicholls, Anal. Chim. Acta 2001, 435, 115120.
41
(a) J. Ou, X. Li, S. Feng, J. Dong, X. Dong, L. Kong, M. Ye, H. Zou, Anal. Chem. 2007, 79, 639-646; (b) J.
Ou, J. Dong, T. Tian, J. Hu, M. Ye, H. Zou, J. Biochem. Biophys. 2007, 70, 71-76.
42
F. Xu, R. D. Tillyer, D. M. Tschaen, E. J. J. Grabowski, P. J. Reider, Tetrahedron: Asymmetry 1998, 9, 16511655.
43
Minder, B.; Schurch, M.; Mallat, T.; Baiker, A. Catal. Lett. 1995, 31, 143.
-11-
Equation 1-6
Harmata et al have prepared 14, a derivative of 1 and used it for the addition of Et2Zn to
aromatic aldehydes lead to rather encouraging results (ee up to 86%, Equation 1-7).44. To the
best of our knowledge, this is the only report where a functionalized derivative of 1 was
prepared to develop a rational design of catalyst derived from Tröger base 1.
Equation 1-7 Tröger base as a ligand for diethyl zinc addition
There are few other reports where methano-Tröger base 1 was used as a ligand in
organometallic processes, though poor results were obtained possibly due to racemization of
1. Sigman et al has used it as a ligand for palladium catalyzed kinetic resolution of secondary
alcohols with little success;45 and Hermann et al have also used a preformed complex of
methyltrioxorhenium(VII) and 1 for epoxidation of olefins and oxidation of sulfides. This
catalytic system results in good reactivity although no induction of asymmetry was
observed.46
Hermann et al also reported in the same paper, the XRD-spectrum of the yellow complex of
(+)-1 with methyltrioxorhenium(VII). This represents the first XRD-spectrum of a complex
containing TB as ligand ever been reported. It’s noteworthy that the presence of the rhenium
atom in the XRD-spectrum could also allow them to assign the absolute configuration of
Tröger base (+)-1 as (5S,11S). Surprisingly, wrong absolute configuration was assigned in this
44
M. Harmata, M. Kahraman, Tetrahedron: Asymmetry 2000, 11, 2875-2879.
M. S. Sigman, D. R. Jensen, Acc. Chem. Res. 2006, 39, 221-229.
46
W. A. Herrmann, F. E. Kuhn, M. R. Mattner, G. R. J. Artus, M. R. Geisberger, J. D. G. Correia, J. Organomet.
Chem. 1997, 538, 203-209.
45
-12-
Chapter 1
paper possibly due to the wrong absolute configuration written on the commercially available
product.47
Shi et al have also used (+)-1 as an organocatalyst for the aziridination of chalcones
derivatives using MSH. The authors have also shown that this aziridination process can be
done using catalytic amount of (+)-1 (60 mol%) to provide corresponding aziridine up to 67%
ee (Equation 1-8).48
Equation 1-8
Recently, Sergeyev et al made thiourea derivatives of TB for Michael additions of malonate
derivatives to trans-β-nitrostyrene.49 The authors also observed partial racemization of the
methano-TB derived 15 isolated after the reaction (Equation 1-9).
Equation 1-9
These examples demonstrated the potential of the Tröger base motif for asymmetric
transformations which is however counter balanced by facile racemization of 1.
47
As mentioned, (+)-1 was bought from sigma-aldrich with wrong. (5R,11R) absolute configuration.
Y. M. Shen, M. X. Zhao, J. Xu, Y. Shi, Angew. Chem. Int. Ed. 2006, 45, 8005-8008.
49
D. Didier, S. Sergeyev, Arkivoc 2009, 124-134.
48
-13-
1.4
Configurationally Stable Analogues
As mentioned above, Tröger base chemistry has developed in a remarkable manner in terms
of structural diversity and applications in supramolecular chemistry but a general lack of
applications are observed in terms of asymmetric transformations. The most likely reason
being facile racemization of methano-TB.
Few solutions have been brought forward to address this issue. In 1996, Hamada et al
introduce an ethano-bridge analogue 16 of methano-bridge-1 (Figure 1-4).50 The presence of
this ethano-bridge avoids the formation of classically proposed racemization pathway through
iminium intermediates of type 2.12c This analogue 16 also has the same global geometry as 1
and hence could be used for similar applications in supramolecular chemistry or synthesis.
Later, in 1999 Demeunynck et al introduced acridine substituted Tröger base 17.51 This
derivative possesses two more basic acridine nitrogen atoms, and under acidic conditions,
these are protonated prior to the aniline N-atom which renders the molecule electron deficient,
it hence avoids the protonation of bridged nitrogen and prevents the racemization.
Figure 1-4 Possible pathways to avoid racemization
Recently, Kostyanovsky and co-workers have also reported a considerable increase in the
racemization barrier of Tröger base analogues with substitution in the ortho-positions relative
to the N-atoms (racemization barrier was found to be 130.4 kJ.mol-1 for (S,S)-18 as compared
to 101.4 kJ.mol-1 for (S,S)-1).52 The presence of these ortho-substituents adds a large steric
hindrance to the planarization of iminium species of type 2, responsible for the racemization.
This substitution also leads to a strong steric hindrance at the nitrogen centers and this could
limit applications involving these nitrogen atoms. Lützen et al have used nevertheless this
50
Y. Hamada, S. Mukai, Tetrahedron: Asymmetry 1996, 7, 2671-2674.
M. Demeunynck, A. Tatibouet, C. Andraud, A. Collet, J. Lhomme, Chem Commun 1999, 161-162.
52
D. A. Lenev, K. A. Lyssenko, D. G. Golovanov, V. Buss, R. G. Kostyanovsky, Chem. Eur. J. 2006, 12, 64126418.
51
-14-
Chapter 1
solution to make different analogues of Tröger bases which upon further derivatization lead to
self-assembly of metallohelicates, as shown in previous section.
Of all these different candidates for the substitution of methano-bridge-1, ethano-derivatives
seemed to be the most promising in terms of configurational stability. Moreover, introduction
of ethano-bridge resulted in minimal change in V-shaped geometry and also avoids any
addition steric congestion which could have limited its applications using anilinic nitrogen
atoms.
1.5
Ethano-Tröger Bases
As mentioned above, ethano-Tröger base 16 was first introduced by Hamada et al in 1996 as
a configurationally stable analog of methano-Tröger base.50 Compound 16 was prepared by
the direct reaction of 1 with 1,2-dibromoethane at 105 °C (Equation 1-10). This reaction most
probably proceeds via formation of a quaternary ammonium salt which then undergoes
cleavage of the methylene-bridge and then subsequent intramolecular nucleophilic substitution
to make the ethano-bridge. Quite surprisingly, the same reaction with diazocine core 19
(without the methylene-bridge) failed to give the desired product 16. A series of molecules
bearing different substituent at the 2- and 8-positions was prepared using this method.
Equation 1-10 Classical synthesis of ethano-Tröger base (16)
Interestingly, this ethano-Tröger base was also resolved using (−)-di-p-toluoyl-L-tartaric acid.
The absolute configuration of (−) and (+)-16 thus assigned as (−)-(R,R) and (+)-(S,S)-16, by
direct correlation with CD spectra of (−)-(R,R)-1 and (+)-(S,S)-1.
Later, Lenev et al have shown that the reaction of enantiomerically pure (+)-16 with MeI
produces the corresponding quaternary salt 20 in nearly quantitative yield without
racemization while in case methano-Tröger bases, racemic products are observed.
Furthermore, heating 20 in DMF results in quantitative recovery of enantiomerically pure (+)-15-
16 (Equation 1-11), which is not the case with 1. These experiments clearly demonstrated the
favored configuration stability of ethano-Tröger bases and also the availability of the nitrogen
lone pairs for participation in applications for asymmetric transformations.
More importantly, the crystal structure of (+)-20 was obtained, which allowed the authors to
determine the absolute configuration. According to the X-ray structure determination, the
absolute configuration of (+)-20 was (R,R). By correlation, the absolute configuration as (+)16 is (R,R) as well. This assignment was clearly opposite to the one assigned by Hamada
using similarity in the CD spectra. This is thus another example which clearly highlights the
danger of using ECD only in determination of the absolute configuration of these molecules.
Equation 1-11 Configuration stability of ethano-Tröger base
The X-ray structure also indicated a global V-shaped structure for 20, characteristic of Tröger
bases. This together with all the prior information was a strong indication that compounds 16
ought to be effective alternatives to Tröger base of type-1. The remainder of thesis will
therefore be devoted to ethano-Tröger chemistry and to novel direct routes to access them, in
particular.
-16-
Chapter 2
2. Rearrangement Reactions of Ammonium Ylides
In organic synthesis, rearrangement reactions occupy a special place by virtue of their
inherent high efficiency and selectivity, and to the wide spectrum of possible chemical
transformations that they afford.1 In this broad class of rearrangement processes, sigmatropic
reactions have also a unique place, being in general highly regio- and stereoselective and
hence providing powerful tools for the synthetic chemists.1e,2 In all-carbon molecules, under
non-catalyzed processes, these rearrangements however require high temperatures and
stringent conditions. The introduction of suitably-placed electron-withdrawing heteroatoms
renders these reactions more accessible by activating same initial step and lowering the
activation energy of the rate determining step.
This chapter will cover rearrangement reactions activated by positive nitrogen atoms and
particularly examples of [1,2]-Stevens rearrangement. This reaction provides unique access to
different types of nitrogen containing compounds that are often natural products belonging to
different classes of alkaloids.2,3 This chapter will also cover a brief introduction on different
methods to make ammonium ylides, the key synthetic intermediate of these reactions.
2.1
[1,2]-Stevens Rearrangement
In 1928, while studying amine-protecting groups, Stevens and co-workers discovered a novel
[1,2]-shift reaction of ammonium ylides (Equation 2-1).4 They reported that on treatment of
quaternary ammonium salt 21 (phenacylbenzyldimethyl-ammonium bromide) with sodium
1
(a) H. E. Zimmerman, D. Armesto, Chem. Rev. 1996, 96, 3065-3112; (b) A. H. Li, L. X. Dai, V. K. Aggarwal,
Chem. Rev. 1997, 97, 2341-2372; (c) T. Ibuka, Chem. Soc. Rev. 1998, 27, 145-154; (d) W. Adam, T.
Heidenfelder, Chem. Soc. Rev. 1999, 28, 359-365; (e) H. Ito, T. Taguchi, Chem. Soc. Rev. 1999, 28, 43-50; (f) A.
M. M. Castro, Chem. Rev. 2004, 104, 2939-3002; (g) T. J. Snape, Chem. Soc. Rev. 2008, 37, 2452-2458; (h) T.
X. Metro, B. Duthion, D. G. Pardo, J. Cossy, Chem. Soc. Rev. 2010, 39, 89-102.
2
(a) J. A. Vanecko, H. Wan, F. G. West, Tetrahedron 2006, 62, 1043-1062; (b) J. Sweeney, Chem. Soc. Rev.
2009, 38, 1027-1038.
3
(a) M. P. Doyle, M. A. McKervey, T. Ye, Modern Catalytic Methods for Organic Synthesis with Diazo
Compounds: From Cyclopropanes to Ylides. Wiley, New York, N. Y., 1998; (b) F. G. West, J. S. Clark,
Nitrogen, Oxygen and Sulfur Ylide Chemistry 2002, 115-134.
4
T. S. Stevens, E. M. Creigton, A. B. Gordon, M. MacNicol, J. Chem. Soc. 1928, 3193-3197.
-17-
Rearrangement Reactions of Ammonium Ylides
hydroxide led to the formation of tertiary amine product 22 (the 2-(dimethylamino)-3phenylpropiophenone) in 90% isolated yield (Equation 2-1).
Equation 2-1 [1,2]-Stevens rearrangement.
From the start, this reaction was considered to involve the formation of an ammonium ylide
intermediate and then a migration of the benzyl group from the nitrogen atom towards the
adjacent carbon. Later, this process was extended by same group to sulfonium salts which
provide a similar rearrangement.5 Since then, the rearrangement has been further extended to
allylic systems to afford [2,3]-Sigmatropic type of rearrangements that involve the formation
of similar ylide intermediate.6 Catalytic procedures have also been developed to generalize the
process and to extend the range of tertiary amines that can be produced. Details are given in
following paragraphs.
2.2
Mechanism of the [1,2]-Stevens Rearrangement
Based on experimental results and the progress in understanding of the concerted reactions,
several hypotheses have been proposed for the [1,2]-Stevens rearrangement. These
mechanisms involve either concerted [1,2]-shift,7 ion-pair5,8 and biradical-pair intermediates
(Scheme 2-1).2
5
6
T. Thomson, T. S. Stevens, J. Chem. Soc. 1932, 55-69.
T. Nakai, K. Mikami, Chem. Rev. 1986, 86, 885-902.
7
(a) T. S. Stevens, J. Chem. Soc. 1930, 2107-2119; (b) R. A. W. Johnstone, T. S. Stevens, J. Chem. Soc. 1955,
4487-4488; (c) A. Campbell, A. H. J. Houston, J. Kenyon, J. Chem. Soc. 1947, 93-95; (d) C. R. Hauser, S. W.
Kantor, J. Am. Chem. Soc. 1951, 73, 1437-1441.
8
J. L. Dunn, T. S. Stevens, J. Chem. Soc. 1932, 1926-1931.
-18-
Chapter 2
Scheme 2-1 Proposed mechanisms: (a) Ion pair; (b) concerted [1,2]-shift; (c) biradical.
Initial studies from Stevens et al using
14
C-labeled starting materials suggested that the
rearrangement is intramolecular as no cross-over product was observed (Equation 2-2).7a,7b,9
Subsequent studies, using various substitutions on the phenyl part of the migrating benzyl
group demonstrated an increase in the migrating ability for the groups containing electronpoor rather than electron-rich substituents.8 These observations led Stevens et al to postulate
an ion-pair mechanism (Scheme 2-1, path a).
Equation 2-2 Cross-over experiment
Later, Wittig10 and Hauser7d proposed a different mechanism (Scheme 2-1, b) to account for
an observation of Kenyon,7c who had reported the retention of configuration of stereogenic
migrating groups during such type of reaction. For them, the ion pair intermediate proposed
9
T. S. Stevens, W. W. Snedden, E. T. Stiller, T. Thomson, J. Chem. Soc. 1930, 2119-2125.
G. Wittig, R. Mangold, G. Felletschin, Liebigs Ann Chem 1948, 560, 116-127.
10
-19-
by Stevens should normally undergo a full loss of the chiral information upon its migration to
the adjacent carbon. They proposed an attack by the carbanionic center of ylide intermediate
at the migrating group intra-molecularly in concerted pathway (Scheme 2-1, path b). For
them, this would entail retention of configuration. Brewster et al also observed retention of
configuration in a chiral spiro bicyclic ammonium system. 11 Schollko12 and Stevens13 also
arrived to the same conclusions that the reaction proceed with high enantiomeric excess with
retention of configuration of migrating center.
Scheme 2-2 First stereoselective Stevens rearrangement.
This proposed mechanism involves a four-active-electrons [1,2]-shift, which according to
Woodward and Hoffmann rules, is symmetry forbidden. Under thermal conditions, this type
of rearrangement should favor thus a supra-antarafacial mode of action rather than suprasupra process. Furthermore, such an approach introduces a very high degree of strain in a
cyclic transition state (TS) and renders the concerted mechanism kinetically unfavorable.
As in other supra-antarafacial processes involving stereogenic migrating groups, two different
pathways can be proposed to rationalized the results, i.e. a suprafacial migration of the R
group that would see its configuration inverted or an antarafacial migration on the negative
carbon with the retention of configuration of the migrating group. Experimentally, this second
type of process seems to be observed. These results in conjunction with Woodward and
Hoffmann rules, opposes the possibility of a concerted mechanism (Scheme 2-1, path b).
Yet, this ambiguity led chemists to consider other mechanisms. In 1983, Ollis et al
demonstrated that [1,2]-Stevens rearrangements can also proceed via non-concerted
mechanism involving radical pairs (Scheme 2-1, path c).14 Evidences were provided by the
isolation and study of the secondary products formed in the process and also using CIDNP15
NMR experiments (Scheme 2-3). The radical pair formed by homolytic cleavage of the most
11
J. H. Brewster, M. W. Kline, J. Am. Chem. Soc. 1952, 74, 5179-5182.
Schollko.U, U. Ludwig, Osterman.G, M. Patsch, Tetrahedron Lett. 1969, 3415.
13
B. J. Millard, T. S. Stevens, J. Chem. Soc. 1963, 3397.
14
(a) W. D. Ollis, M. Rey, I. O. Sutherland, G. L. Closs, J. Chem. Soc., Chem. Commun. 1975, 543-545; (b) K.
Chantrapromma, W. D. Ollis, I. O. Sutherland, J. Chem. Soc., Perkin Trans. 1 1983, 1049-1061; (c) W. D. Ollis,
M. Rey, I. O. Sutherland, J. Chem. Soc., Perkin Trans. 1 1983, 1009-1027.
15
CIDNP stands for chemically induced dynamic nuclear polarisation.
12
-20-
Chapter 2
reactive carbon-nitrogen bond is held tightly together in a solvent cage. This results in the
retention of absolute stereochemistry of the migrating center which upon rapid recombination
of radical pair, provide the major Stevens [1,2]-products (24a and 24b).
Scheme 2-3 Ollis’s radical pair mechanism.
This solvent cage/rapid recombination pathway would also account for the high degree of
intramolecularity and stereoselectivity observed in this rearrangement. The radicals escaping
the solvent cage then undergo a loss of the stereochemical information and their random
recombination to give meso and racemic products 25 and 26.
Nevertheless, this solvent cage model can also be applied for the first heterolytic (iminiumcarbanion) mechanism proposed by Stevens (Scheme 1-1, a). In this case, the retention of the
configuration would occur if the recombination of the intermediates, iminium and benzylic
anion, is fast as the ion pair dissociation. Consequently, both homo-heterolytic pathways are
possible and can be implicated in the [1,2]-Stevens rearrangements depending on substrates
and reaction conditions. Yet, Ollis’ hypothesis in favor of a radical cleavage–recombination
mechanism still remains the most probable and the most considered in the literature.2 This
discussion of both possible mechanisms will be continued in later sections.
-21-
2.3
Methods to generate the Ammonium Ylides
Thanks to the potential of the reaction listed above, ylide intermediates attracted major
attention over the past 50 years, and thus ensuring the development of a wide spectrum of
reactions.16 An ylide is a neutral dipolar molecule containing a formally negatively charged
atom directly attached to a hetero-atom with a formal positive charge. The reactivity of these
intermediates is governed by the nature of the heteroatom and also by the substitution patterns
on the atom containing the formal negative charge.
In recent years, many stable analogues of these intermediates have been isolated. This
provides useful information for understanding of their often diverse reactivity. Discovery of
[1,2]-Stevens rearrangement of ammonium salts, also stimulated the development [1,2]- and
[2,3]- rearrangement using nitrogen, sulfur, and oxygen ylides.3b,6,17 Phosphonium ylides were
used in the Wittig reaction for preparing alkenes from carbonyl groups (C=O). 18 In recent
years, processes involving the formation of ylide intermediates for asymmetric induction were
also developed e.g. epoxides formation (using sulfur, phosphorus ylides) and aziridination
(using nitrogen, sulfur ylides).19
The discovery of milder catalytic conditions for the direct preparation of such entities from
neutral substrates had generated a further interest in these intermediates. The following
paragraph will focus on the reactions involving formation of ammonium ylides exclusively.
2.3.1 Base-induced ylide generation.
After the initial discovery of the [1,2]-Stevens rearrangement, base-induced ylide generation
became the most commonly used method to prepare nitrogen ylides.
16
J. McMurry, Organic chemistry, 7e. ed., Thomson Brooks/Cole, Belmont, CA, 2008.
(a) R. D. Grant, C. J. Moody, C. W. Rees, S. C. Tsoi, J. Chem. Soc., Chem. Commun. 1982, 884-885; (b) R. S.
Gairns, R. D. Grant, C. J. Moody, C. W. Rees, S. C. Tsoi, J. Chem. Soc., Perkin Trans. 1 1986, 491-495; (c) R.
S. Gairns, R. D. Grant, C. J. Moody, C. W. Rees, S. C. Tsoi, J. Chem. Soc., Perkin Trans. 1 1986, 483-489.
18
G. Wittig, U. Schollkopf, Chem Ber-Recl 1954, 87, 1318-1330.
19
(a) G. Y. Fang, O. A. Wallner, N. Di Blasio, X. Ginesta, J. N. Harvey, V. K. Aggarwal, J. Am. Chem. Soc.
2007, 129, 14632-14639; (b) E. M. McGarrigle, E. L. Myers, O. Illa, M. A. Shaw, S. L. Riches, V. K. Aggarwal,
Chem. Rev. 2007, 107, 5841-5883; (c) V. K. Aggarwal, 'Ylide Based Reaction' in Handbook of Organocatalysis,
P. Dalko (Ed.), E. McGarrigle, Wiley-VCH, 2007.
17
-22-
Chapter 2
Scheme 2-4 Ylide intermediate responsible for [1,2]-Stevens rearrangement reaction.
This approach provides the nitrogen ylide directly by deprotonation of the acidic proton
present on the carbon atom  to the quaternary nitrogen atom (Scheme 2-4). This
methodology has been applied for the synthesis of many heterocyclic compounds.2,3,20
The main drawback of the process described above lies in a lack of selectivity in forming the
ylide in presence of more than one acidic site adjacent to the quaternary nitrogen atom.
Another problem lies in the generation of side products such as those resulting from
Hoffmann elimination pathway of the quaternary ammonium salts (Scheme 2-5).
Scheme 2-5 Hoffmann elimination of ammonium salts in basic conditions
2.3.2 Ylide formation via desilylation
To address the regioselectivity problem, Vedejs21 and Sato22 introduced a desilylation
approach using fluoride ion to form the ylide in a controlled manner. This protocol also
reduced the chances of elimination (Equation 2-3).
Equation 2-3 Regioselective formation of ylide intermediate
20
I. E. Markó, in Comprehensive Organic Synthesis, Vol. 3 (Eds.: B. M. Trost, I. Fleming), Pergamon Press,
Oxford, 1991, pp. 913-974.
21
(a) E. Vedejs, G. R. Martinez, J. Am. Chem. Soc. 1979, 101, 6452-6454; (b) E. Vedejs, S. P. Singer, J. Org.
Chem. 1978, 43, 4831-4837.
22
(a) Y. Sato, Y. Yagi, M. Koto, J. Org. Chem. 1980, 45, 613-617; (b) Y. Sato, H. Sakakibara, J. Organomet.
Chem. 1979, 166, 303-307.
-23-
In a few cases, though regioselectivity was nevertheless lost due to proton exchange reactions
scrambling the reaction in favor of more stable ylides prior to migratory insertion.23
2.3.3 Ylide formation under Mitsunobu conditions
In 1966, Walker et al discovered a novel method for ylide formation via intramolecular amine
quaternization under standard Mitsunobu conditions.24 The authors reported the formation of
products 27 and 28 instead of desired product 29, under these reaction conditions (Scheme
2-6). They proposed that the reaction proceed via formation of phosphonium salt 30. This
compound 30 undergoes further intramolecular amine quaternization to form ammonium salt
which under reaction conditions provides the necessary ylide intermediate 31. This ylide
intermediate then rearrange to give the observed major product 27 and a side product 28.
Scheme 2-6
Further, reactions in absence of maleimide and neopentyl alcohol prevented the formation of
28 in the reaction mixture. Further optimizations by replacing DEAD with 2.2 equiv of ADDP
(1,10-(azodicarbonyl)dipiperidine) lead to the product 27 in 77% isolated yield (Scheme 2-6).
23
24
E. Vedejs, F. G. West, Chem. Rev. 1986, 86, 941-955.
M. A. Walker, Tetrahedron Lett. 1996, 37, 8133-8136.
-24-
Chapter 2
2.3.4 Ylide formation via carbenes
Also of the importance of this thesis, ylide intermediates can also made directly through the
reaction of diazo-compounds with tertiary amines. This methodology was first reported by
Stevens et al in 1952 using diazo-compounds as precursors.25 They reported that the nitrogen
ylide can be achieved directly by heating the diazofluorene in the presence of N,Ndimethylbenzylamine, which then subsequently rearrange in tertiary amine 32 in 30% yield
(Scheme 2-7).
Scheme 2-7 Ylide generation via thermal decomposition of diazo-compounds
Later, other methods were developed using photochemical and thermal decomposition of
mainly diazo-compounds, to generate the corresponding carbene intermediates which reacted
as expected with tertiary amines to provide the ylide intermediates. These processes are, not
always efficient and selective.20,26 Discovery of metal catalyzed decomposition of diazocompound provided a milder and efficient access to reactive nitrogen ylides.
2.4
Ylides via metallocarbenes and subsequent rearrangement
As mentioned above, metal catalyzed decomposition of diazo-compound is a viable
alternative to base-promoted ammonium ylide formation or photochemical and thermal diazodecomposition.2a,3b,27 The reaction proceeds through the formation of electrophilic
metallocarbenes which then undergoes nucleophilic addition of tertiary amines to provide
metal bond or metal-free nitrogen ylide intermediates. This process in general, provide highly
reactive ylide intermediates which then undergo subsequent ring-expansion or ring
25
W. R. Bamford, T. S. Stevens, J Chem Soc 1952, 4675-4678.
I. A. D'Yakonov, T. V. Domareva, Zhurnal Obshchei Khimii 1955, 25, 1486-1493.
27
F. F. West, Modern Rhodium-Catalyzed Organic Reactions 2005, 417-431.
26
-25-
contraction reactions to give [1,2]-Stevens like rearrangement.2a,3b This direct generation of
the ylides is highly efficient process. Interestingly, its reactivity and selectivity for migration
can be influenced by different factors including the nature of the diazo-precursor and of the
metal used for the diazo-decomposition.
2.4.1 Diazo-compound influence
Diazo-compounds constitute an important class of molecules widely used in organic
synthesis.2a,3,28 These substrates react with metals or metal complexes to form electrophilic
metal carbenes. The first step, in the generally accepted mechanism of decomposition,
involves the formation of zwitterionic intermediate by electrophilic addition of unsaturated
Lewis acidic metal onto the diazo substrates. The intermediates than undergo spontaneous loss
of dinitrogen to form the metal carbenes (Scheme 2-8). It should be noted that, in C-H
insertion processes the release of nitrogen gas can be the rate determining step of the
reaction.29
Scheme 2-8 General mechanism for metal-catalyzed decomposition of diazo compounds
The carbene intermediates are electrophilic at the carbon center and can hence react with
electron-rich or  bonds and Lewis bases, alike. Then, the metal moiety is released and a
catalytic cycle can then be established by recycling the reactive Lewis acid metal entity. In
some cases, it is described that the metal catalyst remains bounded with the ylide and is only
released after migration or insertion reactions. This method to generate electrophilic
carbenoids is widely used for a variety of reactions among which are C-H, O-H, N-H
insertions,
28
29
cyclopropanation,
addition
reactions,
ylide
formation
M. P. Doyle, R. Duffy, M. Ratnikov, L. Zhou, Chem. Rev. 2010, 110, 704-724.
H. M. L. Davies, J. Du Bois, J. Q. Yu, Chem. Soc. Rev. 2011, 40, 1855-1856.
-26-
and
subsequent
Chapter 2
reactions.2a,3,27 Among these classes of reactions, the focus of this thesis will be the formation
of ammonium ylides and their subsequent reactivity.
The activity of metal complexes towards diazo-decomposition depends both on the
electrophilicity of the metal moiety and on the stability of the diazo-compounds. Generally
observed reactivity trend for diazo-compounds is shown in Figure 2-1.
Figure 2-1 Stability and reactivity of diazo compounds
The most stable are the 1,3-carbonyl-2-diazo compounds and especially 1,3-diesters.
Reactivity increases progressively with decreasing acceptor (electron-withdrawing)30 ability
of the substituents next to the N2 moiety. The most reactive and less stable are those having
the dinitrogen moiety at a terminal carbon.
2.4.2 Metal catalyst influence
As mentioned above, the nature of the metal can influence the outcome of the reaction of
diazo-derivatives. Here, we will discuss only the classical classes of active metal complexes
derived from rhodium and copper. Other metals and their complexes have also been used for
diazo-decomposition reactions will not be covered.31 In this section, the main focus will be to
elaborate the influence of a change of metal sources on the reaction outcome. This discussion
is of importance for results displayed in Chapter 4.
30
G. Maas, Angew. Chem. Int. Ed. 2009, 48, 8186-8195.
(a) G. Maas, Chem. Soc. Rev. 2004, 33, 183-190. (b) C.-Y. Zhou, W.-Y. Yu, P. W. H. Chan, C.-M. Che, J.
Org. Chem. 2004, 69, 7072-7082.
31
-27-
Rh(II) derived catalysts are probably the most effective and widely studied complexes for the
diazo-decomposition reactions.3 Dirhodium (II) tetraacetate was the first complex introduced
in this field.32 Since then, many applications have been reported. This complex has four
bridging acetates as ligands and has D4h symmetry, which leave one vacant coordination site
at each metal position, as shown in Figure 2-2.
Figure 2-2 Dirodhium tetracetate
Many different dirhodium complexes have been synthesized by the exchange of the acetate
ligands, and those made from enantiopure carboxylates to generate chiral catalysts (Figure
2-3).33
Figure 2-3 Dirhodium carboxylates catalysts: examples
Dirhodium carboxamidates constitute another important family of catalysts in the context of
diazo decomposition.34 These complexes have a more rigid structure respect to the
carboxylate derivatives and each ligand is bound to the rhodium atoms through oxygen and a
nitrogen atom. Depending on the nature of the chiral ligand four major families of complexes
32
Paulisse.R, Reimling.H, E. Hayez, A. J. Hubert, P. Teyssie, Tetrahedron Lett. 1973, 2233-2236.
(a) H. Brunner, H. Kluschanzoff, K. Wutz, Bull. Chem. Soc. Belg. 1989, 98, 63-72. (b) M. Kennedy, M. A.
McKervey, A. R. Maguire, G. H. P. Roos, J. Chem. Soc., Chem. Commun. 1990, 361. (c) S. Hashimoto, N.
Watanabe, S. Ikegami, Tetrahedron Lett. 1990, 31, 5173-5174. (d) S. Kitagaki, M. Anada, O. Kataoka, K.
Matsuno, C. Umeda, N. Watanabe, S. Hashimoto, J. Am. Chem. Soc. 1999, 121, 1417-1418. (e) H. Tsutsui, Y.
Yamaguchi, S. Kitagaki, S. Nakamura, M. Anada, S. Hashimoto, Tetrahedron: Asymmetry 2003, 14, 817-821.
34
M. P. Doyle, B. D. Brandes, A. P. Kazala, R. J. Pieters, M. B. Jarstfer, L. M. Watkins, C. T. Eagle,
Tetrahedron Lett. 1990, 31, 6613-6616. (a) M. P. Doyle, W. R. Winchester, J. A. A. Hoorn, V. Lynch, S. H.
Simonsen, R. Ghosh, J. Am. Chem. Soc. 1993, 115, 9968-9978.
33
-28-
Chapter 2
can be considered: pyrrolidinone,35 oxazolidinone,36 imidazolidinone37 and azetidinone
(Figure 2-4).38 Studies from different groups had revealed chemoselectivity of carbenoid
transformations highly influenced by type and electrophilicity of the ligands.39
Figure 2-4 Dirhodium carboxamidate catalysts
Copper derived catalysts are also often used to make ammonium ylides. Oswald Silberrad
discovered that metallic copper is able to catalyze the decomposition of ethyl diazoacetate
(EDA). In the 1960s, some soluble copper catalysts such as copper(I) chloride, copper(II)
acetylacetonate or copper(I) triaryl phosphate were developed for the decomposition of diazo
compounds. .40
In 1972, Hata and Watanabe showed that the ammonium ylide derived from azetidine 33 can
undergo a [1,2]-rearrangement resulting in a ring expansion to provide cyclic amine 34
(Scheme 2-9). 41
Scheme 2-9
35
M. P. Doyle, W. R. Winchester, M. N. Protopopova, A. P. Kazala, L. Westrum, Org. Synth. 1996, 73, 13.
M. P. Doyle, W. R. Winchester, M. N. Protopopova, P. Müller, G. Bernardinelli, D. Ene, S. Motallebi, Helv.
Chim. Acta 1993, 76, 2227-2235.
37
M. P. Doyle, Q.-L. Zhou, C. E. Raab, G. H. P. Roos, S. H. Simonsen, V. Lynch, Inorg. Chem. 1996, 35, 60646073.
38
M. P. Doyle, W. Hu, I. M. Phillips, C. J. Moody, A. G. Pepper, A. G. Z. Slawin, Adv.Synth. Cat. 2001, 343,
112-117.
39
(a) A. Padwa, D. J. Austin, S. F. Hornbuckle, M. A. Semones, M. P. Doyle, M. N. Protopopova, J. Am. Chem.
Soc. 1992, 114, 1874-1876; (b) G. G. Cox, C. J. Moody, D. J. Austin, A. Padwa, Tetrahedron 1993, 49, 51095126; (c) A. Padwa, D. J. Austin, Angew. Chem. Int. Ed. 1994, 33, 1797-1815; (d) S. Miah, A. M. Z. Slawin, C.
J. Moody, S. M. Sheehan, J. P. Marino, Jr., M. A. Semones, A. Padwa, I. C. Richards, Tetrahedron 1996, 52,
2489-2514; (e) C. J. Moody, S. Miah, A. M. Z. Slawin, D. J. Mansfield, I. C. Richards, Tetrahedron 1998, 54,
9689-9700; (f) C. Y. Im, T. Okuyama, T. Sugimura, Chem. Lett. 2007, 36, 314-315.
40
(a) H. Nozaki, S. Moriuti, M. Yamabe, R. Noyori, Tetrahedron Lett. 1966, 7, 59-63. (b) H. Nozaki, S. Moriuti,
H. Takaya, R. Noyori, ibid., 7, 5239-5244. (c) H. Nozaki, H. Takaya, S. Moriuti, R. Noyori, Tetrahedron 1968,
24, 3655-3669. (d) R. G. Salomon, J. K. Kochi, J. Am. Chem. Soc. 1973, 95, 3300-3310.
41
Y. Hata, M. Watanabe, Tetrahedron Lett. 1972, 4659.
36
-29-
In this context, it has often been observed that the Rh(II) catalysts are less efficient relative to
copper salts to generate ammonium ylides.2a,3b This is probably due to a passivation by the
strongly Lewis basic nitrogen of the tertiary amines precursors (or the final product which is
also a tertiary amine) which coordinate tightly to the dirhodium moieties. In most cases, this
leads to coordinative saturation of the metal center and causes a loss of catalytic activity.
West and co-workers have for instance extensively studied the [1,2]-rearrangement of
ammonium ylides using both copper- and dirhodium-based catalysts. The authors in this
context
investigated
the
intramolecular
preparation
of
piperidines
and
tetrahydroisoquinolines.42 Their studies revealed that this transformation of 35 to product 36
can be achieved in better yields using Rh2(OAc)4 than corresponding copper source, as
catalyst (Table 2-1, entry 1). However, the substrates with longer side chain (n = 2,3) provide
mostly C-H activation product 37 with Rh2(OAc)4. This transformation was than achieved by
using copper based catalyst, which provide better yield with no C-H activation impurities
(entries 2,3). This particular result also highlighted the differences in mode of action of the
two catalysts.
Table 2-1 Metal sources affecting the outcome of the reaction
Entry
1
2
3
a
(n)
1
2
3
Substrate
Product
35a
35b
35c
36a
36b
36c
Yields of 36
Rha
99
35
0
Cub
69
61
58
3 mol% Rh2(OAc)4, DCM, rt; b 5 mol% Cu(acac)2, Toluene, reflux.
Interestingly, in contrast to copper catalysis, reactions with Rh2(OAc)4 provided also homocoupling products which supports a biradical mechanism. The authors have suggested that the
drastic differences in reactivity between the two catalytic systems is due to the greater
electrophilicity of copper carbenoids which may work in concert with a diminished propensity
for C-H insertion to favor the ylide formation.
42
F. G. West, B. N. Naidu, J. Am. Chem. Soc. 1993, 115, 1177-1178.
-30-
Chapter 2
In 1998, Moody et al published their studies on effect of ligands on chemoselectivity of
rhodium carbenoids, towards ylide formation and insertion reactions.39e They showed that
(Rh2(pfm)4)43 resulted in mainly the indole type products (39a and 39b) by insertion reaction
whereas with Rh2(OAc)4, mixture of different products were formed (Equation 2-4).
Equation 2-4
During these studies authors also tested copper (II) catalyst which selectively form the ylide
intermediate which subsequently resulted in benzo-fused heterocycles (38) by [2,3]rearrangement (Scheme 2-10).
Scheme 2-10
Padwa has also used metal catalyzed decomposition of diazo derivatives to create tandem
ammonium ylides and then [1,2]-rearrangement reactions to access cephalotaxine skeletons.44
Overall, with highly Lewis-basic nitrogen derivatives, better yields have been acquired with
copper salts than with dirhodium catalysts.45 In one of these cases shown below, it was
possible to isolate the ammonium ylide 40 which, when heated in absence of the catalyst,
provide [1,2]-rearranged product 41 in high yield (Scheme 2-11). This observation also
supports the formation of ammonium ylide as an intermediate but more importantly, it clearly
demonstrates that the rearrangement of ammonium ylide can be achieved selectively without
the participation of metal catalyst.
43
Pfm stands for perfluorbutyramide (NHCOC3F7).
A. Padwa, L. S. Beall, C. K. Eidell, K. J. Worsencroft, J. Org. Chem. 2001, 66, 2414-2421.
45
(a) F. G. West, B. N. Naidu, R. W. Tester, J. Org. Chem. 1994, 59, 6892-6894; (b) J. S. Clark, S. A. Krowiak,
L. J. Street, Tetrahedron Lett. 1993, 34, 4385-4388.
44
-31-
Scheme 2-11
Clark and Hodgson studied the [2,3]-sigmatropic rearrangement of ammonium ylides, formed
by decomposition of diazo-compounds using both copper and rhodium based catalysts.46
Interestingly, in this case also better yield were observed when Cu(acac)2 was employed as a
catalyst in place of as Rh2(OAc)4. (Equation 2-5).
Equation 2-5
This methodology was employed to efficient synthesis of azabicyclo[6.3.0]undecane, which
correspond to the CE ring system of alkaloid Manzamine A (Equation 2-6).47
Equation 2-6
2.5
Enantiospecific [1,2]-Stevens Rearrangements
As mentioned before in section 2.2, the mechanism of [1,2]-Stevens rearrangements is neither
unified nor trivial and despite all reported investigations, it still being the subject of
discussion. This lack of understanding associated to mechanistic difficulties is again reflected
46
47
J. S. Clark, P. B. Hodgson, J. Chem. Soc., Chem. Commun. 1994, 2701-2702.
J. S. Clark, P. B. Hodgson, Tetrahedron Lett. 1995, 36, 2519-2522.
-32-
Chapter 2
in the fact that a general asymmetric version of this reaction still remains to be developed.
Though several effective diastereoselective processes have been described, enantioselective
[1,2]-Stevens rearrangements of ammonium ylides using metallocarbenes or other means for
that matter is yet to be found.2,3 This was somewhat surprising considering the asymmetric
transformations already developed using metallocarbenes derived from diazo-decomposition
reactions. In this section, selected articles of the literature will be outlined together with
studies from our group on the topic.
2.5.1 Base promoted processes
In 1966, Hill et. al. demonstrated the first example of asymmetric Stevens rearrangement of
an enantioenriched salt (+)-(R)-42 as substrate. Its reaction with t-BuOK in DMSO gave
optically active (–)-(S)-43 in low yield (15%, Equation 2-7).48 This reaction was the first
example of transfer of asymmetry from a stereogenic quaternary nitrogen atom to an adjacent
carbon in [1,2]-Stevens rearrangements.
Equation 2-7 Enantiospecific [1,2]-Stevens rearrangement
Several subsequent efforts were detailed in literature.2a,49 In 1999, West and Glaeske50
reported the [1,2]-Stevens rearrangement of proline-derived salts 44. Compound 44 led under
basic conditions the formation of ylide 45, which rearranged by migration of the N-benzylic
substituent to the adjacent carbon affording the rearranged product 46 in a good yield (73%)
and a moderate enantiomeric excess 54% (Scheme 2-12).
48
R. K. Hill, T. H. Chan, J. Am. Chem. Soc. 1966, 88, 866-867.
J. H. Brewster, R. S. Jones, J. Org. Chem. 1969, 34, 354-358.
50
K. W. Glaeske, F. G. West, Org. Lett. 1999, 1, 31-33.
49
-33-
Scheme 2-12 Asymmetric [1,2]-Stevens rearrangement of a proline derived cation.
This transformation established clearly the feasibility of the approach but also its
shortcomings that is the loss of stereochemical information during the [1,2]-Stevens
rearrangement. The methodology provided nevertheless an interesting stereoselective route to
α-quaternary amino acid derivatives albeit with a moderate level of N-to-C chirality transfer.
More recently, Tayama et al revisited the process described above, using more hindered tbutyl ester ammonium salt (1S,2S)-47.51 This diastereomerically pure salt had provided, under
West’s conditions (t-BuOK, THF), the rearranged Stevens-product (R)-48 with 80% yield and
72% ee (Equation 2-8). Moreover it was shown that biphasic solid/liquid conditions (CsOH,
1,2-dichloroethane) afforded a better outcome with the production of (R)-47 in 73% yield and
92% ee.
Equation 2-8
Both West and Tayama suggested similar mechanism for this process. They proposed the
formation of the ylide by deprotonation of the α-stereogenic proton, as initial step. The ylide
then react through a homolytic fragmentation pathway to give rise to a biradical pair. A rapid
recombination in a solvent cage then afforded 48 preferentially in a stereo-retentive fashion.
The rate of this recombination and the rate of diffusion of the free radical to the opposite face
of the other depend on the substrate and the reaction conditions. Also, these results
demonstrated that the enantioselectivity depends upon the ionization conditions and to some
extent upon the migrating group.
51
E. Tayama, S. Nanbara, T. Nakai, Chem. Lett. 2006, 35, 478-479.
-34-
Chapter 2
Later, in a different approach, Somfai et al reported an asymmetric Lewis acid-mediated
[1,2]-Stevens-like rearrangement of N-benzyl glycinamides.52 The rearrangement proceeds by
complexation of the chiral Lewis acid to the amine 49 followed by deprotonation, which then
undergoes a rearrangement to give product 50 in 72% yield.
Equation 2-9
In 2007, our group reported the first example of enantioselective [1,2]-Stevens rearrangement
using enantiopure anionic counterions as asymmetric auxiliaries.53
Scheme 2-13
The methodology constituted an interesting example of double transmission of chirality: (i) a
supramolecular transfer of the helical chirality of anion BINPHAT 51 to the axial chirality of
cation 52 and then, (ii) its very effective translation (90 to 100%) during the [1,2]-Stevens
rearrangement to the centered chirality of the resulting amine (Scheme 2-13).
52
(a) J. Blid, O. Panknin, P. Somfai, J. Am. Chem. Soc. 2005, 127, 9352-9353; bJ. Blid, O. Panknin, P. Tuzina,
P. Somfai, J. Org. Chem. 2007, 72, 1294-1300
53
M.-H. Gonçalves-Farbos, L. Vial, J. Lacour, Chem. Commun. 2008, 829-831.
-35-
Though a moderate level of ee was obtained (eemax 55%), this example still represent an
interesting direction in the search for high control of selectivity in [1,2]-Stevens
rearrangement. Importantly, only a moderate loss of chirality (0-10%) was observed during
the translation step from axial to centered chirality.
2.5.2 Metallocarbenes promoted processes
After the pioneering work of Hata and Watanabe on [1,2]-rearrangement of the ammonium
ylides derived from diazo-decomposition reactions resulting in a ring expansion to provide
cyclic amine. In 1994, West and Naidu applied this approach in the efficient enantiospecific
total synthesis of epilupinine using Cu(acac)2 and Rh2(OAc)4 as catalysts.44a Proline derived
diazo-compound 53 was prepared, which afforded diastereomeric spiro-cyclic ylide
intermediates 54 and 55. These ylides then underwent a [1,2]-shift to provide the major
diastereomer 56 containing desired quinolizidine ring system in high yield (Scheme 2-14).
The authors also reported that better diastereo- and enantioselectivity were obtained when
Cu(acac)2 was used as catalyst in place of Rh2(OAc)4. In all cases, retention of configuration
was observed for the migrating group.
Scheme 2-14
West and Naidu postulated that resulting diastereoselectivity was due the selectivity in
formation of one of the diastereomeric ylides. This selectivity was due to the steric
interactions caused by the ester group. They also suggested that the slower-recombination of
radical-pair resulting in loss of chiral information Scheme 2-15.
-36-
Chapter 2
Scheme 2-15
Inspired from these results, Saba and co-workers,54 also utilized a [1,2]-Stevens
rearrangement/ring expansion approach to make swainsonine analogues using Cu(acac)2 and
Rh2(OAc)4 as catalyst (Table 2-2).
Table 2-2
Entry
1
2
3
4
Substrate
58a
58a
58b
58b
Catalyst
Rh2(OAc)4
Cu(acac)2
Rh2(OAc)4
Cu(acac)2
Yield (%)
84
90
85
90
dr
60:40
53:47
72:28
65:35
ee
80
68
84
90
Interestingly, slightly better diastereoselectivity was obtained with Rh2(OAc)4 than with
Cu(acac)2. Although transfer of chirality was highly dependent on ester side chain and was
better with Rh2(OAc)4 when methyl ester was used (entries 1 and 2) whereas for benzyl ester
better ee was obtained with Cu(acac)2 (entries 3 and 4).
The authors also reported that the ylide intermediate can be isolated by changing the reaction
conditions to reflux in DCM. The major diastereomer of the ylide intermediate was isolated
which, on heating in absence of catalyst, afforded alkaloid 58 as a single diastereomer with
95% ee in both cases (Scheme 2-16).
54
D. Muroni, A. Saba, N. Culeddu, Tetrahedron: Asymmetry 2004, 15, 2609-2614
-37-
Scheme 2-16
In summary, we have seen that base catalyzed [1,2]-Stevens rearrangements can generate
many different types of nitrogen containing compounds which have been used in mechanistic
studies and to make different biologically active compounds. This rearrangement involves the
formation of a nitrogen ylide which then undergoes a rearrangement. We have also seen that
these nitrogen ylides can be made directly, often more efficiently and selectively using the
reaction metallocarbenes with tertiary amines. Few examples were shown to explain the
dependence of the outcome of reaction on the nature of the metal source. These examples
have highlighted the difficulties in choosing one catalyst (copper or rhodium) over the other
providing better selectivity for the substrate. Also, isolation of metal-free nitrogen ylides
using this protocol and their thermal rearrangement in absence of metal catalyst suggested the
possibilities of metal-free rearrangement in general. These results again highlighted the
difficulties in making this process enantioselective as compare to other processes using
metallocarbenes where metal remained bounding in enantio-discrimination step.
This information will be important in this manuscript and in particular in the studies that will
be discussed in chapter 4.
-38-
Chapter 3
3. Two Step Synthesis of Ethano-Tröger Bases
3.1
Preamble
As mentioned in chapter 1, methano-Tröger bases undergo facile racemization under mild
acidic conditions which limits possible applications. One of the solutions to this problem is
the use of ethano-Tröger bases. In fact, the ethano-bridge provides a configurational stability
to the compounds of type 16, while keeping the V-shape geometry essentially intact.1 In view
of these (added) advantages, it was thus our intention to use 16 and its derivatives as surrogate
of methano-Tröger bases. Also, in light of the previous experience in the group on [1,2]Stevens rearrangements (see chapter 2),2 an approach using this transformation was
considered for making the compounds of type 16.
The results mentioned in this chapter were obtained with the cooperation of Dr. Christophe
Michon, a former post-doctorate fellow of the group.
Figure 3-1
To our initial surprise, a [1,2]-Stevens rearrangement of methano-Tröger derivatives had
never been reported in the literature, although several quaternary ammonium salt of titled
compound 1 had been described.1a,3 This provided an excellent opportunity to utilize the
rearrangement to prepare new scaffolds.
1
(a) D. A. Lenev, D. G. Golovanov, K. A. Lyssenko, R. G. Kostyanovsky, Tetrahedron: Asymmetry 2006, 17,
2191-2194; (b) Y. Hamada, S. Mukai, Tetrahedron: Asymmetry 1996, 7, 2671-2674.
2
M.-H. Gonçalves-Farbos, L. Vial, J. Lacour, Chem. Commun. 2008, 829-831.
3
(a) M. Häring, Helv. Chim. Acta 1963, 46, 2970-2982; (b) O. Trapp, G. Trapp, J. W. Kong, U. Hahn, F. Vögtle,
V. Schurig, Chem. Eur. J. 2002, 8, 3629-3634.
-39-
Two Step Synthesis of Ethano-Tröger Bases
3.2
Base Catalyzed Rearrangement
As already mentioned (chapter 2), over the last few years our group has shown that anionic
chiral counterions can impose, by Pfeiffer effect, a particular configuration to labile
quaternary chiral cations.4 In this context, we envisioned that a chiral anions may also induce
asymmetry on quaternary ammonium salts of methano-Tröger bases considering their
relatively low racemization barrier of 1 (G‡298 = 21.5 kcal.mol-1) The presence of a chiral
anion such as BINPHAT would indeed form diastereomeric salts which might then, in low
polarity solvent, undergo a epimerization in favor of possibly more stable diastereoisomer
(Scheme 3-1). This process coupled with a stereospecific Stevens rearrangement, would then
lead to the formation of ethano-Tröger base hopefully in a high enantiomeric excess.
Scheme 3-1 Initial Strategy: Induction of asymmetry using a chiral anion and coupled with a stereospecific
[1,2]-Stevens rearrangement.
For the initial experiments, two different hexafluorophosphate salts of TB-1 were made by
treatment of rac-1 in refluxing benzene with the corresponding alkyl bromide/iodide,5
followed by ion exchange metathesis of the halide ions (NH4PF6, MeOH) (Equation 3-1).
4
5
J. Lacour, D. Moraleda, Chem. Commun. 2009, 7073-7089.
R. G. Kostyanovsky, V. R. Kostyanovsky, G. K. Kadorkina, K. A. Lyssenko, Mendeleev Commun 2001, 1-5.
-40-
Chapter 3
Equation 3-1 Synthesis of quaternary ammonium salts of Tröger bases
With these salts of 59 and 60 in hand, further ion exchange was realized to make the
quaternary ammonium salts of (,S)-BINPHAT anion. Initial NMR analyses of these salts
were promising as the association with the chiral anion caused a large separation in the proton
signals of the diastereomeric salts (max 0.40 ppm in benzene, Figure 3-2). However, to our
disappointment, this NMR-signals splitting efficacy was not translated into a supramolecular
stereocontrol of the chiral cations (as a 1:1 ratio of signals was always observed). This ratio
remained same when the salts were heated at 60 °C.6 These results were not favorable but, we
nevertheless continued with the hypothesis of a possible asymmetric induction from the chiral
anion as upon [1,2]-rearrangement, selectivity might still occur due to this step of the process.
(a)
(b)
(c)
(d)
(e)
(f)
2.30
2.20
2.10
2.00
1.90
1.80
1.70
1.60
Figure 3-2 1H NMR spectra (400 MHz,  2.38-1.55) of (a) [rac-60][Br] in CD3CN. [rac-60][(S-51] in (b)
CD3CN, (c) acetone-d6, (d) CD2Cl2, (e) CDCl3 and (f) C6D6. Solvent signals are masked for clarity.
6
(a) S. Kirschner, N. Ahmad, C. Munir, R. J. Pollock, Pure Appl. Chem. 1979, 51, 913-923; (b) P. Pfeiffer, K.
Quehl, Chem. Ber. 1931, 64, 2667-2671.
-41-
To induce the desired rearrangement, salt 60 was first treated under basic conditions using
CHCl3 as solvent. However, as stated in previous reports,3a,7 only starting material and/or
demethylenated product 61 (up to 93% using NaOH as base, Scheme 3-2) were/was observed,
with no trace of any rearranged product. This exclusive formation of 61 was explained by a
preferred nucleophilic reactivity of the introduced bases adding to the bridge-methylene
carbon (Scheme 3-2, path a),3a,7 rather than acting as bases to form the intended ylide and
subsequent rearrangement (path b).
Scheme 3-2 Possible reactivity pathways of quaternary ammonium salts of 1 in the presence of bases.
To solve this issue and to direct the reaction towards path b, it was decided to introduce side
chains with more acidic -protons. A quick survey of pKa values in literature suggested that
introduction of more acidic ArCOCH2- side chains could be useful (pKa values for
PhCOCH2N+Me3 and PhCH2N+Me3 are 14.6 and 31.9 respectively). Salt 62 was then
prepared by treatment of Tröger base rac-1 with -bromoacetophenone at reflux in benzene
(87%, Equation 3-1). Again however, initial attempt using standard conditions for Stevens
rearrangements (NaOH, KOH, t-BuOK as bases) were disappointing and led to
dihydrodibenzodiazocine (diazocine) product 63 through probably path a (Scheme 3-2). Yet,
with other bases like CsOH, Cs2CO3, NaOMe, NaOAc, traces of a new product was seen in
1
H-NMR spectrum which was considered to be possibly the desired rearranged derivatives.
To increase further the selectivity towards path b, the use of organic bases was considered.
Different tertiary amines were used and an isolable amount of this novel species was achieved
7
F. C. Cooper, M. W. Partridge, J. Chem. Soc. 1957, 2888-2893.
-42-
Chapter 3
in 5-18% yield (Table 3-1, entry 1-4).The best result was afforded with hindered Hünig base
(i-Pr2NEt), although, in this case, diazocine product 63 was still the major component of the
reaction. These results indicate the requirement for a possibly less-nucleophlic base.
Table 3-1 Screening of bases for [1,2]-Stevens rearrangementa
Entry
Salt
Base
equiv
64:63c
Yieldb
1
62
DABCO
5.0
34:64
10%
2
62
Et3N
5.0
43:57
8%
3
62
Proton Sponge
5.0
27:73
5%
4
62
Hünig’s base
5.0
40:60
18%
5d
62
Al2O3 (pH=9.5 ±0.5)
20xe
94:6
45%
6d
62
Al2O3 (pH=9.5 ±0.5)
40xe
89:11
85%
a
[62][Br], Base, CHCl3, 10 hours, 25 °C . b Isolated yields. c Determined by 1H NMR (400 MHz) on the crude reaction
mixture. d reaction was finished in 2.5 hours, 25 °C. e Mass equivalents are used.
In this context, basic alumina (Brockman activity I, pH 9.5 ± 0.5) was considered. To our
delight, treatment of salt 62 with Al2O3 in chloroform resulted in novel scaffold 64 and upon
optimization in excellent yield 85% (Table 3-1, entries 5 and 6). Further, NMR spectroscopic
analysis confirmed that product 64 was formed by the migration of methylene-bridge rather
than the phenyl or benzyl group attached to the positive N-atom. Due to this migration the
ethano-bridge was constructed and a new stereogenic center was added  to the carbonyl. To
our delight, NMR spectroscopy of crude reaction mixture suggested the formation of a single
stereoisomer of compound 64 (dr > 98:2). A precise determination of the configuration of this
newly generated stereogenic center turned out to be difficult using NMR spectroscopy. This
was achieved through an X-ray diffraction analysis of a single crystal of rac-64. The preferred
relative configuration of this ethano-Tröger base is (5S,11S,14R)/(5R,11R,14S) for the
tricyclic moiety.
N11
N11
C14
N5
C14
N5
Figure 3-3 ORTEP view of the Ethano-TB 64 crystal structures; enantiomers (5SN,11SN,14RC) are shown.
Hydrogen atoms are omitted for clarity (except one at 14C) and thermal ellipsoids are drawn at 50% probability.
-43-
From this cliché, it was clear that the functionalized ethano-bridged TB retains the folded
geometry of classical methano-TB, the mean planes containing the tolyl rings are almost
perpendicular to each other 83° vs. ca. 90° for compounds 1.
At this stage, resolution of enantiomers of rac-64 was achieved in the collaboration with Dr.
Eric Francotte (Novartis). The two enantiomers were obtained on a preparative scale
chromatography resolution using cellulose-based phase Chiralcel OJ and a mixture of nheptane:ethanol 90:10, as eluent. From a batch of 190 mg of rac-64, two separated
enantiomers 53 mg (ee >99%, 28%) and 60 mg (ee >99%, 32%) were afforded corresponding
to the first and second eluted fractions respectively. These fractions corresponded to the (+)64 and (−)-64 enantiomers. The electronic circular dichroism (ECD) spectrum were recorded
and are displayed in Figure 3-4.
Figure 3-4 ECD spectra of (a) (+)-64 and (b) (−)-64
With these separate enantiomers of compounds 64 in hand, the question of their
configurational stability could be studied. To our delight, solutions of each enantiomers of 64
(MeOH or CHCl3) were stirred for 2 h at 25 or at 60 °C in the presence of a strong acid such
as (+)-camphorsulfonic acid (0.5 or 2 equiv.) with no visible loss of enantiomeric purity. Only
under more forcing conditions in DMF, at 100 °C for 2 h, in presence of 2 equiv. of
camphorsulfonic acid, a drop of enantiomeric purity was observed (from >99% to 88%).8 In
any case, these results clearly demonstrate the better configurational stability of these
compound 64 as compared to corresponding methano-Tröger bases. An extension of this
chemistry was then looked for.
8
Most probably this racemization of 64 involves an acid-mediated retro-Mannich type reaction through an
iminium/enolate intermediate. Please consider section4.5 for detailed discussion.
-44-
Chapter 3
To extend the scope of this novel rearrangement, substituents were introduced on the phenyl
chain; the salt being prepared according to the alkylation procedure already described
(Equation 3-1). Upon treatment with basic alumina in chloroform, these salts (62a to 62d)
afforded corresponding rearranged products (63a to 63d) with relatively lower yields (3465%; Table 3-2, entries 2 to 5), but excellent stereoselectivity was observed for the newly
generated stereogenic center (dr >98:2). Not too surprisingly, in the line of increased acidity
of the methylene protons, better yields were obtained with electron-withdrawing rather than
electron-donating substituents on the phenyl part of the side chain. NMR studies are
consistent with the same relative configuration of the corresponding compounds formed. Salt
[62e][Br] was prepared from rac-2,3,8,9-tetramethoxy-Tröger with -bromo-acetophenone,
also led to rearranged product (35% yield, Table 3-2); the lower outcome being probably due
to the formation of a diazocine side product of type 63.
Table 3-2 Rearrangement of ammoniums 64a to 64b, side-chain influencea
Entry
Salt
X,Y
R
Product
Yield
drb
1
62
Me,H
C6H4
64
85%
>98:2
2
62a
Me,H
C6H4-pMe
64a
51%
>98:2
3
62b
Me,H
C6H4-pOMe
64b
34%
>98:2
4
62c
Me,H
C6H4-pF
64c
65%
>98:2
5
62d
Me,H
C6H4-oBr
64d
40%
>98:2
7
62e
OMe,OMe
Ph
64e
35%
>98:2
a
Ammonium salt, Al2O3 (pH 9.5 ±0.5), CHCl3, 2 hours, 25 °C. b Determined by 1H NMR spectroscopy (400 MHz).
As mentioned above, this basic-Al2O3 induced condition provide excellent stereocontrol at the
newly generated stereogenic center, but often poor to moderate yields of ethano-Tröger bases
results, due to the formation of diazocine side-product 63 (Scheme 3-2, path a).
In an attempt to improve the situation, hexamine (hexamethylenetetramine) was considered as
a possible alternative of basic alumina, as it has a propensity to release (decompose into)
formaldehyde on quaternization (protonation) of one of the nitrogen atom. This can thus avoid
-45-
decomposition of ylide intermediate toward demethylenated diazocine side-product 63
(Scheme 3-2, path a). In fact, treatment of salt 62 with 2 equivalents of hexamine in
chloroform, results in its quantitative conversion without formation of diazocine impurity 63.
However, the stereocontrol was poor this time (dr 4:1) but this was seen as an advantage as,
for the first time the minor diastereomer was observed. Isolation of minor diastereomer turned
out to be tedious, as it converts into major diastereomer on silica and also spontaneously on
basic alumina. More detailed studies on this phenomenon will be covered in section-3.4. In
most of the cases, however, the isolation was not deemed necessary and the crude reaction
mixtures were directly treated with basic-alumina to afford diastereomerically pure
compounds of type 64.
In conclusion, a simple three step protocol was developed starting from rac-1 to make
functionalized ethano-Tröger bases, using an alkylation step, a rearrangement induced by
hexamine and then a final epimerization with basic alumina. This three step protocol provides
good overall yields of the major diastereomer (Table 3-3).
Table 3-3 Rearrangement of ammoniums 62a to 62g, via 2 step protocola
Entry
Salt
X,Y
R’
Yield
drb
1
62
Me,H
C6H4
91%
>98:2
2
62a
Me,H
C6H4-pMe
63%
>98:2
3
62b
Me,H
C6H4-pOMe
72%
>98:2
4
62c
Me,H
C6H4-pF
71%
>98:2
5
62d
Me,H
C6H4-oBr
54%
>98:2
6
62e
OMe,OMe
Ph
85%
>98:2
a
Ammonium salt, hexamine, CHCl3, 6 hours, 25 °C, then Al2O3 (pH 9.5 ±0.5), 0.5 hours, 25 °C.
spectroscopy (400 MHz).
-46-
b
Determined by 1H NMR
Chapter 3
3.3
Acid Catalyzed Rearrangement
As mentioned earlier, initial approach in this project considered the use of a chiral anion to
induce a stereoselective equilibration prior to the rearrangement. Our attention was thus
drawn on the related studies by Wilen et al, which claimed an asymmetric synthesis of Tröger
base 1 in the presence of BINOL-derived phosphoric acid (−)-5, as mentioned in chapter 1.9
Salt [62][Br] was then treated with acid (−)-5 at room temperature in chloroform.
Unfortunately, no change was observed in 1H-NMR spectrum at 25 °C. This solution was then
heated at 60 °C to force an enantiomerization of 1, but under these forcing conditions, the
substrate underwent a rearrangement to give the protonated-product 64-H+ instead in the
conjugated acid form with very high selectivity at the newly generated stereogenic center (dr
>98:2). The rearranged product 64 was recovered after basic workup. However the product
64 was racemic and no transfer of chirality from the chiral acid (−)-5 was observed (Equation
3-2). For comparison, heating the salt 62 in chloroform, in the absence of acid, resulted in
only starting material or some degradation products.
Equation 3-2 Acid catalyzed rearrangement
Table 3-4 Screening of chiral acid and chiral alcoholsa
Temp. (°C)
Yield (%)b
None
60
Np
2
(−)-5
60
95
>98:2
0
3
(+)-Camphorsulfonic acid
60
92
>98:2
0
4
(+)-L-Tartaric acid
60
95
>98:2
0
5
(-)-6
60
98
>98:2
0
6
(R,R)-TADOL
60
94
>98:2
0
Entry
Reagent
1
a
drb
eec
0
Ammonium salt (0.1 mmol), acid or alcohol (0.1 mmol), CHCl3, 6 hours, 60 °C. b NMR yields, determined by 1H NMR
(400 MHz) c Determined by CSP-HPLC analysis on a chiral stationary phase.
9
S. H. Wilen, J. Z. Qi, P. G. Williard, J. Org. Chem. 1991, 56, 485-487.
-47-
Encouraged by these results, different chiral acids were tested to hopefully induce an
asymmetry into the titled rearrangement (Table 3-4). Unfortunately, all our attempts results in
a complete lack of induction and the only racemic rearranged product was isolated (entries 25). It was later found that changing the conditions to MeOH at 60 °C, in absence of acid also
led to the synthesis of compound 64. A possible role for hydrogen bonding interactions was
considered but the use of enantiopure TADOL as chiral diol in the medium did not bring an
induction either (Table 3-4, entries 5-6).
In summary, at this stage, the two step alkylation/rearrangement protocol was established
which can be easily fine-tuned towards new functionalized ethano-bridged Tröger bases.
Configurationally stable derivatives are prepared as single diastereomers (de > 96%).
However, our attempts failed so far to provide enantioenriched ethano-Tröger bases using
either chiral anions, or with chiral acids and alcohols. To improve our understanding of this
rearrangement and to possibly develop this methodology further, we decided to perform some
control experiments.
3.4
“Control” Experiments
Two questions needed to be answered, in particular: (i) the origin of high diastereoselectivity
observed both in base- and acid-catalyzed reactions and (ii) the mechanism of this novel
rearrangement (ion-pair, bi-radical or other).
A brief solvent screening was performed for the rearrangement promoted by basic alumina
(Table 3-5). As evident in Table 3-5, changing to polar protic solvents like methanol or
ethanol led to the formation of significant amount of the usual minor diastereomer epi-64
(entries 4 and 5). On the other hand, in the case of polar aprotic solvents like DMF, DMSO
and dioxane, decomposition of starting material was observed to give diazocine 63 as the
main component (entries 6 to 9). All in all, the nature of the solvent is crucial both in terms of
yield and selectivity.
-48-
Chapter 3
Table 3-5 Solvent screening for rearrangement in basic conditions
Entry
Solvent
yieldb
drb
1
CHCl3
95
>98:2
2
CH2Cl2
93
>20:1
3
CH3CN
87
>25:1
4
MeOH
95
3:1
5
EtOH
90
3:1
6
DMSO
18
ndc
7
DMF
26
ndc
8
Dioxane
29
ndc
9
CD3OD
94
1.5:1
a
Ammonium salt, Al2O3 (pH 9.5 ±0.5), solvent, 2 hours, 25 °C. b determined by 1H NMR spectroscopy (400 MHz).c not
determined.
Interestingly, when CD3OD was used as solvent, a lower diastereomeric ratio (1.5:1) of
ethano-Tröger base was observed as compared to the corresponding reaction in MeOH (Table
3-5, entries 4 and 9). We will come back to this point later in this chapter.
Equation 3-3 Effect of deuterated solvent on diastereoselectivity of rearrangement
These results indicated also that deuterium was completely incorporated in 64-d and epi-64-d.
In our view this deuteration occurs by acid-base reaction on either the starting salt prior to the
rearrangement or on the minor and major diastereomer i.e. the products themselves (Equation
3-3). Cooling the reaction mixture in CD3OD at −20 °C resulted in an incorporation of two
deuterium atoms at the position to carbonyl group of the starting material to provide 62-d2,
-49-
and this without any trace of 64 and epi-64. This deuterated starting material 62-d2 provided
the same diastereomeric ratio (1.5:1) upon treatment with basic alumina in CD3OD.
Finally the possibility of deuterating directly 64 or epi-64 in CD3OD was tested under the
regular conditions (Equation 3-4). Interestingly, no deuterium incorporation into the major
diastereomer was observed 64.
Equation 3-4 Deuteration of final product in basic conditions.
In case of the minor diastereomer epi-64, it was converted into the major diastereomer 64
with only a partial deuteration at the -position of the later (Equation 3-5). This deuteration
experiment turn out to be capricious; percentage of deuteration varying from 50-80%.
Equation 3-5
These results suggest that both diastereomers are most probably formed in the course of
reaction. However, conversion of the less-thermodynamically stable epi-64 into the more
stable 64 appears to be highly dependent upon the nature of solvent. It is obviously fast in
CHCl3 and slow in MeOH (Scheme 3-3).
Scheme 3-3
-50-
Chapter 3
In MeOH, the acid-base reaction that transforms epi-64 into 64 is most probably ratedetermining. As such, when CD3OD is used and deuterium is fully incorporated in epi-64, a
primary kinetic isotopic effect occurs that slow down further the epimerization at the end of
the reaction. This leads to a higher percentage of epi-64.
3.5
Functionalization of Ethano-Tröger Bases
As shown above, we were able to synthesize ethano-Tröger bases of type 64 with very high
diastereoselectivity (dr 49 > 1), using a simple alkylation/rearrangement sequence. With these
compounds in hand, it was interesting to investigate further functionalization. First, attempts
were made to prepare vicinal aminols by simple reduction of carbonyl group. This reduction
will generate a new stereogenic center at carbonyl carbon; thereby could result in two
diastereomers. To our delight, when compound 64 was treated with NaBH4 in MeOH at 0 °C,
a single amino-alcohol 65 was obtained in excellent yield and diastereoselectivity (97%, dr >
49:1, Equation 3-6).
Equation 3-6
This excellent level of diastereocontrol for the addition of the small hydride nucleophile can
be explained by classical Felkin-Anh trajectory, the crystal structure of 64 being an excellent
structural model of the possible transition state (Figure 3-5). As it can be seen, the carbonyl
group points towards the medium size methylene group of the adjacent stereo center and of
even more importance, the bulky nitrogen substituent, with its low energy * C-N orbital, is
perfectly aligned for a antiperiplanar attack of the nucleophile. All combine for the excellent
diastereoselectivity. Thus the configuration could be assigned to the product 65 as predicted
by Felkin-Anh analysis i.e. R-(5S,11S,14R)/S-(5R,11R,14S).
-51-
Nuc
Figure 3-5 Felkin-Anh trajectory for nucleophilic addition on 64
Interestingly, the obtained structural motif present analogous to the well -known cinchona
alkaloid family containing a rigid tricyclic core containing a bridge head N-atom together
with a stereogenic center carrying a secondary benzylic alcohol (Figure 3-6).
Figure 3-6
In an attempt to generate a large number of surrogates, and confirm the excellent stereocontrol
of the hydride addition, the addition reaction of various other nucleophiles to this compound
64 was investigated.
To this end, various lithiated alkyl and aryl nucleophiles were then added to 64 at −78 °C. In
all cases, the corresponding tertiary alcohols 65a to 65d were isolated in good to excellent
yield (89-91%) and excellent diastereoselectivity (dr > 49:1) (Equation 3-7). In this case, due
to the presence of the quaternary centers (less favorable conformational situation) it is much
more difficult to assign the relative configuration after the attack. However, as the
stereoselectivity remains excellent, it is unlikely that a change of the stereochemical
preference would have been observed during the hydride addition.
-52-
Chapter 3
Equation 3-7
The alcohol 65 was further functionalized by treatment with acetic anhydride in presence of
catalytic amount of DMAP. The corresponding acetate was isolated in 51% yield (Equation
3-8): This low yield was due to the highly crowded environment around the hydroxy group of
65. In fact, no conversion was observed reactions with sterically more demanding TsCl, Tf2O,
Ph2PCl.
Equation 3-8
However, interestingly, this alcohol was also smoothly converted using PCl5 into the
corresponding chloride 67 (73% isolated yield, Equation 3-9). Again, this reaction took place
with complete stereocontrol on the reactive carbon center leading to formation of only one
product 67.
Equation 3-9
The determination of the configuration of this newly generated stereogenic center was
difficult using NMR spectroscopy. Nevertheless, was achieved through an X-ray diffraction
analysis of a single crystal of rac-67 (Figure 3-7). The preferred relative configuration of this
-53-
ethano-Tröger base is S-(5S,11S,14R)/R-(5R,11R,14S). These results suggested the anti-attack
of the chloride ion relative to the leaving hydroxyl group.
X-ray
Nuc
Figure 3-7 ORTEP view of the Ethano-TB 67 crystal structures; enantiomers S-(5SN,11SN,14RC) are shown.
Hydrogen atoms are omitted for clarity (except two hydrogen atoms present on stereogenic centers) and thermal
ellipsoids are drawn at 50% probability.
Compound 67 was further functionalized by treatment with silver benzoate in either
acetonitrile or MeOH to give corresponding product of benzoate addition and etherification
with once again, complete stereocontrol on configuration of the carbon stereogenic center
(Equation 3-10).
Equation 3-10
To check, if 67 could be converted back to alcohol 65, the chloride compound 67 was then
treated with AgBF4 in CH2Cl2 containing 10 equivalents of water. The reaction proceeded
well to give yet another new compound 70 in 65% isolated yield (Equation 3-11). This
compound rac-70 was later characterized as the diastereomer of alcohol 65. In other words,
rac-(S-(5S,11S,14R)-70 was obtained from rac-(S-(5S,11S,14R)-67.10 This suggested retention
of configuration at the benzylic carbon atom. This compound rac-(S-(5S,11S,14R)-70 was
10
Configuration of 65 was assigned as R-(5S,11S,14R)/S-(5R,11R,14S)-65 using Felkin-Anh analysis for
reduction of 64 using NaBH4 in MeOH. And hence the configuration of diastereomer 70 could be assigned as S(5S,11S,14R)/R-(5R,11R,14S)-70.
-54-
Chapter 3
then reacted with PCl5, results in formation of product (S-(5S,11S,14R)-67 exclusively
(Equation 3-11).
Equation 3-11
Thus, clearly both diastereomers of the benzylic alcohol rac-(R-(5S,11S,14R)-65 and rac-(S(5S,11S,14R)-70 provided the same diastereomerically pure rac-(S-(5S,11S,14R)-67.
Together with the previous results, all the reactions of 67 and derivatives happen most likely
through an intermediate carbenium ion of which a single diastereotropic face is subjected to
nucleophilic attack leading to product of a single relative configuration (S-(5S,11S,14R)/ R(5R,11R,14S) at the benzylic center. It strongly resembles the results of recent studies from
the laboratory of the Thorsten Bach11 and Giorgio Cozzi12 who have been able to demonstrate
the existence of highly stereoselective SN1-like reactions. Our group has also recently
reported selective addition of hydride and organolithium reagents on the unsymmetrical
cationic [4]helicenes and the two diastereotropic faces were discriminated with high
efficiency (dr up to and higher than 49:1).13
Scheme 3-4
In our case here and also in the case of Cozzi, the presence of stereogenic centers  to the
carbenium ion fully controls the stereochemical future of the nucleophilic attack. By analogy
with the Felkin-Anh transition state, it is most likely that the hydrogen atom of the secondary
carbenium ion points towards the medium-sized methylene bridge or has the large aromatic
11
F.Mühlthau, O. Schuster and T. Bach, J. Am. Chem. Soc. 2005, 127, 9348-9349.
P. G. Cozzi and F. Benfatti, Angew. Chem., Int. Ed. 2010, 49, 256-259.
13
J. Guin, C. Besnard, P. Pattison, J. Lacour, Chem Sci 2011, 2, 425-428.
12
-55-
substituent of the cation is closer to the H-atom of the -stereogenic center. This is indicated
in Scheme 3-4. The nucleophilic attack occurs anti to the bulky and electron-rich nitrogen –
containing fragment to lead to the observed selectivity.
3.6
Enantiospecificity/Chirality Transfer
So far, we have shown that treatment of quaternary ammonium ions of Tröger bases, under
basic conditions may afford configurationally stable ethano-bridged Tröger derivatives and
this with very high to good selectively. These derivatives can themselves be further
functionalized with high selectivity. Yet, all the transformations were performed on racemic
material. Care was then taken to try to validate the approach starting from enantiopure Tröger
base, as a enantiospecific transformation would have an obvious interest in this field of
chemistry.
The quaternary ammonium salt 62 of enantiopure methano-Tröger base ( )-(R,R)-1 was then
prepared and it was treated immediately with basic alumina. To our disappointment, using the
usual conditions only rac-64 was observed (Equation 3-12).
Equation 3-12
In retrospect, the outcome was perfectly logical in light of the intermediacy of cation 62. In
fact, as mentioned in chapter 1, quaternary ammonium salts methano-Tröger bases of type 1
possess rather low racemization barrier (90.2 kJ mol-1).14 One can than extrapolate that, at 80
°C, the half-life for the racemization is only 2.1 second and hence provide racemic salt
(Scheme 3-5). As the alkylation of 1 by PhCOCH2Br to yield 62 takes a minimum of 12 hours
at that temperature, it was logical to obtain the racemic material.
14
O. Trapp, G. Trapp, J. W. Kong, U. Hahn, F. Vögtle, V. Schurig, Chem. Eur. J. 2002, 8, 3629-3634.
-56-
Chapter 3
Scheme 3-5
Attempts were made at intercepting the immediately quaternary ammonium ion after its
formation achieving by adding basic alumina to the reaction of (R,R)-1 and PhCOCH2Br in
benzene. However, this tandem alkylation/rearrangement protocol in one pot led again to
racemic product only in a poor yield due to the predominant formation of diazocine 63
(Equation 3-13).
Equation 3-13
In search for a general solution to this problem, we realized that actually only the ylide
intermediate 71 was necessarily required for the reaction to occur and not the labile
quaternary ammonium ion precursor (Scheme 3-6).
Scheme 3-6
Therefore, any reaction affording directly the ylide intermediate from Tröger base 1 being
then likely to yield an enantioenriched product. Since, as shown in chapter 2, nitrogen ylides
can be made directly through the reaction of tertiary amines with metallocarbenes, we decided
to use this more direct approach and develop hopefully efficient enantiospecific
transformation.
-57-
Chapter 4
4. One-Step Synthesis using Metallocarbenes
4.1
General Considerations
As shown in the previous chapter, the two-step methodology provided ethano-Tröger bases
with excellent stereocontrol on newly generated tertiary stereogenic centers. However, it
failed to provide an enantiospecific transformation i.e. an enantio-enriched ethano-Tröger
base from an enantiopure methano-Tröger base precursor. These studies also suggested that
the intermediacy of the quaternary ammonium ion of 1 was the reason for racemization. To
overcome this shortcoming, other means to generate the ammonium ylide intermediate
directly were then considered. As mentioned in chapter 2, this can be achieved in principle
through metal-catalyzed decomposition of diazo-compounds. This is a powerful method for
the generation of electrophilic metal carbenes, which are known to react with tertiary amines
to form the desired nitrogen–ylide intermediates, as shown in Scheme 3-7.
Scheme 3-7
Such transformations have also been reported in literature on enantiopure tertiary amines.
However, when the migrating carbon atom or the quaternized nitrogen atom are the only
stereocenters present on the substrate, important losses in enantiomeric purity are normally
observed as a result of the mechanism involved. Hence, this was not a straight forward
solution, and in fact, it offers another challenge to the problem in hand and an opportunity to
develop a methodology which could provide excellent chirality transfer on Tröger base.
Different acceptor-acceptor or donor-acceptor types of diazo-compounds were synthesized
according to the procedures described below and tested as reactants for the synthesis of
ethano-Tröger bases.
-59-
4.2
Synthesis of Diazo-Compounds
All diazo-compounds were easily synthesized using the classical diazo-transfer procedure
from p-acetamidobenzenesulfonyl azide (p-ABSA) in acetonitrile. Two different reaction
conditions were generally utilized.1 Acceptor-acceptor type of diazo-compounds were
prepared by treatment of the corresponding diketones, diesters and ketoesters, with ABSA in
presence of 2 equivalents of triethylamine (Equation 3-14). The product was purified by
column chromatography resulting in corresponding light yellow diazo-compounds in 85-95%
yields. These types of diazo-compounds are particularly easy to prepare and handle thanks to
their relatively high chemical stability promoted by the acceptor substituents. No degradation
was observed after several months upon storage at – 20 °C.
Equation 3-14
On the other hand, -phenylketones or -phenylesters required 1.5 equivalents of the stronger
DBU as a base to afford donor-acceptor types of diazo-compounds (Equation 3-15). These
reactions were usually carried out in an inert dinitrogen atmosphere to provide corresponding
yellow diazo-compounds in 60-90% yield, after purification on flash column chromatography.
These compounds are more reactive, relatively unstable and were stored at – 20 °C flushed
with argon. Special care was taken while handling these compounds.
Equation 3-15
1
J. S. Baum, D. A. Shook, H. M. L. Davies, H. D. Smith, Synth. Commun. 1987, 17, 1709 – 1716.
-60-
Chapter 4
4.3
Rhodium-catalyzed Synthesis of Ethano-Tröger Bases
4.3.1
Initial discovery
To begin the studies, dirhodium catalysts were selected together with symmetric acceptoracceptor types of diazo-compounds (-diazo--diketone and -diazo--diester) to avoid
diastereomeric mixtures. Diazo-compounds 72a and 72b were selected to check the feasibility
of this transformation (Table 3-6).
Table 3-6 Initial screening to achieve desired producta
Product
Yieldb
25
73a
npc
Toluene
25
73b
Np
Me
Toluene
60
73a
Np
72b
OMe
Toluene
60
73b
Np
5
72a
Me
CH2Cl2
25
73a
Np
6
72b
OMe
CH2Cl2
25
73b
Np
7
72a
Me
CH2Cl2
60
73a
ndd
8
72b
OMe
CH2Cl2
60
73b
Nd
9
72a
Me
Toluene
100
73a
Np
10
72b
OMe
Toluene
100
73b
87
11
72c
Ph
Toluene
100
73c
Np
12
72d
OEt
Toluene
100
73d
85
Entry
Diazo
R
Solvent
Temp. (°C)
1
72a
Me
Toluene
2
72b
OMe
3
72a
4
a
Typical reaction conditions: rac-1 (0.4 mmol), diazo-compound (0.8 mmol), [Rh2(OAc)4] (1 mol%), Solvent (1 mL), T °C,
16 h; reported results are the average of at least two experiments. b Yield of isolated product. c np stands for no product
formation. d nd stands for not determined.
Treatment of rac-1 with 72a and 72b at room temperature in toluene did not yield any of the
desired products. Only a violet precipitate was observed which is probably the coordination
-61-
complex of methano-Tröger base with Rh2(OAc)4 (entries 1 and 2). Heating the reaction
mixtures to 60 °C provided similar result in toluene (entries 3 and 4).
To check whether the absence of reactivity was due to the insolubility of the catalyst, a more
polar CH2Cl2 was used as solvent. Changing solvent at room temperature was futile (entries 5
and 6) as no product was formed. However, at 60 °C, traces of some products were identified
in ESI mass spectroscopy but the reaction mostly led to decomposition to diazocine of type 63
(entries 7 and 8). To further drive the reaction towards ethano-Tröger base derivatives, the
reaction mixture was then heated to 100 °C using toluene as a solvent. While the use of 72a
was unproductive (entry 9), 72b afforded the corresponding ethano-bridged Tröger base in
excellent yield (87%, entry 10). This result was particularly important as, it proved our initial
hypothesis that ethano-Tröger bases could be indeed prepared using metallocarbenes. Other
symmetric diazo-compounds were immediately tested under these conditions. As seen before,
no product was observed with -diazo--diketone 72c though a good reactivity was observed
with -diazo--diester 72d (85%, entries 11 and 12).
4.3.2
Substrate Screening
With these initially established conditions, we decided to increase somewhat the complexity
of the process by using non-symmetrically substituted diazo-derivatives. In these cases,
presence of two different groups on either sides of diazo-compound would introduce a new
stereogenic center at the constructed ethano-bridge, and hence results in diastereomers.
Further, other classes of acceptor-acceptor types of diazo-compounds were used. A series of
-diazo -ketone-esters reagents (72e to 72l) were chosen varying the steric environment on
both ester and ketone side chains (Table 3-7). These compounds were then prepared
according the procedure described above in section 4.2.
Initially, less hindered -diazo--ketone-esters 72e and 72f were tested (Table 3-7). These
substrates afforded the corresponding ethano-Tröger base in excellent yield (R1=Me,
R2=CO2Me and CO2Et, 78 and 80% respectively), though poor diastereoselectivity was
observed for the newly generated stereogenic center (2.4:1 and 2.1:1 respectively, entries 1
and 2). We envisaged that increased steric demand on the diazo-compounds might increase
the diastereoselectivity. Consequently, diazo-compounds with more sterically demanding
groups were tested. On introducing bulky substituents on the ester side-chain (R1=Me,
R2=CO2iPr and CO2tBu; 72g and 72h respectively), dramatic loss in yield was observed (73g,
-62-
Chapter 4
21% and 73h, 25%) and poor diastereoselectivity was still obtained (73g, dr 1.2:1 and 73h, dr
1.1:1). A similar trend was observed with diazo-compounds containing more sterically
demanding substituents (R1=Et, nPr and Ph R2=CO2Et; 72i, 72j, and 72k being generated in
similar poor isolated yield (73i, 17%; 73j, 25% and 73k, 45%), although this time a slight
increase in diastereoselectivity was observed (Table 3-7, entries 5 to 8). For diazo-compound
72l (R1=iPr), product was not even formed (entry 8).
Table 3-7 Screening of acceptor-acceptor type of diazo-compoundsa
Entry
Diazo
R1
R2
Product
Yieldb
drc
1
72e
Me
CO2Me
73e
78
2.4:1
2
72f
Me
CO2Et
73f
80
2:1
3
72g
Me
CO2iPr
73g
21
1.2:1
4
72h
Me
CO2tBu
73h
25
1.1:1
5
72i
Et
CO2Et
73i
17
2.4:1
6
72j
nPr
CO2Et
73j
25
2.7:1
7
72k
Ph
CO2Et
73k
45
3.3
8
72l
iPr
CO2Et
73l
npd
-
a
Typical reaction conditions: rac-1 (0.4 mmol), diazo-compound (0.8 mmol), [Rh2(OAc)4] (1 mol%), Toluene (1 mL), 100
°C, 16 h; reported results are the average of at least two experiments. Substituents indicated at positions R 1 and R2 correspond
to that of the major diastereomer of 73. b Yield of isolated product (both diastereomers). c Determined by 1H NMR analysis
(400 MHz) of the crude reaction mixtures. d np stands for no product formation.
After this first screening, we decided to address this problem of low productivity and
diastereoselectivity by using more reactive donor-acceptor type of diazo-compounds (
Table 3-8) generated from arylketones and arylesters.
-63-
Table 3-8 Screening of donor-acceptor type of diazo-compoundsa
Entry
Diazo
R1
R2
Product
Yieldb
drc
1
72m
Me
Ph
73m
70
5:1
2
72n
OEt
Ph
73n
50
>49:1
3
72o
OMe
Ph
73o
75
10:1
4
72p
OEt
H
73p
Nd
Nd
5
72q
Ph
H
64
25
Nd
a
Typical reaction conditions: rac-1 (0.4 mmol), [Rh2(OAc)4] (1 mol%), Toluene (1 mL), 100 °C, 16 h; reported results are
the average of at least two experiments. Substituents indicated at positions R1 and R2 correspond to that of the major
diastereomer of 73. b Yield of isolated product (both diastereomers). c Determined by 1H NMR analysis (400 MHz) of the
crude reaction mixtures. d np stands for no product formation.
This was a rewarding choice. Indeed, using diazo-compound 72m (R1=Me, R2=Ph), the
corresponding ethano-Tröger base 73m was obtained in good yield and higher
diastereoselectivity (70%, dr 5:1, entry 1). Importantly, switching to -diazo -phenyl-ester
72n (R1=OEt, R2=Ph), the corresponding ethano-Tröger base 73n was afforded as essentially
one diastereomer in 50% yield (entry 2). In the same class of donor-acceptor type of diazocompounds, methyl ester 72o (R1=OMe, R2=Ph) afforded excellent yield of major
diastereomer of 73o (71%) with good diastereoselectivity (dr 10:1, entry 3).
These results indicated that the higher reactivity of the diazo-compounds (derived from
donor-acceptor series) and the better steric difference among the substituents2 on the diazocompounds were beneficial factor in achieving and better diastereoselectivity and yield for
compounds of type 73. Thus, we tested even more reactive diazo-compound 72p (R1=OEt,
R2=H), though a complex mixture of compounds was formed giving little clue on what had
happened (entry 4). We also tried diazo-compound 72q (R1=Ph, R2=H, entry 5) which could
have provided 64 again. However, another complex mixture of products was then formed
2
The stereic influence of aryl and ester substituents can be analyzed using the A values derived from the
cyclohexane conformational analysis. The equitorial preference of a phenyl and a CO2Me substituent being 3.0
and 1.7 kcalmol-1, respectively.
-64-
Chapter 4
which make it impossible to determine diastereoselectivity. This mixture then treated with
basic alumina in chloroform resulting in formation of compound 64 as a single diastereomer
in 25% isolated yield.
For some of these derivatives, mono-crystals for the major diastereomers were obtained in
racemic form either after separation from the minor diastereomers or directly from crude
reaction mixtures. The relative configuration of these adducts 73e, 73g and 73o was
established by X-ray diffraction analysis (Figure 3-8). In all cases,3 at the new carbon
stereogenic centers, large and small substituents replace preferentially the pro-(S) and pro-(R)
hydrogen atoms of the parent (S,S)-16, respectively (and vice versa for (R,R)-16) as shown in
Figure 3-8.
N11
Pro-S
C14
N5
Pro-R
Figure 3-8 ORTEP views of the major diastereomers of racemic 72e,72g and 72m crystal structures;
enantiomers (5RN,11SN,14SC) are shown. Hydrogen atoms are omitted for clarity and thermal ellipsoids are
drawn at 50% probability.
All data indicate that this relative configuration is conserved for the major diastereomers
within the series of products 73e to 73o (Tables 4.2 and 4.3, 5SN,11RN,14RC or
5RN,11SN,14SC).
4.3.3
Enantiospecificity / Chirality transfer
With these results in hand, the enantiospecificity of this process was evaluated using
enantiopure (–)-(R,R)-1 as substrate. For initial experiments, diazo-compounds were selected
so as to cover each class of acceptor-acceptor (72d and 72f) and donor-acceptor (72m,72n
and 72o) type of diazo-compounds.
3
X-rays structures of other ethano-Tröger bases will be shown later in this chapter.
-65-
First, 72d and 72f gave unsatisfactory results, as corresponding products 73d and 73f were
afforded in low enantiomeric purity (ee 34% and 5% respectively, Table 3-9, entries 1 and 2).
Table 3-9 Initial screening for enantiospecificitya
Entry
Diazo
R2
R1
Product
Yield b
dr c
ee d
1
72d
CO2Et
OEt
73d
85%
-
34%
2
72f
CO2Et
Me
73f
55%
2:1
5%
3
72m
Ph
Me
73m
60%
5:1
10%
4
72n
Ph
OEt
73n
50%
>49:1
93%
5
72o
Ph
OMe
73o
71%
10:1
99%
a
Typical reaction conditions: rac-1 (0.1 mmol), [Rh2(OAc)4] (1 mol%), Toluene (0.25 mL), 100 °C, 16 h; reported results
are the average of at least two experiments. Substituents indicated at positions R 1 and R2 correspond to that of the major
diastereomer of 73. b Yield of isolated major diastereomers. c Determined by 1H NMR analysis (400 MHz) of the crude
reaction mixtures.
Further, diazo-compounds from donor-acceptor series were tested. Initial test with 72m was
disappointing as the corresponding product 72m was obtained in 10% ee only (entry 3).
However, to our delight, reactions with 72n afforded the products 73n with a first good
enantioselectivity (ee 93%, entry 4). Even, better results were obtained with diazo-compound
72o as essentially enantiopure compound 73o (ee 99%, entry 5) was resulted from the
reaction, the compound being levorotatory (−)-73o enantiomer. As expected, the reaction with
(+)-(S,S)-1 also gave the antipodal product (+)-73o with the same ee value and with exactly
the same yield and diastereoselectivity as in the racemic series. After these clear evidence of
enantiospecificity and with these enantiomers in hand, the electronic circular dichroism
(ECD) spectras were recorded and are displayed in Figure 3-9.
-66-
Chapter 4
Figure 3-9 ECD spectra of (a) + 73o (red) and (b) (−)-73o (blue).
The higher enantiomeric purity of 73o and the fact that its major diastereomer was obtained in
quite better yield than that of 73n (71% vs 50%) led us to use methyl rather than ethyl ester
diazo derivatives for the remainder of the study.
To detect a possible influence of electronic and steric parameters, various -diazo-methylester compounds containing different substituents on the aryl part (72oa to 72ol) were
prepared according to the procedure described in section 4.2. The results are summarized in
Table 3-10.
Clearly, both electron-donating (entries 1 to 4) and electron-withdrawing (entries 8 to 11)
substituents are amenable on the aryl group of the diazo-reagents and excellent transfer of
chirality was observed (except for entry 6, this will be discussed shortly). In all cases, the
corresponding products were obtained with good isolated yields (70-80%). The only
parameter which is fluctuating was the diastereoselectivity. The ratios varied from moderate
(4:1, entry 12) to high (20:1, entry 6).
In the case of ortho substituted diazo-compounds 72oe and 72og, complex mixtures of
products were formed (entries 5 and 7). These results demonstrate again the sensitivity of the
reaction on the steric-environment. Also, a slight decrease in enantiomeric purity was
observed while going from products with donor substituents (entries 1 to 4; OMe, Me) to
electron-deficient groups (entries 8-11; Cl, Br, CF3, F). This effect being particularly
noticeable in the reaction of (–)-(R,R)-1 and 72of in which, (–)-73of was isolated in only 64%
enantiomeric purity (entry 6). This result although disappointing, have importance for the
determination of the mechanism of the process which will be discussed later in this chapter.
-67-
Table 3-10
Entry
1 (X)
Diazo
R1
Product
Yield b
dr c
ee d
1
Me
72oa
p-MeOC6H4
73oa
83
10:1
98%
2
Me
72ob
2-Naphthyl
73ob
80
9:1
98%
3
Me
72oc
p-MeC6H4
73oc
72
-e
98%
4
Me
72od
m-MeC6H4
73od
70
8.0:1
98%
5
Me
72oe
o-MeC6H4
73oe
nd
nd
Nd
6
Me
72of
p-NO2C6H4
73of
82
20:1
64%
7
Me
72og
o-ClC6H4
73og
nd
nd
Nd
8
Me
72oh
p-ClC6H4
73oh
78
7.5:1
97%
9
Me
72oi
m-ClC6H4
73oi
75
6:1
97%
10
Me
72oj
p-BrC6H4
73oj
75
8:1
97%
11
Me
72ok
p-CF3C6H4
73ok
70
12:1
97%
12
Me
72ol
p-FC6H4
73ol
50
4:1
96%
a
Conditions: (–)-(R,R)-1 0.1 mmol, 1 mol % Rh2(OAc)4, 0.25 mL of dry toluene, 100 °C, 6 h. Reported results are the
average of at least two experiments b Isolated yield (%) of the major diastereomer. c dr ratios determined by 1H-NMR
spectroscopy (400 MHz) on the crude reactions mixtures. d ee values determined by CSP-HPLC. e not determinable. f ee 98%.
Furthermore, as the field of Tröger bases has been crumbled with mistakes concerning the
absolute configuration of the derivatives (see chapter 1), we decided to assign the absolute
configuration of major diastereomer of 73o with certainty. In view of the rigidity (only 2
different possible conformers) of conformation within 73o and the precise knowledge of its
structure though XRD (Figure 3-11), we decided to use a VCD (Vibrational circular
dichroism) approach.
These measurements and calculations were performed in collaboration of Dr. J.-V. Naubron
(University of Aix-Marseille). Infrared absorption and VCD spectra were measured for
solutions (CCl4) of the major diastereomer of both (+)-73o and (−)-73o and compared to the
averaged spectrum calculated for (5SN,11RN,14RC)-73o (Figure 3-10).
-68-
Chapter 4
Figure 3-10 Experimental IR absorption (top) and VCD (bottom) spectra (CCl 4, 298 K) of (−)-73o (red) and (+)73o (blue). Calculated spectrum of (5SN,11RN,14RC)-72o (green).
As briefly mentioned, Conformational analysis showed only two possible conformations for
the ester moiety the molecule being totally devoid of liberty, the most stable conformer
having essentially the geometry determined by X-ray crystallographic analysis (Figure 3-11).
The geometry optimizations, vibrational frequencies, IR absorption and VCD intensities were
calculated by using density functional theory (DFT).4 Overall, a good agreement between the
experimental and theoretical spectra was observed, thus allowing the assignment of a
(5SN,11RN,14RC) configuration for (+)-73o.
Later, this absolute configuration assignment was confirmed through a Flack parameter
analysis of the X-ray structure of compound (−)-73oh containing a chlorine heavy atom as
(−)- (5RN,11SN,14SC) 73oh as shown in Figure 3-11.
4
B3LYP and CAM-B3LYP functionals combined with 6-311+ G(d,p) and/or TZVP basis set were used.
Frequencies were scaled by a factor of 0.98 with B3LYP and 0.96 with CAMB3LYP. Infrared absorption and
VCD spectra were constructed from calculated dipole and rotational strengths assuming Lorentzian band shape
with a half-width at half maximum of 6 cm-1. All calculations were performed using Gaussian09 Revision A02.
-69-
N11
N11
C14
C14
N5
N5
73o
73oh
73oh
Figure 3-11 ORTEP view of crystal structure of rac-73o and (−)-(5RN,11SN,14SC) enantiomer shown. Hydrogen
atoms are omitted for clarity and thermal ellipsoids are drawn at 50% probability.
To conclude, cleché indicates that this dirhodium catalyzed enantiospecific ring expansion
reaction occurs with retention of configuration.
4.3.4
Scope and limitations of the reaction
The scope of this methodology was further examined using other methano-Tröger bases than
the classical 2,8-dimethylated derivatives of 1. First, an electron-rich Tröger base 74,5 was
used as a substrate (X=OMe, 98% ee) together with diazo-compound 72o. The reaction
worked even better than with 1 as the corresponding product 75 was isolated with excellent
diastereoselectivity (16:1 dr) and enantiospecificity (97% ee, Equation 3-16), Thus the
transfer of chirality (= 97/98) being 99% in the reaction.
Equation 3-16
5
Satishkumar, S.; Periasamy, M. Tetrahedron: Asymmetry 2009, 20, 2257.
-70-
Chapter 4
Then, other methano-Tröger bases were prepared.6 These derivatives (74a to 74f) where then
reacted with diazo-compound 72o. Electron-rich tetramethoxy-Tröger base 74a afforded the
corresponding ethano-Tröger base 75a in good yield and excellent diastereoselectivity (
Table 3-11, entry 1). Further, Tröger bases with electron-withdrawing substituents were tested.
The reaction of 2,8-dibromo-substituted Tröger bases 74b with 72o led to poor yield of
corresponding ethano-tröger analogue 75b. Yet, this was the better result for the rest of the
series. The reaction with 2,8-diodo-substituted Tröger bases 74c provide a complex mixture
of products. Further, more sterically demanding bisortho-substituted Tröger bases (74d, 74e
and 74f) were tested. No reaction was observed (entries 4 to 6) with these Tröger bases. This
was not unexpected in view of the strong steric hindrance around the N-atom (see chapter 1).
Table 3-11 General reactivity with methano-Tröger bases.a
Entry
(X)
Methano-TB
Y
Z
Product
Yieldb
drc
1
OMe
74a
OMe
H
75a
72
12:1
2
Br
74b
H
H
75b
32
nd
3
I
74c
H
H
75c
nd
nd
4
Br
74d
H
Me
75d
nd
nd
5
Me
74e
H
Br
75e
nd
nd
6
Me
74f
H
Me
75f
nd
nd
a
Conditions: 74 (0.1 mmol), 1 mol % Rh2(OAc)4, 2 equiv. of 72o, 0.25 mL of dry toluene, 100 °C, 6 h; Reported results are
the average of at least two experiments b Isolated yield (%) of the major diastereomer. c dr ratios determined by 1H-NMR
spectroscopy (400 MHz) on the crude reactions mixtures.
Furthermore, various hetero-aromatic diazo-compounds were prepared and tested (Figure
3-12). The results were again unsatisfactory, as in all these cases either no reaction or
complex mixtures of products were observed. Then again these results were not surprising in
6
D. A. Lenev, K. A. Lyssenko, D. G. Golovanov, V. Buss, R. G. Kostyanovsky, Chem. Eur. J. 2006, 12, 64126418.
-71-
particular for pyridine containing reactants as it can coordination to the dirhodium catalyst
pretty strongly and hence can quenched its reactivity.
Figure 3-12
In summary, we have developed a single step methodology to access ethano-Tröger bases
using diazo-compound and Rh2(OAc)4 as catalyst. This process provides very high level of
chirality transfer (ee up to 99%). A careful analysis of the results showed above indicated
nevertheless some shortcomings that could possibly be addressed. Indeed, high levels of
diastereoselectivity and chirality transfer were obtained only using diazocarbonyl reagents
derived from methyl arylesters. For all other tested diazo-compounds, selectivity ratios and
sometimes yields were poor. As, it had been demonstrated by other groups (see section-2.5.2)
that switching the nature of the catalyst might be beneficial so for instance, we decided to test
other metal sources, and copper in particular with these problematic reactions.
4.4 Copper-catalyzed Synthesis of Ethano-Tröger Bases
Copper sources (metal or salts) are in fact used more frequently than Rh(II) complexes in the
context of nitrogen ylide chemistry.7 A screening of different copper sources was therefore
preformed in the ring-expansion methodology, described above.
4.4.1 Screening of copper sources
First, the reaction of (+)-(S,S)-1 with diazo arylketone 72m (table 2, R1=Me, R2=Ph) was
selected as a reference test reaction, as it had previously with Rh2(OAc)4 afforded a moderate
7
(a) M. P. Doyle, M. A. McKervey, T. Ye, Modern Catalytic Methods for Organic Synthesis with Diazo
Compounds: From Cyclopropanes to Ylides. , Wiley, New York, N. Y., 1998; (b) J. S. Clark, Nitrogen, Oxygen
and Sulfur Ylide Chemistry: A Practical Approach, Oxford University Press, Oxford 2002, pp. 1 – 98.
-72-
Chapter 4
yield of product 73m (70%) and diastereoselectivity (dr 5:1) together with poor transfer of
chirality (ee 10%). The results are summarized in Table 3-12.
Table 3-12 Screening of copper sources a
Entry
Catalyst
Yield (%)b
drb
eec
1
Cu powder
60
1.0:1
25
2
Cu(OTf)2
89
3.5:1
5
3
[Cu(CH3CN)4][BF4]
85
12:1
15
4
Cu2(OTf)2·Toluene
72
3.0:1
16
5
CuBr
91
5.1:1
76
6
Cu2(OAc)2
82
5.5:1
78
7
CuCl
89
8.7:1
75
8
Cu(hfacac)2·H2O
85
5.0:1
72
9
Cu(acac)2
90
4.3:1
77
10
CuI
87
10:1
82
11
CuTC
85
11:1
85
12
Rh2(OAc)4
75
5:1
10
a
Conditions (+)-(S,S)-1 0.1 mmol, 5 mol% of catalyst, 0.2 mmol of diazo-compound 72m, 0.25 mL of toluene, 100 °C, 6 h;
1,3,5-trimethoxybenezene then added as internal standard. b By 1H-NMR analysis (400 MHz) of the crude reactions mixtures
using the internal standard. c Of the major diastereomer; determined by CSP-HPLC analysis.
To our satisfaction, all the copper sources afforded the desired ethano-Tröger base 73m in
moderate to excellent NMR yields (60-91%, entries 1 to 11). Yet, very strong differences
were noticed both in terms of diastereoselectivity and enantiospecificity. Only in the case of
copper powder, relatively poor results were obtained in term of diastereoselectivity and yields
as compared to corresponding reaction with dirhodium catalyst (entry 12), still the level of
enantiospecificity was better with copper powder (entry 1). CuTC, which, to best of our
-73-
knowledge, had never been reported in diazo-decomposition chemistry,8 turned out to be the
most effective copper catalyst for this process (73m, 85%, dr 11:1, ee 85%, entry 11). The
more classical CuI also provided comparable yield and selectivity (73m, 87%, dr 10:1, ee
82%, entry 10). This inspired us to select both CuTC and CuI for the remainder of the studies.
Further, Rh2(OAc)4 was also tested with same catalyst loading (5 mol%) as the copper
sources. In this case, it afforded compound 73m with same yield, diastereoselectivity and
10% ee (entry 12). Here, this increased catalyst loading of the dirhodium catalyst has no effect
on yield and diastereoselectivity of this reaction.
4.4.2 Dichotomy in copper and rhodium based catalysis
As shown in section 4.2, different acceptor-acceptor and donor-acceptor type of diazocompounds were synthesized. These substrates were then treated with rac-1 at 100 °C in
presence of 5 mol% of CuTC or CuI. These results were then compared with the reaction
using 5 mol% of Rh2(OAc)4 as catalyst (Table 3-13). Interestingly, the symmetric α-diazo βdiketone 72a (R1=Me, R2=COMe), which had previously not been able to afford a product
with Rh2(OAc)4 as catalyst (1 or 5 mol%), afforded now 73a in moderate to good yields.
Indeed, both copper sources provided 73a with better yield with CuTC (70%) as compared to
CuI (40%, entry 2).
Encouraged by this result, non-symmetrically substituted α-diazo-β-ketoester derivatives were
then examined. Compounds 72e and 72f (R1=Me, R2=CO2Me and CO2Et, respectively)
reacted equally well with all three catalysts to provide 73e and 73f in good isolated yields
(Table 3-13, entries 3 and 4). Although, quite better diastereoselectivity ratios were obtained
(dr 4.3:1 and 5.5:1, respectively) with CuTC in comparison with that attained with CuI (3.8:1
and 3.4:1 respectively) or Rh2(OAc)4 (2.4:1 and 2.1:1 respectively).Yet, it is important to note
that, both copper catalysts provided better diastereoselectivity relative to Rh2(OAc)4. Also, in
all these reactions, the three catalysts afforded the same diastereomer as the major component
of the crude reaction mixtures.
This better reactivity and selectivity with CuTC was further confirmed with more steric
demanding reactants 72g to 72l. Excellent yields and better dr values were obtained (73g to
73l: 78 to 91%, dr up to 6.0:1, entries 5 to 10) and this irrespectively of the increase in size
8
CuTC was only mentioned in the context of carbene self-dimerization using Fischer carbene complexes as
substrates and a transmetalation from Cr to Cu to generate the carbenoids intermediate. See J. C. del Amo, M. J.
Mancheno, M. Gomez-Gallego and M. A. Sierra, Organometallics, 2004, 23, 5021.
-74-
Chapter 4
for R1 or R2. These results contrast sharply with that obtained in the presence of Rh2(OAc)4
for which products 73g to 73l are isolated on average with a 25% yield only. Particularly in
the case of 72l, in which no product was formed using Rh2(OAc)4, CuTC provided the desired
product 73l with 78% isolated yield and 6.0:1 diastereoselectivity.
Table 3-13 Copper vs. rhodium catalysisa
Entry
Diazo
R1
R2
Product
CuI
CuTC
Rh2(OAc)4
Yieldb
drc
Yieldb
drc
Yieldb
drc
1
72m
Me
Ph
73m
81
10:1
80
11:1
70
5:1
2
72a
Me
COMe
73a
40
-
70
-
np
-
3
72e
Me
CO2Me
73e
78
3.8:1
80
4.3:1
78
2.2:1
4
72f
Me
CO2Et
73f
70
3.4:1
82
5.5:1
80
2.0:1
5
72g
Me
CO2iPr
73g
78
3.3:1
85
3.9:1
20
1.2:1
6
72h
Me
CO2tBu
73h
81
2.6:1
87
4.0:1
25
1.0:1
7
72i
Et
CO2Et
73i
87
4.3:1
91
5.0:1
19
2.4:1
8
72j
nPr
CO2Et
73j
83
3.7:1
80
5.0:1
25
2.7:1
9
72k
Ph
CO2Et
73k
65
4.2:1
80
4.3:1
49
3.3:1
10
72l
iPr
CO2Et
73l
46
4.8:1
78
6.0:1
np
-
11
72d
OEt
CO2Et
73d
13
-
40
-
85
-
12
72o
OMe
Ph
73o
nd
71
10:1
Nd
a
Conditions: rac-1 0.4 mmol, diazo reagent (0.8 mmol), CuTC (5 mol%) or Rh 2(OAc)4 (5 mol%), 1.0 mL of toluene, 100 °C,
16 h; reported results are the average of at least two experiments. The relative configuration displayed is that of the major
diastereomers of compounds 73. np stands for no product of type 73. b Isolated yield (%, both diastereomers). c dr ratios
determined by 1H-NMR analysis (400 MHz) of the crude reactions mixtures.
In the case of α-diazo β-diester 72d on the contrary, Rh2(OAc)4 outperform both the copper
catalysts (85 vs 40%, 13% for 73d, entry 11). Also in the case of 72o, Rh2(OAc)4 provided a
good yield and diastereoselectivity, whereas complex mixtures of products were observed
with both copper catalysts (entry 12). All in all, there seems to be a dichotomy among the
catalyst. Dirhodium complexes are better for catalyzing reactions of arylesters and diesters
-75-
diazo-compounds whereas CuI/CuTC are more efficient with all other class of reactants.
Studies to explain this dichotomous behavior of these catalysts will be shown in section 4.4.
Finally, if one compare the results obtained from 5 mol% loading of Rh2(OAc)4 (Table 3-13)
with that observed with 1 mol% loading (Table 3-7) section 4.3.2, it is evident that catalyst
loading has minimal or no effect on the outcome of the reaction within the dirhodium series.
4.4.3 Enantiospecificity
It was clear from the studies shown above that CuTC was a better catalyst than CuI and thus
was selected to investigate the chirality transfer. First, enantiopure (+)-(S,S)-1 was treated
with acceptor-acceptor type of diazo-compounds at 100 °C using CuTC as catalyst (Table
3-14).
Table 3-14 Screening of different classes of diazo-compounds for chirality transfer a
Entry
Diazo
R1
R2
Product
Yieldb
drc
eed
1
72a
Me
COMe
73a
70
-
64
2
72f
Me
CO2Et
73f
67
5.5:1
40
3
72m
Me
Ph
72m
73
11:1
85
4
72d
COEt CO2Et
73d
40
-
10
a
Conditions: rac-1 0.4 mmol, diazo reagent 72 (0.8 mmol), CuTC (5 mol%), 1.0 mL of toluene, 100 °C, 30 min; reported
results are the average of at least two experiments. The relative configuration displayed is that of the major diastereomers of
compounds 73. b Isolated yield (%, major diastereomers). c dr Ratios determined by 1H-NMR analysis (400 MHz) of the
crude reactions mixtures.
The reactions of -diketones 72a and ketoesters 72f with (+)-1, provided moderate level of
chirality transfer (ee 64 and 40% respectively, entries 1 and 2). A quite better transfer of
chirality was obtained using 72m (donor-acceptor) derived from phenylpropanone (ee 85%,
-76-
Chapter 4
entry 3). Only For symmetric -diazo--diester 72d, which had provided product 73d in 34%
ee with Rh2(OAc)4, we did observed a lower ee value of 10% with CuTC (entry 4).
Overall, Although moderate in most cases, an improvement was still noticed with CuTC over
the previously used Rh2(OAc)4, particularly for the diazo-compounds derived from phenyl
ketone 72m. This class of arylketone diazo-compounds was then selected for a scope
extension studies using CuTC as catalyst.
So far, even though results are better with CuTC, relatively extensive losses in enantiopurity
are observed for the products and it was then making sense to try to understand more clearly
the reasons for this loss prior to perform a full study on the scope of reaction. It of course
occurred to us that the decrease in enantioselectivity could come either by a racemization of
the starting material under the reaction conditions or even possibly of the product. As, such
(+)-1 and (+)-73m were independently subjected to 5 mol% of CuTC in toluene at 100 °C for
1 hour. Whereas no racemization was observed for and (+)-73m, a decrease of ee value (9798%) was seen for (+)-1.9 However, these rather small decrease in the enantiopurity of (+)-1
not enough to explain the large loss in chirality transfer (detailed previously). Reactions were
thus performed again (toluene, 100 °C, 1 h) with this time in the 0.5 equiv. of -diazo-phenylketone 72m, instead of 2 equiv. Due to the sub-stoichiometric amount of diazo-reagent,
it was possible to isolate both the product 73m and the substrate 1 at the end of the reaction.
The results were bit puzzling. As (+)-1 was isolated in a lower ee value than previously (93%
vs 97%, 45% yield) and on the contrary, 73m (20% yield) was isolated in 90% ee (vs 85%
with 2 equiv. of 72m). Clearly, from the loss of ee of (+)-1, a more acidic species is generated
in the presence of diazo-compound. To slow this racemization during the reaction,
temperature was then lowered to 90 °C (Table 3-15).
To our satisfaction, under the classical conditions using 2 equiv. of 72m, an increase in
enantiospecificity of 73m (90%) was observed (entry 2). But lowering further the temperature
to 80 °C did not produce the awaited results as this time a lower ee value was obtained (entry
3). The reason for this latter observation was not clear.
9
These results are in line with Rh2(OAc)4, where no racemization was observed for and (+)-73m and (+)-1.
-77-
Table 3-15 Study of enantiospecific transformation with temperature
Entry
Diazo
Temp. (°C)
Product
Yield b
dr c
Ee
1
72m
100
73m
73
11:1
85
2
72m
90
73m
70
11:1
90
3
72m
80
73m
67
11:1
87
a
Conditions: rac-1 0.4 mmol, diazo reagent 72 (0.8 mmol), CuTC (5 mol%), 1.0 mL of toluene, T °C, 30 min; reported
results are the average of at least two experiments. The relative configuration displayed is that of the major diastereomers of
compounds 73. b Isolated yield (%, major diastereomers). c dr Ratios determined by 1H-NMR analysis (400 MHz) of the
crude reactions mixtures.
Yet, from this short study, we could determine that a 90 °C temperature is quite better than
100 °C for this class of reactions.
Further, several diazo reactants made from arylketones (72ma to 72mf) were prepared and
according to the procedure mentioned in section 4.3.2 and tested (Table 3-16). In all cases, the
major diastereomers of 73ma to 73mf were obtained with good diastereoselectivity (11:1> dr
> 8:1) and isolated yields (80-85%).
The substitution of methyl by ethyl or phenyl substituents  to the carbonyl group had little
effect in terms of enantiospecificity as the products 73ma and 73mb (entries 4 and 5) being
afforded with the same 90% ee value as 73m. In the case of 73mb, a simple recrystallization
in acetone by slow evaporation of the solvent afforded the product in 95% ee (entry 5).
A series of diazoketones carrying different substituents on the aryl moiety namely reagents
72mc to 72mf (R2=p-MePh, m-MePh, p-ClPh, p-NO2Ph) were then used. These diazoreagents afforded similar yields and diastereoselectivity ratios (entries 6 to 9), in line with
those previously discussed.
-78-
Chapter 4
Table 3-16 Efficient chirality transfer a
Entry
diazo
R1
R2
X
Product
Yield b
dr c
ee d
1
72m
Me
Ph
Me
73m
73
11:1
90
3
72f
Me
CO2Et
Me
73f
67
5.5:1
50
4
72ma
Et
Ph
Me
73ma
83
11:1
90
5
72mb
Ph
Ph
Me
73mb
85 (80) f
8:1
90 (95) f
6
72mc
Me
p-MeC6H4
Me
73mc
80
9:1
82
7
72md
Me
m-MeC6H4
Me
73md
83
10:1
84
8
72me
Me
p-ClC6H4
Me
73me
83 (75) f
10:1
93 (99) f
9
72mf
Me
p-NO2C6H4
Me
73mf
85
11:1
95
10
72m
Me
Ph
OMe e
73mg
80
12:1
92
a
Conditions: (+)-(S,S)-methano-Tröger bases 0.1 mmol, diazo reagent (0.2 mmol), 5 mol % CuTC, 0.25 mL of dry toluene,
90 °C, 30 min. Reported results are the average of at least two experiments b Isolated yield (%) of the major diastereomer. c
dr ratios determined by 1H-NMR spectroscopy (400 MHz) on the crude reactions mixtures. d ee values determined by CSPHPLC. e ee 98%. f After crystallization.
Better enantiomeric purity values (ee ≥ 93%) were however obtained with reagents carrying
electron-withdrawing groups (p-Cl and p-NO2: 72me and 72mf respectively) instead of
electron-donating ones (ee ≤ 84%, p- and m-Me: 72mc and 72md). The product 73me on
single recrystallization in acetone afforded in 99% ee. Clearly, within this series of
diazocarbonyl derived from arylketones, the highest levels of enantioselectivity are observed
with the most reactive electrophilic carbenoids. Interestingly, this trend is exactly opposite to
that observed with arylester diazo derivatives under Rh2(OAc)4 catalysis. These results again
highlighted the dichotomous behavior of these two catalytic systems.
Furthermore, non-racemic 2,8-dimethoxy methano-Tröger base (ee 98%) was also treated
with diazo 72m under the new conditions and good yield (80%), dr (12:1) and ee (92%)
values were obtained.
-79-
Finally, mono-crystals for the major diastereomer of (+)-(5SN,11RN,14RC)-72me were
obtained. Here again, the presence of chlorine atom led to an assignment of the absolute
configuration. This indicates that the CuTC catalyzed enantiospecific ring expansion reaction
of (+)-1 also occurs with retention of configuration (Figure 3-13).
N11
C14
N5
Figure 3-13
4.5 Mechanistic Insight
Now, with all the gathered information, a mechanistic rationale can be proposed for these
reactions, and for the rhodium catalyzed process in particular.
4.5.1. Proposition
It is widely accepted in the literature that the first step in rhodium-catalyzed reactions is the
decomposition of diazo-compounds to form electrophilic metallocarbenoid. We believe that it
is also the case in our catalytic process. The generated carbenoid reacts subsequently with
methano-Tröger base 1 to form a metal-bound nitrogen ylide 76´ (Scheme 3-8); this ylide can
also exist possibly in a metal free state 76.
-80-
Chapter 4
Scheme 3-8
This ylide intermediates, 76` or 76, can then undergo [1,2]-Stevens type rearrangement via
either a biradical 77 or iminium/enolate 78 intermediates as shown in Scheme 3-9. These ringopened intermediates, biradical or iminium/enolate, can then collapse to form the ethanobridge, and this at a rate sometimes faster than the conformational relaxation to afford then
the desired ethano-Tröger base with high enantiospecificity. We will discuss each step of the
proposal in more details in coming paragraphs.
Scheme 3-9
4.5.2. Elements confirming the mechanistic rationale
The first considered step was the decomposition of the diazo-reactants by the metal catalyst.
This proposition is supported by the observation of rapid gas evolution upon the reaction at 90
°C or 100 °C and by the in-situ monitoring the decomposition of diazo-compounds using
React-IR™ instrument (these results will be discussed later in this section 4.5.3).
-81-
The next step is the formation of nitrogen ylides 76` or 76 which then undergo ring expansion
reaction through biradical or iminium/enolate intermediates. In fact, these intermediates 77 or
78 are key to the process and we believe that the efficacy of chirality transfer is essentially
related to the stability of these ring-opened intermediates (Scheme 3-10).
Scheme 3-10
If these open-ring intermediates 77 and 78 are long lived, then the entities generated with a
chiral puckering can relax conformationally as shown in Scheme 3-10. The consequence is
the formation of two enantiomeric conformations of 77 and 78 and consequently of the
product in a lower enantiomeric purity. The ee values depending on the conformational
relaxation.
Indeed as shown in section 4.4.3, under optimized reaction conditions, no racemization is
observed for 1 and also for the product 73m and hence we rationalize that the loss of
enantiomeric purity is occurring from either at the ylide stage (76` or 76) or more likely at the
level of intermediates 77 and 78.
This fits then the observation of a rather strong racemization with acceptor-acceptor type of
diazo-compounds which highly stabilized.10 With donor-acceptor type of diazo-compounds
derived from phenylesters and phenylketone, the stability of radical intermediate 77 would
remain almost the same and for the intermediate 78 possess a more reactive enolate capable of
reacting immediately after its formation with the pendent iminium moiety.9
The pKa value of a -diester is 16.4 in DMSO, and for aryl ester an aryl ketones (22.7 and 19.8). Diester
enolate is therefore more stable than with aryl ester and ketones.
10
-82-
Chapter 4
As shown in Table 3-10, -diazo--phenylesters containing electron-donating groups provide
better enantiospecificity (ee 98%) than those containing electron-withdrawing substituents (ee
97%). The presence electron-withdrawing groups on the aryl substituents lead (again) to a
stabilization of the ring opened intermediates and hence to a large racemization, in
particularly for the stronger p-NO2 group. This gradual loss in ee values with stronger
electron-withdrawing substituents on the aryl part also supported the involvement of
intermediate 78.
In summary, experimental results are well supported by the involvement of an
iminium/enolate type of intermediate 78, in particular.
To further support our mechanistic proposal, an unsymmetrically-substituted methano-Tröger
derivative 79 was prepared.11 This compound 79 has both electron-donating methoxy group
(OMe) and electron-withdrawing nitro (NO2) groups para to the N-atoms on the two
perpendicular aromatic rings (Equation 3-17). This unsymmetrical substitution pattern on 79
makes the two nitrogen atoms electronically different. This compound was then reacted under
both rhodium and copper catalysis. To avoid complex mixtures of diastereomers in addition to
the expected regioisomers, a symmetric -diazo--diester diazo-compound 72d was selected.
Treatment of methano-Tröger base 79 with 72d at 100 °C using both CuTC and Rh2(OAc)4
were performed. While the reaction with CuTC did not lead to any product, Rh2(OAc)4
provided 80 in 71% yield.
Equation 3-17
Moreover, the reaction was highly regioselective, a single regioisomer 80 being obtained of
which the structure could only be determined by a careful NMR analysis.
12
Interestingly,
product 80 arises from the single reaction of the electron-poor nitrogen atom (para to the nitro
11
Pardo, C.; Ramos, M.; Fruchier, A.; Elguero, J. Magn. Reson. Chem. 1996, 34, 708.
12
In the Experiment part, the HMBC spectrum leading to the described structure are detailed
-83-
group) rather than the more electron-rich nitrogen atom (para to the methoxy group, Equation
3-17).
This result was, at first glance, surprising. Yet, on second thought, it could be quite easily
rationalized by considering the intermediates already detailed and the iminium/enolate
intermediate in particular. For this compound, the presence of two different types of nitrogen
atoms should result in two different ammonium ylides 81 and re-81.13 Ylide re-81 is formed
by the reaction of more electron-rich nitrogen atom with the electrophilic metallocarbenoid. In
principle, its formation should be preferred over that of ylide 81 which forms via the reaction
of electron-poor nitrogen atom instead (Scheme 3-11).
Scheme 3-11 Radical mechanism of regioselective reaction
These ylide intermediates 81 and re-81 can then undergo the ring expansion reaction through
either biradical (Scheme 3-12) or through iminium/enolate intermediates (Scheme 3-13). As
we will see, this experiment provides the mean to distinguish between these two pathways
that are detailed below sequentially.
If biradical pairs are formed through ylides intermediates 81 and re-81 (Scheme 3-12), then
each resulting pair of biradicals 83 and re-83 will be most likely equally stabilized by the
nitrogen atoms and their substituents.14 And such, the proportion of 81 and re-81 should be
reflected in that of 80 and re-80 if this mechanism is involved. Moreover, as ylide re-81 has a
higher probability of being formed, re-80 should be the major product of the reaction, which
was not the case experimentally. This evidence suggests that a biradical pathway is probably
not involved.
13
For simplification, 81 and 82 are presented as metal free ylides in the following diagrams.
The bond between the radical and the nitrogen atoms are the 3-electron bond. They can be as easily described
by a mesomeric form with a single e– on the N-atom and 2 e– on the carbon instead.
14
-84-
Chapter 4
Scheme 3-12
The other possibility is the formation of iminium/enolate intermediates, as shown in Scheme
3-13. Interestingly, although more stabilized, ylide re-81 will provide a highly disfavored
iminium/enolate intermediate re-84 in which the iminium ion is para to the nitro-group and
the electron-rich nitrogen atom is adjacent to the produced enolate. The electronic
contributions are therefore totally unfavorable on both nitrogen atoms of re-84.
Scheme 3-13
-85-
While, with 84 derived from the less-stable ylide 81, the iminium ion is para to methoxy
group and the carbanion electrons can be drowned towards the nitro group. Therefore,
intermediate 84 is highly favored and will yield after the closure, the observed product.
At this stage, it is still not clear why the less-stable ylide 81 is formed preferentially. It cannot
be explained by a preferred formation under kinetic control, in particular. Since, 79 react with
MeI to generate 85, corresponding ammonium ion on the electron-rich N-atom (Equation
3-18).
Equation 3-18
The best explanation so far is a reversible formation of ylide intermediates re-81 and 81 and a
Curtin-Hammett situation. The most stable ylide 81 is formed preferentially but it is 81 that
react further only to yield selectively 84 and then product 80. A reversible formation of
nitrogen ylide has been reported twice in the literature and it was mentioned that it could be
promoted by the presence of the metal catalyst (Scheme 3-14).
Scheme 3-14
Finally, as mentioned the experiment advocated for the participation of iminium/enolate
intermediates. More evidence for this could be gathered by studying the side products of the
reaction mixtures. As mentioned before, diazocines of type 63 are major impurities formed in
the reaction (Scheme 3-15). These diazocine impurities are formed by the loss of the
methylene-bridge presumably by addition of water molecule to the iminium or the ylide
-86-
Chapter 4
intermediate and then by subsequent liberation of formaldehyde. If a biradical mechanism is
considered, hydrogen abstraction reactions should occur instead. However, N-methylated
products of type 90 never been observed. This observation supports again the iminium/enolate
intermediates rather than the biradical one (Scheme 3-15).
Scheme 3-15
In summary, evidences supports better the iminium/enolate type of intermediate for this
reaction, and the global mechanistic rationale accounts for all the major observation made
during the study.
However some things are still to be explained including the influence that the nature of metal
catalyst has on this transformation. As shown in sections 4.4.2 and 4.4.3, for the same diazoreactant 72m (R1=Me, R2=Ph), we have observed very different reactivity and
enantiospecificity using different catalysts. It was decide then to attempt to understand these
differences.
-87-
4.5.3. In-situ monitoring
Finally, to achieve a better understanding of the differences observed with CuTC and
Rh2(Oct)4 as catalysts,15 we tried to gather more information by following the reactions using
an in-situ method of monitoring of the consumption of the reactants and the synthesis of the
products. We selected an infrared spectroscopy as both starting diazo reagent (72m) and its
corresponding ethano-TB product 73m can be distinguished by characteristic absorbance in
the studied spectral window. As you will see, rather than clarify the situation presented in the
previous section, the collected data give interesting yet complex information about the
mechanistic situation.
Equation 3-19
All the reactions were carried out using a React-IR™ 45m instrument in a special Schlenk
tube fitted with an immersible in-situ FT-IR DiComp probe. The following reaction
conditions (toluene, 100 °C, 5 mol% of catalyst) were used and a stirring speed of 1000 rpm
was selected.16 Signals at 2108 and 1708 cm-1 were utilized for the monitoring as those are
characteristic absorbance of 72m and ethano-Tröger base 73m respectively. Data collection
was performed every 10 seconds.
The results are summarized below. As shown in Figure 3-14, diazo 72m is rapidly consumed
at 100 °C. The decomposition is faster with Rh2(Oct)4 over CuTC; the process being over in
25 and 100 seconds respectively!17 Whereas the profile of the diazo consumption is somewhat
classical with Rh2(Oct)4, it is not the case for the reaction catalyzed by CuTC. This is due to
the minimal amount of data collected in this short span of the decomposition but not only.
. 15 for Solubility reasons, Rh2(Oct)4 was used in place of Rh2(OAc)4, also no noticeable effect was seen in
changing the dirhodium catalyst for all the classes of diazo-compounds.
16
The stirring speed was set to 1000 rpm for all of the kinetic measurements. A cylindrical-shaped magnetic
stirrer was used throughout the present study. For the effect of stirring speed: Sun, Y.; Landau, R. N.; Wang, J.;
LeBlond, C.; Blackmond, D. G. J. Am. Chem. Soc. 1996, 118, 1348-1353
17
These results are in agreement with the known high reactivity of dirhodium complexes as compared to copper
salts for diazo-decomposition reactions.
-88-
Chapter 4
1.2
1
Rh2(oct)4
Conversion
0.8
CuTC
0.6
0.4
0.2
0
0
50
-0.2
100
150
200
Time (in seconds)
Figure 3-14 Decomposition of diazocompoundn72m with Rh2(oct)4 (in blue) and CuTC (in red)
The progress of the reactions was also studied by plotting graphs corresponding to the
formation of 73m with respect to time. There are shown in Figure 3-15 and Figure 3-16.
Unexpectedly, for both catalytic systems, there are clearly two different catalytic regimes. An
initial regime (Figure 3-15), before 100 seconds, that sees a rapid formation of 73m and then,
after this period, a much slower synthesis of 73m in comparison (Figure 3-16).
1
1
Rh2(oct)4
0.9
CuTC
0.8
0.8
0.7
0.7
Conversion
Conversion
0.9
0.6
0.5
0.4
0.6
0.5
0.4
0.3
0.3
0.2
0.2
0.1
0.1
Rh2(oct)4
CuTC
0
0
0
20
40
60
80
0
100
400
600
800
1000
1200
Time (in seconds)
Time (in seconds)
Figure 3-15 Formation of product 73m in “initial
regime” (0 – 50 seconds) using Rh2(Oct)4 (in blue)
and CuTC (in red) as catalysts.
200
Figure 3-16 Over all reaction profile for the formation of
73m using Rh2(Oct)4 (in blue) and CuTC (in red) as
catalysts.
-89-
With such decomposition and synthesis profiles, only qualitative conclusions can be made. As
just seen, in both systems, the rapid formation of 73m occurs in the “first regime” while the
process of diazo decomposition is actually still occurring. After that period, when there is no
more diazo reactant, production of 73m is much slower.18 As such, it looks like there are two
different mechanistic pathways for the formation of 73m from 72m – one fast (before 100s)
and a much slower one afterwards. A crude mathematical analysis considering the mean slope
of the two regimes indicates that the first is 14 (± 10%) times faster than the second. The
origin of these two regimes – and their implication in terms of the enantiospecificity of the
reactions in particular – is hard to assess under the current situation.19 Some differences can
be noted for CuTC and Rh2(Oct)4 but we are not sure of their relevance for the moment.
4.5.4. Conclusion
In summary, we have developed an enantiospecific process to prepare different types of
configurationally stable ethano-Tröger bases using one step protocol. A mechanistic rationale
involving iminium species as the key intermediate has been proposed by studying the effect of
various substituents on the diazo-compounds for the enantiospecificity of the reaction. A brief
study on the effect of metal source on enantiospecificity was also presented. However, these
studies of reaction profile using React-IR turn out to be rather complicated to draw any firm
conclusions.
Next, we decide to look for the applications of this easily accessible enantioenriched ethanoTröger bases. This will be shown in next chapter.
18
It could be accounted by a “massive” production of ylides intermediates in the first regime that react more
slowly in the second. However, signatures for the ylide(s) intermediate(s) could not be found in the React-IR
spectra.
19
An interesting hypothesis would be that the first regime corresponds to the synthesis of the ethano-TB under
kinetic control while the second period is that of a more thermodynamic approach. This is to be considered in
view of the results obtained in the reaction of 72m with (+)-(S,S)-1.
-90-
Chapter 5
5. Applications of Ethano-Tröger Bases
As mentioned in the previous chapters, we have developed a methodology which provides
ethano-Tröger bases in high enantiomeric purity. The next logical step was the development
of applications for these derivatives. We looked for them in the fields of asymmetric catalysis
and chiral stationary phase (CSP)-chromatography.
5.1
Asymmetric Catalysis
Concerning catalysis, we selected a reaction that had been previously described with
enantiopure methano-Tröger base 1 as organocatalyst. As mentioned in chapter 1, Shi and
coworkers have shown that enantiopure Tröger base 1 can be used as “catalyst” for the
aziridination of chalcone derivatives.1 For this reaction to occur at a decent rate and with
moderate enantioselectivity, a rather large catalyst loading (60 mol%) was necessary. This,
together with the mechanism involving a quaternary ammonium intermediate, led to us think
that a racemization of 1 ought to occur in the reaction. If this would be indeed the case, then
configurationally stable ethano-Tröger bases might provide a solution.
Equation 5-1
First, the reaction was reproduced using Shi’s conditions. In our hands, the product was
obtained with the same ee value as reported and, as expected, catalyst (+)-1 was recovered
with 93% ee only (Table 5-1, entry 1). This result indicated that part of the moderate
enantioselectivity could be due to the partial racemization of (+)-1. When this reaction was
tried at ambient temperature (25 °C vs.-20 °C previously), the product was isolated in 45% ee
and (+)-1 was recovered with 81% ee – a quite lower value than before.
1
Y. M. Shen, M. X. Zhao, J. Xu, Y. Shi, Angew. Chem. Int. Ed. 2006, 45, 8005-8008.
-91-
Applications of Ethano-Tröger Bases
In view of these results, the reaction was tested using ethano-Tröger base (+)-73o (ee 99%)
instead. Initial results at -20 °C were disappointing as results were worse than that obtained
with (+)-1 (entry 3). The product 86 was isolated with 10% yield only and a 60% ee value.
The only positive observation was the lack of racemization of catalyst (+)-73o since it was
recovered, after chromatography, in 90% yield and 99% ee. Also, catalysts (+)-1 and (+)-73o
afford the same predominant enantiomer, which is quite logical considering the known
absolute configurations of the compounds, (5S,11S) and(5SN,11RN,14RC) respectively.
In an attempt to optimize the yield, we decided to increase the temperature of the reaction to
25 °C. Again, no racemization of “catalyst” (+)-73o was observed and the product was
isolated with 58% yield and 57% ee (entry 4). This result was encouraging as (+)-73o is this
time more selective than (+)-1 at that temperature (entry 2) and, after purification, it could
also be recovered with the same enantiomeric purity unlike its parent compound (+)-1.
Catalyst loading was further reduced to 30 mol% to provide the same selectivity and yields
after just a small modification of the experimental procedure.
Table 5-1
Entry
Catalyst
Temp. (T °C)
Product (86)
Catalyst recovered
Yield (%)
ee (%)
Yield (%)
ee (%)
1
(+)-1
-20
66
67
75
93
2
(+)-1
25
69
45
60
81
3
(+)-73o
-20
10
60
90
99
4
(+)-73o
25
58
57
90
99
5
(+)-73o
25
55
57
90
99
Although these results are not in the current standards of the scientific community in terms of
selectivity and reactivity, they support conceptually our hypothesis that ethano-Tröger bases
have the potential to replace methano-Tröger bases in many of their asymmetric applications
– especially those that require the formation of quaternary ammonium intermediates and
-92-
Chapter 5
elevated temperatures to proceed efficiently. Efforts are still happening in the group to
optimize the reaction conditions and to screen different derivatives and reactants to achieve
more effective catalysis.
5.2
Applications in CSP-Chromatography
The second application that was considered lies more in the field of analytical chemistry.
Today, chiral stationary phases (CSPs) are recognized as the most effective tools for the
analytical and preparative enantiomer separation of chiral compounds.2As mentioned in
chapter 1, after the initial report of Prelog and Wieland of the chiral nature of 1 and its
resolution on a cellulose-based material,3 the methano-Tröger base has become the most
popular analyte for the evaluation of novel CSPs. Interestingly, essentially all these new
phases have been able to separate the enantiomers of 1. Then, by application of Pirkle’s
reversibility principle,4 a chiral stationary phase containing a derivative of 1 ought to be itself
a most effective CSP for all the molecules that were grafted on the new columns.
Unfortunately, until now, chemists were never able to prepare such a CSP due to the
configurational instability of Tröger base 1 and analogues. This status drew our attention as
the configurationally stability of the ethano-Tröger base motif could upset the situation. The
work presented in this part was performed in collaboration with the group of Prof Dr
Wolfgang Lindner (Universität Wien) and Dr Eric Francotte (Novartis Institutes for
BioMedical Research).
5.2.1 Grafting of ethano-Tröger base
To develop the CSP, it was necessary to elaborate first the design and the synthesis of a novel
highly enantioenriched 5 ethano-Tröger base containing a synthetic handle for the grafting on
silica gel; this extra functional group being introduced a priori at a position remote from the
expected sites where stereoselective interactions should occur. As the chiral molecular cleft
2
(a) Y. Okamoto, T. Ikai, Chem. Soc. Rev. 2008, 37, 2593. (b) E.R. Francotte, J. Chromatogr. A 2001, 906, 379.
(c) M.R. Buchmeiser, J. Chromatogr. A 2001, 918, 233.
3
V. Prelog, P. Wieland, Helv. Chim. Acta 1944, 27, 1127-1134.
4
(a) W. H. Pirkle and D. J. Hoover, Top. Stereochem. 1982, 13, 263–331. (b)W. H. Pirkle, D. W. House and J.
M. Finn, J. Chromatogr., A 1980, 192, 143–158.
5
In CSP chromatography, the chiral selector does not need to be enantiopure to achieve the perfect resolution of
chiral analytes.
-93-
constituted by the two “perpendicular” aromatic groups,6 and the two basic nitrogen atoms
might be key elements for recognition events, it was decided to introduce the extra group at
the distance from them, and at the para-position of the aromatic ring added during the ethanoTB synthesis in particular. For the handle, we selected a double-bond as precedents from the
group of Prof Lindner indicated that grafting should occur readily using sulphide-modified
silica gel. Also, importantly, the introduction of the double bond at that remote position of the
substrate should not divert the metallocarbene reaction. Finally, as diazo derivatives derived
from electron-rich aryl esters afford better selectivity and yields in Rh2-catalyzed syntheses,
we elected 87 containing a p-O-allyl moiety as target molecule.
The diazo reactant 87 necessary for its synthesis was made in three steps from commercially
available 4-hydroxyphenylacetic acid by a selective esterification of the carboxylic acid under
standard conditions (MeOH, H2SO4, 98%). Allylation of the phenol residue (allyl bromide,
K2CO3, 95%) afforded the functionalized ester which was then treated with p-ABSA and
DBU to obtain the target reactant.
Then, compounds (+)-88 and (−)-88 were synthesized using (+)-(S,S) and (−)-(R,R)-1 as
substrates respectively and the standard protocol described in section 4.3.3. In these reactions,
which could be performed on rather large scale (2.0 mmol, 0.50 g) due to the intended
application, the major diastereomer of compound 88 was obtained in 70% isolated yield and
98% ee (Equation 5-2). A total of 500 mg of (+)-(5SN,11RN,14RC)-88 were then sent to
Vienna.
6
(a) M. J. Crossley, L. G. Mackay, A. C. Try, J. Chem. Soc., Chem. Commun. 1995, 1925-1927; (b) E. B. Veale,
T. Gunnlaugsson, J. Org. Chem. 2010, 75, 5513-5525. (c) J. Artacho, P. Nilsson, K.-E. Bergquist, O. F. Wendt,
K. Wärnmark, Chem. – Eur. J. 2006, 12, 2692. (d) J. Jensen, J. Tejler, K. Wärnmark, J. Org. Chem. 2002, 67,
6008.
-94-
Chapter 5
Equation 5-2
There, in the group of Prof Lindner, the chiral stationary phase was prepared by
immobilization of the ethano-Tröger base onto a 3-mercaptopropyl-modified silica (4.58% C,
1.12% H corresponding to 0.95 mmol SH per gram of modified silica) using a radical addition
reaction shown in Equation 5-3. The solid phase was further characterized by elemental
analysis to estimate the selector coverage. A total of 360 µmol/g of 88 was immobilized which
ascertained the quality of the CSP grafted with ethano-Tröger base 88.
Equation 5-3
Finally, the packing material was slurry packed at 600 bar into a 250 x 3 mm stainless steel
column, which was then evaluated for its enantioselectivity and efficiency with chiral analytes
of diverse structures. This screening was realized by Dr. Georg Schuster in Basle in the team
of Dr. Eric Francotte (Novartis).
-95-
5.2.2 Chiral resolution using Tröger base grafted CSP (TB-CSP)
As just mentioned, the TB-CSP was tested with a collection of different chiral racemic
analytes and initial tests were rather discouraging as hardly any separation was observed.
Considering the -basicity of the TB selector, different -acidic analytes were tested. Again,
these efforts led to poor results. Then, methano-Tröger base 1 was tested itself and, to our
delight, it was recognized quite nicely by the ethano-TB-CSP as shown in Figure 5-1.
800 Troeg erBasenStudy #69 [modified by Lab_DIEHLGE1]
mAU
TB
UV_VIS_3
WVL:254 nm
1 - 5.567
2 - 5.990
600
(−)-(R,R)-1
(+)-(S,S)-1
400
200
-100
0.00
k1
k2
α
0.69
0.82
1.19
R
2.01
min
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
5.50
6.00
6.50
7.00
7.50
8.00
Figure 5-1
This result suggested the possibility of a “self-recognition” of the Tröger base scaffold.7 To
verify this hypothesis, it was decided to test other methano-TBs containing various
substituents at 2,8-positions (para to the aniline N-atoms). Satisfactorily, the TB-CSP
provided a separation for all methano-TBs tested irrespective of the electron-donating or
withdrawing ability of the substituents (Figure 5-2). Differences were nevertheless noted
between the various analytes as TBs containing electron-withdrawing groups (CN and CHO)
presented better resolution factors (R = 5.36 and 4.48 respectively) than those with donating
groups (R = 2.13 (OMe) and 1.75 (Et)).
7
(a) U. Kiehne, T. Weilandt, A. Lützen, Org. Lett. 2007, 9, 1283. (b) U. Kiehne, T. Weilandt, A. Lützen, Eur. J.
Org. Chem. 2008, 2056.
-96-
Chapter 5
mAU
Nat8 16-35*
UV_VIS_3
WVL:254 nm
1 - 9.464
200
2 - 10.337
Eluent: Hexane/EtOH
150
k1
k2
α
1.87
2.13
1.14
100
85/15
R
2.13
50
0
min
-50
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
1'200 Troeg erBasenStudy #67 [modified by Lab_DIEHLGE1]
mAU
8.0
9.0
10.0
11.0
12.0
13.0
SS-43
14.0
UV_VIS_3
WVL:254 nm
1 - 5.141
2 - 5.464
1'000
Eluent: Hexane/EtOH
750
k1
k2
α
0.56
0.66
1.18
500
85/15
R
1.75
250
0
-200
0.00
min
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
mAU
4.50
5.00
5.50
6.00
6.50
7.00
7.50
TB38*
UV_VIS_3
WVL:254 nm
2 - 29.022
150
Eluent: Hexane/EtOH
k1
k2
α
7.79
9.60
1.23
8.00
85/15
3 - 34.993
R
100
5.36
50
1 - 25.165
-20
0.0
2.0
4.0
6.0
8.0
mAU
min
10.0
12.0
14.0
16.0
18.0
20.0
22.0
SS-70*
24.0
Eluent: Hexane/EtOH
26.0
28.0
30.0
32.0
34.0
36.0
38.0
40.0
42.0
45.0
UV_VIS_3
WVL:254 nm
5 - 22.596
85/15
6 - 27.584
60
k1
k2
α
5.85
7.36
1.26
R
4.84
40
20
1 - 8.3502 - 9.383
-10
0.0
3 - 10.9404 - 12.168
min
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
22.0
24.0
26.0
28.0
30.0
32.0
35.0
Figure 5-2 Chiral resolution of methano-Tröger bases of type 1
Unsymmetrically substituted TBs, containing different substituents at positions 2 and 8, were
also considered and a better separation was observed for the substrate containing OMe and
NO2 groups (R = 5.62) in comparison with that having Br and H atoms para to the nitrogen
atoms (R = 2.10) (Figure 5-3).
-97-
mAU
Mono Br
UV_VIS_3
WVL:254 nm
1 - 6.814
2 - 7.297
600
Eluent: Hexane/EtOH
k1
400
k2
α
85/15
R
1.21 1.14 2.10
1.06
200
3 - 7.789
-100
0.0
4 - 8.595
min
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
mAU
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
AS840R
10.5
11.0
UV_VIS_3
WVL:254 nm
1 - 15.625
2 - 18.658
Eluent: Hexane/EtOH
80
k1
60
k2
α
85/15
R
3.73 4.65 1.25 5.62
40
20
-10
0.0
min
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
17.0
18.0
19.0
20.0
21.0
22.0
Figure 5-3 Chiral resolution of unsymmetric methano-Tröger bases
Then, ethano-TBs were tested as analytes. From the experiments (see below), similar
conclusions can be drawn. As shown in Figure 5-4, a better recognition was observed with
TBs containing better electron-withdrawing rather than electron-donating groups (R = 5.9,
3.84 and 3.41 for 73mf (NO2), 73me(Cl) and 73oa (OMe), respectively).
Figure 5-4 Effect of change in electronic parameters of para substitution on chiral resolution using TB-CSP
-98-
Chapter 5
All in all, these results support the “self-recognition” hypothesis. The fact that it is more
effective with TBs containing electron-withdrawing groups suggests that a -acidic (analyte) /
-basic (selector) interaction is capable of reinforcing the stereorecognition event.
Finally, the racemic and levorotatory forms of the selector 88 were tested. The (+)(5SN,11RN,14RC) and (−)-(5RN,11SN,14SC) enantiomers are less- and more-retained by the
TB-CSP respectively (Figure 5-5). As the stationary phase has been prepared from (+)(5SN,11RN,14RC)-88, this result tends to indicate a preference for a heterochiral mode of
association of the Tröger fragments.
14C
11N
(−)-(5RN,11SN,14SC)
(+)-(5SN,11RN,14RC)
5N
Figure 5-5
Care was then taken to analyze again the results from the experiment performed with rac-1
and identify with certainty the elution order of the enantiomers; the results were shown
already on Figure 5-1. A similar analysis was then performed with Hamada’s ethano-TB 16.
mAU
Et TB R
CH3
600
UV_VIS_3
WVL:254 nm
Eluent: Hexane/EtOH
1 - 5.499
2 - 5.746
k1
k2
α
0.67
0.74
1.11
N
400
85/15
R
1.12
N
H3C
200
-100
0.0
min
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
1 - Troeg erBasenStudy #54
1'200 2 - Troeg erBasenStudy #55
mAU
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
Et TB P1
Et TB P2
1 - 5.503
1'000
Eluent: Hexane/EtOH
(S,S)
k1
(R,R)
750
k2
α
85/15
9.5
10.0
UV_VIS_3
UV_VIS_3
WVL:254 nm
R
0.67
0.74
500
1.11
1.11
250
2
0 1
-200
0.0
min
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Figure 5-6
-99-
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
In these two cases, the most retained enantiomers are the (−)-(5RN,11RN)-1 and (−)(5SN,11SN)-16. These enantiomers, despite an apparent inversion of the absolute configuration
due to a change of CIP priorities, have the same tridimensional (3D) disposition of the groups
around the stereogenic atoms as (−)-88. We can thus conclude that heterochiral association is
definitely preferred.
Overall, these results suggest that the TB selector that was used is a very stiff chiral molecule
which hardly undergoes “induced fit” phenomena upon interactions with other molecules. It is
in line with the rigidity of the molecule observed by VCD-analysis of 73o. Nevertheless, an
interesting self-recognition process could be noticed which is probably general for all TB
molecules. Further studies are performed to ascertain this observation.
.
-100-
Chapter 6
6. Conclusion and Perspectives
6.1
Conclusions
In conclusion, we have developed a methodology which can provide a quick access to highly
enantioenriched configurationally-stable ethano-Tröger derivatives using one step protocol.
The process is general, enantiospecific (ee up to 99%, retention of configuration),
diastereoselective (quaternary carbon center introduction, dr up to 49:1) and it presents an
interesting regioselectivity.
A mechanistic rationale involving iminium/enolate species as key intermediates has also been
proposed by studying the effect on enantiospecificity of various substituents on the diazocompounds. A brief study on the catalytic effect of different metal sources was also presented
which highlighted the dichotomous behavior of copper and dirhodium based catalysts.
We have also described the use of ethano-TBs for asymmetric aziridination of chalcone
derivatives. Although moderate levels of selectivity were observed, it was shown that these
compounds are configurationally stable in the reaction unlike their methano-precursors and
they can be recovered (>90%) at the end. Further, we have also developed a new chiral
stationary phase (TB-CSP) using a designed ethano-TB. Though poor resolution was
observed for most chiral analytes, the study highlighted an interesting self-recognition hetero
chiral process which is probably general for all TB molecules.
6.2
Perspectives
The major drawback of the methodology described above is the need for enantiopure starting
material. Resolution of methano-TB is limited to a few examples by traditional separation
methods. For instance, the process works poorly with halogenated TBs which are able to
undergo cross-coupling reactions essential to the development of applications in
supramolecular chemistry. These issues still need to be addressed.
-101-
Conclusion and Perspectives
One of the obvious solution to this problem is to develop an enantioselective process, which
could provide enantioenriched ethano-TBs starting from racemic-TBs. This rather difficult
transformation was mentioned in chapter 2. To best of knowledge, it has never been reported
on racemic tertiary amines. To achieve this transformation with rac-1 specifically, several
chiral rhodium and copper catalysts were tested. So far, after initial screening and
optimization of reaction conditions, we are able to achieve up to 30% ee of product 73m,
starting from rac-1 and using Pirrung’s catalyst (Equation 6-1). The result, although not
synthetically useful so far, still represents a first evidence that the enantioselective
transformation is possible and could be achieved with higher selectivity, with the right
catalyst.
Equation 6-1
In terms of applications of these scaffolds, few examples of their possible applications were
shown in this thesis, including, a new first generation of novel chiral stationary phase (TBCSP) using ethano-Tröger bases as chiral selector.1 Initial results were bit discouraging as the
TB-CSP hardly separated any chiral analyte. To achieve better separation, we will introduce
electron-poor or electron-rich substituents on the aromatic rings. We will also make CSP
containing free carboxylic acid moieties rather than ester group.
Finally, we believe that compounds of type 75 (ee 97%) are an interesting entry point for
further applications and that in the field of supramolecular chemistry in particular. Compound
75 can selectively deprotected to yield a diphenolic moiety in good yield and
enantioselectivity as shown in.
1
This work was performed in collaboration with the group of Prof Dr Wolfgang Lindner (Universität Wien) and
Dr Eric Francotte (Novartis Institutes for BioMedical Research).
-102-
Chapter 6
Equation 6-2
These phenolic moieties can be converted into the corresponding triflate functional groups.
This TB- 91 can then acts as a precursor for cross-coupling reactions and hence it opens the
doors for the mentioned applications.
-103-
Chapter 7
7. Experimental Part
7.1
Generalities
All reactions were carried out under dry N2 or Ar by means of an inert gas/vacuum double
manifold line and standard Schlenk techniques with magnetic stirring, unless otherwise
stated. Dry toluene, dichloromethane, hexane and tetrahydrofuran were obtained by filtration
through appropriate activated-alumina drying columns. Analytical thin-layer chromatography
(TLC) was performed with Merck SIL G/UV254 plates. Unless otherwise stated, column
chromatography (Fluka silicagel 60, 40 µm or Fluka basic alumina type 5016A) was
performed in air and under pressure (0.1-0.3 bar). NMR spectra were recorded on Bruker
ARX-300 or AMX-400 or ARX-500 at room temperature unless otherwise stated. 1H-NMR:
chemical shifts are given in ppm relative to Me4Si with the solvent resonance used as the
internal standard. 31P-NMR: chemical shifts are reported in ppm relative to H3PO4. 13C-NMR:
chemical shifts are given in ppm relative to Me4Si, with the solvent resonance used as the
internal standard. Assignments may have been achieved using DEPT, COSY, HSQC, HMBC
and/or NOESY experiments. IR spectra were recorded with a Perkin-Elmer 1650 FT-IR
spectrometer using a diamond ATR Golden Gate sampling. Melting points (M.p.) were
measured in open capillary tubes on a Stuart Scientific SMP3 melting point apparatus and are
uncorrected. React-IR™ 45m instrument was used for in-situ measurement in a special
Schlenk tube fitted with an immersible in-situ FT-IR DiComp probe. Electrospray mass
spectra (ESI-MS) were obtained on a Finnigan SSQ 7000 spectrometer. GC-MS were
obtained spectra on a Hewlett Packard 6890 GC chromatograph coupled with an Agilent
5973 Network mass-selective detector. HPLC analyses were performed on an Agilent 1100
apparatus (binary pump, auto-sampler, column thermostat and diode array detector). Unless
otherwise stated, all chemicals were obtained from Fluka, Aldrich, Pressure Chemicals or
Acros and used as received or purified according to standard literature procedures.
148
D. D. Perrin, W. L. F. Armarego, Purification of Laboratory Chemicals, 3
1988.
-105-
rd
148
ed., Pergamon Press: Oxford,
Experimental Part
7.2
General Procedure for the Alkylation of rac-1
4 mmol (1 g) of rac-Tröger base (rac-1) and 4 mmol of 2-bromo-1-phenylethan-1-one were
dissolved in benzene (40 ml). The resulting solution was refluxed for 12 h under dinitrogen
atmosphere and then left to cool to room temperature. The solvent was removed under
reduced pressure. The resulting solid was then dissolved in CH2Cl2 and purified by selective
precipitation with n-Hexane (sequence repeated twice) to afford the desired bromide salts in
good to moderate yields.
rac-2,8-Dimethyl-5-(2-oxo-2-phenyl-ethyl)-6H,12H-5,11-methano-dibenzo[b,f][1,5]
diazocin-5-ium bromide (62)
1.57 g (87% yield) of 62 was obtained starting from 1 g of
rac-1. 1H NMR (CDCl3, 400 MHz):  = 8.30 (d, J = 7.9 Hz,
1H), 8.1610 (d, 3J = 8.6 Hz, 1H), 7.7014 (d, 2J = 18.6 Hz, 1H),
7.51 (t, J = 7.3 Hz, 1H), 7.39 – 7.29 (m, 3H), 7.16 – 7.033,4,9
(m, 3H), 6.837 (s, 1H), 6.6814 (d, J = 18.6 Hz, 1H), 6.671 (s,
1H), 6.4012 (d, J = 15.2 Hz, 1H), 5.9413 (d, J = 10.7 Hz, 1H), 5.7213 (d, J = 10.4 Hz, 1H),
5.5612 (d, J = 15.1 Hz, 1H), 4.956 (d, J = 16.8 Hz, 1H), 4.246 (d, J = 16.8 Hz, 1H), 2.19 (s,
2H), 2.13 (s, 2H). 13C NMR (CDCl3, 100 MHz): = 191.8 (CO), 140.5, 140.3, 138.5, 135.7,
134.9, 134.8, 131.0, 130.5, 129.4 (2C), 129.2 (2C), 128.3, 127.7, 127.6, 124.3, 122.1, 120.610,
77.513, 68.512, 66.114, 57.86, 21.2, 21.0. M.P.: 182 °C. IR (neat): 1693 cm-1. HRMS (ESI+):
calculated for C25H25N2O 369.1961, found 369.1950.
-106-
Chapter 7
rac-2,8-Dimethyl-5-(2-oxo-2-p-tolyl-ethyl)-6H,12H-5,11-methano-dibenzo[b,f][1,5]
diazocin-5-ium bromide (62a)
0.18 g (49% yield) of 62a was obtained starting from 0.20 g
rac-1. 1H NMR (CDCl3, 400 MHz):  = 8.22 (d, 2H, 3J =
7.6), 8.00 (d, 1H, 3J = 8.2 Hz), 7.41 (m, 1H), 7.15 (m, 2H),
6.92 (d, 1H, 3J = 7.3 Hz), 6.35 (m, 2H), 6.01 (d, 1H, 2J =
10.6 Hz), 5.45 (d, 1H, 2J = 10.2 Hz), 5.01 (d, 1H, 2J = 17.0 Hz), 4.31 (d, 1H, 2J = 17.0 Hz),
2.40 (s, 3H), 2.27 (s, 3H), 2.24 (s, 3H). 13C NMR (CDCl3, 100 MHz): = 190.9, 145.9, 140.3,
140.1, 138.3, 135.5, 132.1, 130.8, 130.2, 129.9 (2C), 129.4, 129.2 (2C), 128.4, 128.1, 127.7,
127.5, 127.4, 77.3, 68.4, 65.7, 57.5, 21.9, 21.0, 20.8. M.P.: 157 °C. IR (neat): 1683 cm-1.
HRMS (ESI+): calculated for C26H27N2O 383.2117, found 383.2114.
rac-5-[2-(4-Methoxy-phenyl)-2-oxo-ethyl]-2,8-dimethyl-6H,12H-5,11-methano-dibenzo
[b,f][1,5]diazocin-5-ium bromide (62b)
0.19 g (49% yield) of 62b was obtained starting from 0.20
g rac-1. 1H NMR (CDCl3, 400 MHz):  = 8.35 (d, 2H, 3J
= 8.9 Hz), 8.03 (d, 1H, 3J = 8.6 Hz), 7.48 (d, 1H, 3J =
18.4 Hz), 7.15 (m, 2H), 6.93 (t, 3H, 3J = 9.3 Hz), 6.30 (d,
1H, 2J = 18.7 Hz), 6.23 (d, 1H, 2J = 16.2 Hz), 6.07 (d, 1H, 2J = 10.5 Hz), 5.46 (d, 1H, 2J =
3.84 (s, 3H), 2.27 (s, 3H), 2.25 (s, 3H).
13
C NMR (CDCl3, 100 MHz):  = 189.9, 164.9,
140.6, 140.3, 138.4, 135.8 , 131.8 (2C), 131.0, 130.4, 128.2, 127.8, 127.7, 127.6, 124.3,
122.1, 120.6, 114.6 (2C), 77.5, 68.5, 65.7, 57.7, 55.7, 21.2, 21.0. M.P.: 166 °C. IR (neat):
1681 cm-1. HRMS (ESI+): calculated for C26H27N2O2 399.2067, found 399.2058.
rac-5-[2-(4-Fluoro-phenyl)-2-oxo-ethyl]-2,8-dimethyl-6H,12H-5,11-methano-dibenzo
[b,f][1,5]diazocin-5-ium bromide (62c)
0.20 g (54% yield) of 62c was obtained starting from 0.20 g
rac-1. 1H NMR (CDCl3, 400 MHz):  = 8.45 (m, 2H), 8.14
(d, 1H, 3J = 8.6 Hz), 7.75 (d, 1H, 2J = 17.9 Hz), 7.18 (m,
3H), 7.08 (t, 2H, 3J = 8.2 Hz), 6.90 (s, 1H), 6.81 (s, 1H), 6.65
-107-
Experimental Part
(d, 1H, 2J = 18.2 Hz), 6.38 (d, 1H, 2J = 16.0 Hz), 5.99 (d, 1H, 2J = 10.6 Hz), 5.68 (d, 1H, 2J =
2.28 (s, 3H), 2.26 (s, 3H).
13
C NMR (CDCl3, 100 MHz):  = 190.2, 140.4, 140.3, 138.3,
135.9, 132.4, 132.2, 131.2, 131.0, 130.6, 128.3, 127.7, 127.6, 124.3, 121.9, 120.5, 116.8,
116.8, 116.5, 77.4, 68.7, 66.2, 57.7, 21.2, 20.9. M.P.: 168 °C. IR (neat): 1694 cm-1. HRMS
(ESI+): calculated for C25H24N2OF 387.1867, found 387.1858.
rac-5-[2-(2-Bromo-phenyl)-2-oxo-ethyl]-2,8-dimethyl-6H,12H-5,11-methano-dibenzo
[b,f][1,5]diazocin-5-ium hexafluorophosphate (62d)
An additional metathesis step was performed using 1 eq.
NH4PF6 in MeOH. The resulting solid obtained after
evaporation and recrystalization in CH2Cl2 / Et2O the titled
compounds 62d as beige solid. 0.31 g (66% yield) of 62d was
obtained starting from 0.20 g rac-1. 1H NMR (CDCl3, 400
MHz):  = 7.88 (d, 1H, 3J = 7.8 Hz), 7.80 (d, 1H, 3J = 7.6 Hz), 7.55 (m, 2H), 7.47 (t, 1H, 3J =
7.6 Hz), 7.20 (m, 3H), 6.98 (s, 1H), 6.91 (s, 1H), 6.13 (d, 1H, 2J = 18.4 Hz), 5.99 (d, 1H, 2J =
4.95 (m, 2H), 4.34 (d, 1H, 2J = 16.7 Hz), 2.35 (s, 3H), 2.31 (s, 3H).
13
C NMR (CDCl3, 100
MHz): = 191.9, 141.2, 139.9, 137.4, 136.6, 135.7, 134.9, 134.3, 131.1, 131.0, 130.4, 128.9,
128.4, 127.8, 127.6, 124.3, 120.8, 119.9, 119.3, 78.3, 68.6, 66.1, 57.4, 21.2, 21.1. M.P.: 274
°C. IR (neat): 1722 cm-1. HRMS (ESI+): calculated for C25H24N2OBr 447.1004, found
447.1006.
rac-2,3,8,9-Tetramethoxy-5-(2-oxo-2-phenyl-ethyl)-6H,12H-5,11-methano-dibenzo[b,f]
[1,5]diazocin-5-ium bromide (62e)
The starting rac-Tröger base was prepared according to a
literature procedure.149 0.59 g (57% yield) of 62e was
obtained starting from 0.66 g rac-1. 1H NMR (CDCl3, 400
MHz): = 8.45 (d, 2H, 3J = 7.3 Hz), 8.05 (s, 1H), 7.63 (t,
1H, 3J = 7.4 Hz), 7.46 (t, 2H, 3J = 7.8 Hz), 6.77 (s, 1H), 6.50
149
X.-H. Bu, M. Du, L.-J. Zhao, K. Tanaka, M. Shionoya, and M. Shiro, J. Chem. Research (S), 2001, 243-245.
-108-
Chapter 7
(s, 1H), 6.39 (s, 1H), 6.15 (d, 1H, 2J = 10.6 Hz), 5.98 (d, 1H, 2J = 11.6 Hz), 5.94 (d, 1H, 3J =
4.26 (d, 2H, 2J = 16.7 Hz), 3.94 (s, 3H), 3.88 (s, 3H), 3.83 (s, 3H), 3.65 (s, 3H).
13
C NMR
(CDCl3, 100 MHz): = 191.9, 150.3, 150.1, 149.9, 135.8, 135.0, 134.9, 132.4, 129.6 (2C),
129.5 (2C), 119.8, 113.9, 109.2, 108.8, 108.6, 107.1, 104.1, 77.8, 67.6, 65.3, 58.2, 57.2, 56.4,
56.3, 56.1. M.P.: 161 °C. IR (neat): 1699 cm-1. HRMS (ESI+): calculated for C27H29N2O5
461.2070, found 461.2055.
(5R,11R)-2-methoxy-5-methyl-8-nitro-6,12-dihydro-5H-5,11-methanodibenzo[b,f][1,5]
diazocine-5-ium iodide (85)
0.045 mmol (20 mg) of rac-TB-79 was dissolved in methyl iodide (2 ml). The resulting
solution was heated at 40 oC under dinitrogen atmosphere and then left to cool to room
temperature. The solvent was removed under reduced pressure. The resulting solid was then
dissolved in CH2Cl2 and purified by selective precipitation with n-Hexane (sequence repeated
twice) to afford the desired iodide salts 85 in 70% yields.
1
H NMR (500 MHz, acetone-D6):  = 8.21 (dd, J = 8.9, 2.5 Hz, 1H),
8.14 (d, J = 8.9 Hz, 1H), 8.13 (s, 1H), 7.06 (dd, J = 9.3, 2.9 Hz, 1H),
6.84 (d, J = 2.9 Hz, 1H), 5.66 (dd, J = 18.6, 8.9 Hz, 2H), 5.43 – 5.32
(m, 2H), 5.10 (d, J = 17.2 Hz, 1H), 4.65 (d, J = 17.0 Hz, 1H), 4.02 (s,
3H), 3.78 (s, 3H).13C NMR (125 MHz, acetone-D6):  = 162.2, 150.6,
146.7, 135.1, 131.4, 127.7, 126.0, 125.7, 124.9, 124.8, 117.4, 113.4, 78.2, 68.8, 59.0, 57.0,
54.2, 53.1. M.P: 160-162 °C. IR (neat): -1 = 2962, 1432, 1302, cm-1.
-109-
Experimental Part
7.3
General procedure for the rearrangement of ammonium
salts of rac-1
7.3.1 Basic alumina (Al2O3) assisted rearrangement
1.1 mmol of rac Tröger salt (typically 0.50 g of [62][Br]) was dissolved in CHCl3 (60 ml,
p.a) and 20 g of basic alumina (Brockmann activity I, pH 9.5 ±0.5) were added. The reaction
was stirred at room temperature for 2.5 h. The reaction mixture was then filtration over a
cotton plug; the remaining alumina was washed with 50 ml of a 9:1 mixture of EtOAc and
MeOH. Evaporation of solvents and drying under vacuum led to the desired product (64)
with a satisfactory purity for next steps. Additional purification can be performed by column
chromatography on silica gel (eluent: hexanes/acetone=8:2) to give rac-64 as a white solid
(350 mg, 85%).
7.3.2 Hexamine assisted rearrangement
p.a) and Hexamine (2 equivalent) were added. The reaction was stirred at room temperature
for 6 h. Then 10 g of basic alumina (Brockmann activity I, pH 9.5 ±0.5) was added. The
reaction was stirred at room temperature for 0.5 h. The reaction mixture was then filtration
over a cotton plug; the remaining alumina was washed with 30 ml of a 9:1 mixture of EtOAc
and MeOH. Evaporation of solvents and drying under vacuum led to the crude product which
was then purified by column chromatography on silica gel (eluent: hexanes/acetone=8:2) to
give rac-64 as a white solid (375 mg, 91%).
-110-
Chapter 7
7.3.3 Acid/Alcohol assisted rearrangement
p.a) and 1 equivalent of acid/alcohol were added. The reaction was then heated at 60 °C for
12 h. The reaction mixture was then quenched using 10% NaHCO3 solution followed by
extraction. Evaporation of solvents and drying under vacuum led to the desired product (64)
with a satisfactory purity for next steps. Additional purification can be performed by column
chromatography on silica gel (eluent: hexanes/acetone=8:2) to give rac-64 as a white solid
(in 92 to 98% yield).
rac 2,8-dimethyl-6H,12H-5,11-[(R)-benzoyl-ethano]-dibenzo[b,f][1,5]diazocine (64)
Using protocol given in section 7.3.1, starting from 0.50 g of
salt, 0.35 g of product was obtained in 85% yield. 1H NMR
(CDCl3, 400 MHz):  8.05 (d, 2H, 3J = 7.8 Hz), 7.54 (m, 1H, 3J
= 6.9 Hz), 7.44 (t, 2H, 3J = 7.6 Hz), 6.99 (m, 2H), 6.94 (d, 1H,
3
J = 8.1 Hz), 6.87 (d, 1H, 3J = 8.1 Hz), 6.79 (s, 1H), 6.62 (s,
1H), 5.02 (m, 1H), 4.65 (d, 1H, 2J = 17.2 Hz), 4.45 (m, 2H),
4.07 (m, 2H), 3.76 (m, 2H), 2.22 (s, 3H), 2.16 (s, 3H). 13C NMR (CDCl3, 100 MHz): 198.1,
147.3, 146.7, 137.3, 136.3, 136.1, 135.1, 134.5, 133.2, 129.9, 129.5 (2C), 129.1, 128.6 (2C),
128.5, 128.1 (2C), 127.8, 66.71, 59.49, 55.85, 54.87, 21.1, 20.9. M.P.: 201 °C. IR (neat):
1682 cm-1. HRMS (ESI+): calculated for (C25H24N2O + 1H) 369.1961, found 369.1958.
((5S,11S,14S)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14yl)(phenyl)methanone (epi-64)
Using protocol given in section 7.3.2, starting from 0.50 g of salt,
61.7 mg of product epi-64 was obtained in 15% yield. 1H NMR
(CDCl3, 400 MHz): = 7.97 (m, 2H), 7.46 (m, 1H), 7.34 (m, 2H),
7.05 (d, 1H, 3J = 8.08 Hz ), 6.90 (d, 1 H, 3J = 7.84 Hz), 6.73 (s, 1 H),
6.51 (m, 3H), 4.99 (d, 1 H, 3J = 7.32 Hz), 4.88 (d, 1 H, 2J = 17.44
Hz), 4.73 (d, 1H, 2J = 17.16 Hz), 4.56 (m, 2 H), 4.30 (d, 1H, 2J =
17.16 Hz), 3.68 (m, 1H), 2.19 (s, 3 H), 2.04 (s, 3 H).
13
C NMR (CDCl3, 100 MHz): =
197.9, 147.7, 142.7, 137.2, 136.9, 136.6, 135.1, 134.6, 132.6, 130.1, 129.1 (2C), 128.8, 128.7,
-111-
Experimental Part
128.6, 128.2, 128.0 (2C), 127.8, 71.7, 61.4, 60.4, 53.4, 20.90, 20.88. IR (neat): 1675 cm-1.
HRMS (ESI+): calculated for (C25H24N2O + 1H) 369.1961, found 369.1952.
Characteristic peak for
minor diastereomer epi-64
1
7
14 6 12 13,6 12
13
Characteristic peak for
major diastereomer 64
1
10
8.5
8.0
14
3,4,9
7.5
7.0
7
14
7 1,14
6.5
12
13
6.0
13 12
12 12,6
6
5.5
5.0
6,13
13
6
4.5
4.0
3.5
3.0
2.5
2-(2,8-Dimethyl-11,12-dihydro-6H-dibenzo[b,f][1,5]diazocin-5-yl)-1-phenyl-ethanone
(rac-63)
Yellowish solid. 1H NMR (CDCl3, 400 MHz):  = 7.98 (d, 2H, 3J
= 8.5 Hz), 7.62 (t, 1H, 3J = 7.4 Hz), 7.49 (t, 2H, 3J = 7.5 Hz), 6.92
(m, 3H), 6.85 (s, 1H), 6.60 (dd, 2H, 3J = 3.6 Hz), 4.58 (s, 2H),
4.55 (s, 2H), 4.50 (s, 2H), 2.28 (s, 3H), 2.25 (s, 3H).
13
C NMR
(CDCl3, 100 MHz): = 197.7, 146.9, 145.7, 135.7, 133.6, 132.7, 131.6, 129.2, 129.0, 128.9
(2C), 128.8 (2C), 128.6, 128.2 (2C), 127.8, 123.7, 117.4, 58.2, 55.7, 51.0, 20.6, 20.5. ). IR
(neat): 1725 cm-1. M.P.: 88 °C. HRMS (ESI+): calculated for (C24H25N2O + 1H) 357.1961,
found 357.1953.
-112-
Chapter 7
Rac-2,8-dimethyl-6H,12H-5,11-[(R)-(4-methyl-benzoyl)-ethano]-dibenzo[b,f][1,5]
diazocine (64a)
Using protocol given in section 7.3.1, starting from 0.1
mmol of salt, 19.5 mg of orange oil 64a was obtained
in 51% yield. 1H NMR (CDCl3, 400 MHz): 7.90
(d, 2H, 3J = 8.1 Hz), 7.27 (m, 2H), 7.04 (m, 2H), 6.96
(d, 1H, 3J = 8.1 Hz), (d, 1H, 3J = 7.8 Hz), 6.81 (s, 1H),
6.64 (s, 1H), 5.01 (m, 1H), 4.67 (d, 1H, 2J = 17.2 Hz),
4.50 (d, 2H, 2J = 17.7 Hz), 4.11 (m, 2H), 3.75 (m, 1H), 2.43 (s, 3H), 2.24 (s, 3H), 2.17 (s,
3H).
13
C NMR (CDCl3, 100 MHz):  = 197.7, 147.3, 146.8, 144.0, 137.3, 136.4, 135.1,
134.5, 133.5, 129.8, 129.6 (2C), 129.3 (2C), 129.1, 128.5, 128.2, 128.1, 127.8, 66.6, 59.5,
55.9, 54.9, 21.9, 21.1, 21.0. IR (neat): 1677 cm-1. HRMS (ESI+): calculated for (C26H26N2O
+ 1H) 383.2117, found 383.2105.
rac-2,8-dimethyl-6H,12H-5,11-[(R)-(4-methoxy-benzoyl)-ethano]-dibenzo[b,f][1,5]
diazocine (64b)
Using protocol given in section 7.3.1, starting from 0.1
mmol of salt, 13.5 mg of product was obtained in 34%
yield. 1H NMR (CDCl3, 400 MHz): = 8.07 (d, 2H, 3J =
8.8 Hz), 7.08 (m, 2H), 6.98 (m, 4H), 6.84 (s,1H), 6.68 (s,
1H), 5.02 (m, 1H), 4.70 (d, 1H, 2J = 17.2 Hz), 4.54 (d, 2H,
2
J = 17.2 Hz), 4.14 (m, 2H), 3.92 (s, 3H), 3.78 (m, 1H), 2.28 (s, 3H), 2.21 (s, 3H). 13C NMR
(CDCl3, 100 MHz): = 196.6, 163.6, 147.3, 146.8, 137.3, 136.4, 135.1, 134.5, 131.8 (2C),
129.8, 129.1, 129.0, 128.5, 128.1, 128.0, 127.7, 113.7 (2C), 66.5, 59.5, 55.9, 55.7, 54.9, 21.1,
21.0. IR (neat): 1673 cm-1. HRMS (ESI+): calculated for (C26H26N2O2 + 1H) 399.2067,
found 399.2061.
-113-
Experimental Part
rac-2,8-dimethyl-6H,12H-5,11-[(R)-(4-fluoro-benzoyl)-ethano]-dibenzo[b,f][1,5]
diazocine (64c)
Using protocol given in section 7.3.1, starting from 0.1 mmol
of salt, 25.1 mg of product was obtained in 65% yield. 1H
NMR (CDCl3, 400 MHz): = 8.11 (dd, 2H, 3J = 5.6 Hz),
7.15 (t, 2H, 3J = 8.6 Hz), 7.03 (m, 3H), 6.93 (d, 1H, 3J = 7.8
Hz), 6.86 (s, 1H), 6.68 (s, 1H), 5.01 (m, 1H), 4.69 (d, 1H, 2J =
17.4 Hz), 4.54 (d, 2H, 3J = 6.3 Hz), 4.47 (d, 2H, 3J = 6.3 Hz),
4.15 (m, 2H), 3.80 (m, 2H), 2.28 (s, 3H), 2.21 (s, 3H).
13
196.6, 147.3, 146.5, 137.3, 136.1, 135.3, 134.5, 132.4 (2C), 132.2, 132.1, 129.9, 129.1, 128.6,
128.2, 128.0, 127.8, 115.7, 115.5, 66.8, 59.5, 55.9, 54.7, 21.1, 21.0. M.P.: 200 °C. HRMS
(ESI+): calculated for (C25H23N2OF 387.1867, found 387.1858. IR (neat): 1682 cm-1.
rac-2,8-dimethyl-6H,12H-5,11-[(R)-(2-bromo-benzoyl)-ethano]-dibenzo[b,f][1,5]
diazocine (64d)
Using protocol given in section 7.3.1, starting from 0.1 mmol of
salt, 18 mg of product was obtained in40% yield. 1H NMR
(CDCl3, 400 MHz): = 7.70 (d, 1H, 3J = 7.5 Hz), 7.50 (d, 1H,
3
J = 7.6 Hz), 7.37 (m, 2H), 7.06 (d, 1H, 3J = 10.6 Hz), 6.91 (m,
3H), 6.76 (s, 1H), 6.67 (s, 1H), 4.97 (m, 1H), 4.64 (d, 1H, 2J =
17.3 Hz), 4.48 (d, 2H, 2J = 17.1 Hz), 4.13 (d, 1H, 2J = 17.5 Hz), 3.97 (m, 2H), 2.22 (s, 3H),
2.20 (s, 3H).
13
C NMR (CDCl3, 100 MHz): = 203.7, 147.0, 146.7, 140.9, 136.6, 136.2,
135.0, 134.6, 133.6, 131.6, 129.5, 129.4, 129.1, 128.3, 128.2, 128.1, 127.9, 127.1, 119.6,
70.1, 59.4, 56.3, 55.7, 21.1, 21.0. IR (neat): 1699 cm-1. HRMS (ESI+): calculated for
(C25H23N2OBr + 1H) 447.1066, found 447.1054.
-114-
Chapter 7
rac-2,3,8,9-tetramethoxy-6H,12H-5,11-[(R)-benzoyl-ethano]-dibenzo[b,f][1,5]diazocine
(64e)
Using protocol given in section 7.3.1, starting from
0.1 mmol of salt, 16.2 mg of product was obtained in
35% yield. 1H NMR (CDCl3, 400 MHz): = 8.07 (d,
2H, 3J = 7.7), 7.62 (t, 1H, 3J = 7.3), 7.50 (t, 2H, 3J =
7.7 Hz), 6.75 (s, 1H), 6.70 (s, 1H), 6.53 (s, 1H), 6.36
(s, 1H), 5.11 (dd, 1H, 3J = 6.3 Hz), 4.70 (d, 1H, 2J =
17.2 Hz), 4.54 (m, 2H), 4.15 (m, 2H), 3.88 (s, 3H),
3.86 (s, 6H), 3.84 (m, 1H) 3.79 (s, 3H).
13
C NMR (CDCl3, 100 MHz):  = 197.8, 148.1,
147.8, 146.8, 146.5, 142.6, 142.1, 136.0, 133.3, 129.4 (2C), 128.7, 128.6 (2C), 127.8, 111.4,
111.1 (2C), 110.6, 67.0, 59.1, 56.2, 56.1, 56.0 (2C), 55.5, 55.4. IR (neat): 1682 cm-1. HRMS
(ESI+): calculated for (C27H29N2O5 + 1H) 461.2070, found 461.2086.
7.3.4 Reduction of ethano-Tröger bases
Rac-2,8-dimethyl-6H,12H-5,11-[(R)-((S)-hydroxyl-phenyl-methyl)-ethano]-dibenzo[b,f]
[1,5]diazocine (65)
(0.368 g, 1.0 mmol) of compound 64 was dissolved in 40 ml of
EtOH (p.a) and cooled to 0 °C. NaBH4 (78 mg, 2 mmol) was
added and the reaction was stirred at 0 °C. After 2.5 h reaction, an
additional amount of NaBH4 (39 mg, 1 mmol) was added and the
reaction mixture was stirred for 4.5 h at 0 °C. Workup involved
evaporation of solvent followed by extraction using EtOAc and brine. The resulting organic
phase was dried over Na2SO4, concentrated under vacuum and further purified by flash
chromatography over a pad of silica using EtOAc. Evaporation of solvent under reduced
pressure leads to the desired alcohol. For smaller scale, purification can be performed by
preparative TLC using silica plates and (8:1) n-hexane/acetone mixture followed by
extraction with EtOAc and evaporation. White solid, 0.36 g, 97% yield. 1H NMR (CDCl3,
400 MHz): = 7.28 (m, 5H), 7.15 (d, 1H, 3J = 8.1 Hz), 6.96 (d, 1H, 3J = 7.8 Hz), 6.88 (d,
1H, 3J = 8.1 Hz), 6.73 (d, 1H, 3J = 13.6 Hz), 4.82 (d, 1H, 2J = 17.4 Hz), 4.62 (d, 1H, 3J = 9.6
-115-
Experimental Part
Hz), 4.48 (bs, 1H), 4.43 (d, 1H, 2J = 17.4 Hz), 4.31 (d, 1H, 2J = 14.1 Hz), 4.26 (d, 1H, 2J =
14.4 Hz), 3.60 (m, 1H), 3.30 (t, 1H, 3J = 13.1 Hz), 3.16 (dd, 1H, 3J = 5.8 Hz), 2.21 (s, 3H),
2.20 (s, 3H).
13
C NMR (CDCl3, 100 MHz): = 147.4, 147.0, 141.1, 136.9, 135.3, 135.2,
134.7, 129.6, 129.2, 128.7 (2C), 128.4, 128.3, 128.2, 128.1, 128.0, 127.0 (2C), 73.3, 70.3,
59.4, 56.1, 53.8, 21.1, 21.0. M.P.:115 °C. IR (neat): 2914, 1437 cm-1. HRMS (ESI+)
calculated for (C25H26N2O + 1H) 371.2117, found 371.2116.
7.3.5 Nucleophile addition on 64
General procedure for the addition of nucleophiles to 64:
Compound 64 (36.8 mg, 0.1 mmol) was dissolved in 10 ml of dry THF under dinitrogen
atmosphere. After cooling the resulting solution to –78 °C, organolithium reagent RLi(1.5 eq.,
0.15 mmol) was added drop-wise. The reaction mixture was then stirred at –78 °C for 2 h and
then left to warm to 0 °C and quenched with water. Workup involved evaporation of solvent
followed by extraction using EtOAc and brine. The resulting organic phases were collected
together and dried over Na2SO4, concentrated under vacuum and further purified by flash
chromatography over silica using n-Hexane/EtOAc (8:2) mixtures as eluent.
1-((5S,11S,14R)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14-yl)1-phenylethanol (65a)
Using protocol given in described in section 7.3.5, starting from 36
mg of 64, 34 mg of product 65a was obtained in 91% yield. 1H
NMR (CDCl3, 400 MHz): = 7.46(m, 2H), 7.27 (m, 3H), 7.0 (d,
1H, 3J = 7.84 Hz), 6.96 (d, 1H, 3J = 7.84 Hz), 6.87 (m, 2H), 6.64
(s, 2H), 4.76 (d, 1H, 2J = 17.44 Hz), 4.36 (d, 1H, 2J = 17.16 Hz),
4.25 (d, 1H, 2J = 17.16 Hz), 4.17 (d, 1H, 2J = 17.44 Hz), 4.07 (b, 1H), 3.55 (m, 3H), 2.16 (s,
6H), 1.7 (s, 3H).
13
C NMR (CDCl3, 100 MHz): = 148.6, 147.4, 147.3, 136.8, 135.9, 134.9,
134.6, 129.8, 129.0, 128.4, 128.37 (2C), 128.3, 127.9, 127.8, 127.2, 125.9 (2C), 74.8, 72.2,
59.3, 56.5, 53.9, 26.0, 21.12, 21.1. IR (neat): 3375, 1493, 1260, 1018 cm-1. HRMS (ESI+)
calculated for (C26H28N2O + 1H) 385.2274, found 385.2258.
-116-
Chapter 7
(4-bromophenyl)((5S,11S,14R)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]
diazocin-14-yl)(phenyl)methanol (65b)
Using protocol given in described in section 7.3.5, starting
from 36 mg of 64, 45.5 mg of product 65b was obtained in
87% yield.1H NMR (CDCl3, 400 MHz): = 7.39 (m,
4H), 7.27 (m, 5H), 7.0 (d, 1H, 3J = 8.1 Hz), 6.93 (m, 2H),
6.83 (dd, 1H, 3J = 8.08, 4J = 1.76 Hz), 6.73 (d, 1H, 4J =
1.76), 6.54 (s, 1H), 4.61 (d, 1H, 2J = 17.2 Hz), 4.44 (m, 4H), 4.0 (d, 1H, 2J = 17.44 Hz), 3.53
(m, 2H), 2.19 (s, 3H), 2.12 (s, 3H).
13
C NMR (CDCl3, 100 MHz): = 148.2, 147.2 146.3,
144.9, 136.9, 135.9, 135.1, 134.5, 131.4(2C), 129.9, 129.0, 128.8(2C), 128.6, 128.5(2C),
128.2, 128.0, 127.7, 127.6, 127.5(2C), 121.3, 79.0, 67.9, 59.3, 56.1, 55.0, 21.2, 21.1. IR
(neat): 3376, 1494 cm-1. HRMS (ESI+): calculated for (C31H29BrN2O + 1H) 525.1536, found
525.1582.
3-((5S,11S,14R)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14-yl)3-hydroxy-3-phenylpropanenitrile (65c)
Using protocol given in described in section 7.3.5, starting from
36 mg of 64, 36 mg of product 65c was obtained in 90% yield. 1H
NMR (CDCl3, 400 MHz): = 7.5 (d, 2H, 3J = 7.32 Hz), 7.38(m,
3H), 6.95 (m, 2H), 6.87 (d, 1H, 3J = 7.56 Hz), 6.82 (d, 1H, 3J =
7.84 Hz), 6.67 (s, 1H), 6.48 (s, 1H), 4.49 (d, 1H, 2J = 17.4), 4.29
(d, 1H, 2J = 17.2 Hz), 3.88 (m, 4H), 3.50 (d, 1H, 2J = 16.44 Hz), 3.38 (t, 1H, 2J = 12.88 Hz),
3.06 (d, 1H, 2J = 16.44 Hz), 2.16 (s, 3H), 2.10 (s, 3H).
13
148.6, 147.1, 141.7, 136.5, 135.6, 135.1, 134.7, 130.0, 129.0(2C), 128.9, 128.8, 128.5, 128.4,
127.7, 126.4(2C), 117.9, 77.5, 76.9, 70.2, 59.2, 56.1, 53.4, 30.3, 21.1, 21.0. IR (neat): 3437,
2248, 1494, 1259, 1017 cm-1. HRMS (ESI+): calculated for (C27H27N3O + 1H) 410.2226,
found 410.2233.
-117-
Experimental Part
((5S,11S,14R)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14yl)(phenyl)(pyridin-2-yl)methanol (65d)
Using protocol given in described in section 7.3.5, starting
from 36 mg of 64, 38.9 mg of product 65d was obtained in
89% yield. 1H NMR (CDCl3, 400 MHz): = 8.46 (d, 1H, 3J =
4.8 Hz), 7.59 (dt, 1H, 3J = 7.56, 4J = 1.52 Hz), 7.53 (m, 2H),
7.26 (m, 3H), 7.15 (m, 1H), 6.87 (m, 4H), 6.72 (s, 1H), 6.65
(s, 1H), 6.60 (s, 1H), 5.61 (d, 1H, 2J = 17.68 Hz), 4.49 (d, 1H, 2J = 16.92 Hz), 4.43 (dd, 1H,
3
J = 5.56, 3J = 11.6 Hz), 4.36 (d, 1H, 2J = 16.92 Hz), 3.99 (d, 1H, 2J = 17.68 Hz), 3.84 (dd,
1H, 3J = 11.6, 2J =14.4 Hz), 2.92 (dd, 1H, 3J = 5.56, 2J = 14.4 Hz), 2.21 (s, 3H), 2.13 (s,
3H).
13
C NMR (CDCl3, 100 MHz): = 162.1, 148.9, 147.0, 146.8, 145.4, 137.6, 137.4,
137.1, 134.1, 134.0, 129.2, 128.7, 128.4, 128.2, 127.8(2C), 127.4, 127.3(2C), 126.6, 122.1,
120.8, 78.0, 77.25, 67.4, 59.3, 55.8, 53.9, 20.9, 20.8. IR (neat): 3306, 1688, 1592, 1493 cm-1.
7.3.6 Functionalization of ethano-Tröger base
Rac-(R)-((5S,11S,14R)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin14-yl)(phenyl)methyl acetate (66)
Compound 65 (37.0 mg, 0.1 mmol) was dissolved in 3 ml of acetic anhydride and stirred for
12 h at rt. Workup involved evaporation of solvent followed by extraction using CH2Cl2 and
brine. The resulting organic phases were collected together and dried over Na2SO4,
concentrated under vacuum and further purified by flash chromatography over silica using nHexane/EtOAc (8:2) mixtures as eluent to provide compound 66 (21mg, 51% yield).
1
H NMR (CDCl3, 400 MHz): = 7.32 (m, 5H), 6.98 (d, 1 H, 3J =
7.8 Hz), 6.93 (t, 2 H, 3J = 6.8 Hz), 6.85 (d, 1 H, 3J = 8.6 Hz), 6.72
(d, 2 H, 3J = 5.0 Hz), 6.04 (d, 1 H, 3J = 9.3 Hz), 4.86 (d, 1 H, 2J =
17.9 Hz), 4.44 (d, 1 H, 2J = 17.4 Hz), 4.31 (d, 1 H, 2J = 17.2 Hz),
4.18 (d, 1 H, 2J = 17.9 Hz), 3.85 (m, 1H), 3.24 (t, 1 H, 3J = 13.4),
3.00 (dd, 1 H, 3J = 14.6, 4J = 5.6 Hz), 2.20 (s, 3H), 2.19 (s, 3H), 2.16 (s, 3H).
13
C NMR
(CDCl3, 100 MHz): = 170.6, 148.2, 147.3, 138.0, 137.0, 136.4, 134.6, 134.5, 129.4, 129.1,
-118-
Chapter 7
128.9 (2C), 128.7, 128.3, 128.2, 128.0, 127.8, 127.5 (2C), 74.6, 67.5, 59.2, 56.6, 54.5, 21.6,
21.0 (2C). IR (neat) 1732, 1267, 1263, 1253, 1242. IR (neat): 3306, 1688, 1592, 1493 cm-1.
HRMS (ESI+) calculated for (C27H28N2O2 + 1H) 413.2223, found 413.2209.
Rac-2,8-dimethyl-6H,12H-5,11-[(R)-((S)-chloro-phenyl-methyl)-ethano]-dibenzo[b,f]
[1,5]diazocine (67)
Compound 67 (40 mg, 0.11 mmol) and PCl5 (46mg, 0.22 mmol) were dissolved 5 ml CHCl3
under dinitrogen atmosphere and refluxed for 1h. The resulting mixture was extracted with
CH2Cl2 and brine. The reaction mixture was then diluted with CH2Cl2 and then extracted with
brine. The resulting organic phases were collected together and dried over Na2SO4,
concentrated under vacuum and further purified by flash chromatography over silica using nHexane/EtOAc (8:2) mixtures as eluent to provide compound 67 as a white solid (30.6 mg,
73% yield).
1
H NMR (CDCl3, 400 MHz): = 7.4 (m, 5 H), 7.08 (d, 1H, 3J =
8.1 Hz), 6.92 (d, 1H, 3J = 7.0 Hz), 6.78 (d, 1H, 3J = 8.1 Hz), 6.73
(s, 1H), 6.68 (s, 1H), 6.51 (d, 1H, 3J = 7.8 Hz), 5.01 (d, 1H, 3J =
9.8 Hz), 4.69 (d, 1H, 2J = 17.4 Hz), 4.63 (d, 1H, 2J = 17.7 Hz),
4.45 (d, 1H, 2J = 17.2 Hz), 4.27 (dd, 1H, 3J = 5.6 Hz), 4.13 (m,
1H), 4.05 (d, 1H, 2J = 17.7 Hz), 3.62 (m, 1H), 2.22 (s, 3H), 2.19 (s, 3H). 13C NMR (CDCl3,
100 MHz): = 147.4, 140.8, 137.0, 136.0, 134.7, 134.6, 129.3, 128.9, 128.4(2C), 128.3(2C),
128.2, 128.0, 127.9(2C), 67.8, 63.1, 59.4, 58.6, 54.2, 21.0, 20.9. M.P.: 75 °C. IR (neat): 824,
695, 598 cm-1. HRMS (ESI+): calculated for (C25H25N2Cl + 1H) 389.1779, found 389.1768.
Rac-2,8-dimethyl-6H,12H-5,11-[(R)-((R)-methoxy-phenyl-methyl)-ethano]-dibenzo[b,f]
[1,5]diazocine (69)
(70mg, 0.18mmol) of compound 67 was dissolved in 3 ml of MeOH and then 1.1 eq of
AgCO2Ph (46mg, 0.2mmol) was added. This reaction mixture was then refluxed for 12 h.
After cooling, the resulting mixture was filtered over celite, extracted with CH2Cl2 and a
saturated aqueous solution of NaHCO3. The resulting organic phases were collected together
and dried over Na2SO4, concentrated under vacuum and further purified by flash
-119-
Experimental Part
chromatography over silica using n-Hexane/EtOAc (8:2) mixtures as eluent to provide
compound as a pale yellow oil (39 mg, 75% yield).
1
H NMR (CDCl3, 400 MHz) = 7.36 (m, 5H), 7.03 (d, 1H, 3J =
7.8 Hz), (s, 1H),), 6.90 (d, 1H, 3J = 8.1 Hz), 6.79 (d, 1H, 3J = 8.1
Hz), 6.72 (s, 1H), 6.70 (d, 1H, 3J = 8.1 Hz), 6.68 (s, 1H), 4.96 (d,
1H, 2J = 17.4 Hz), 4.53 (d, 1H, 2J = 17.2 Hz), 4.45 (d, 1H, 3J = 6.1
Hz), 4.35 (d, 1H, 2J = 17.2 Hz), 4.18 (d, 1H, 2J = 17.4 Hz), 3.86
(dd, 1H, 3J = 13.0 Hz), 3.67 (m, 2H), 3.31 (s, 3H), 2.22 (s, 3H), 2.17 (s, 3H).
13
C NMR
(CDCl3, 100 MHz) = 148.3, 147.5, 140.7, 136.8, 136.7, 134.1, 133.9, 129.1, 128.7,
128.0(2C), 127.8, 127.7(2C), 127.6, 127.4, 127.1(2C), 86.8, 67.5, 59.2, 57.3, 56.4, 55.4, 20.8,
20.7. HRMS (ESI+) calculated for (C26H28N2O +1H) 385.2274, found 385.2260.
(S)-((5S,11S,14R)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14yl)(phenyl)methanol (70)
0.38 g, 1 mmol of compound (67) was dissolved in 20 ml of acetone (in light protected flask),
AgBF4 (1 eq, 0.194 g) is added. To this solution 2 ml of water is added subsequently and
reaction mixture was left to stir at rt for 5 h. Reaction was monitored by TLC. Later solvent is
removed and crude reaction mixture was purified by flash column chromatography to give
55% yield of 70 (single diastereomers is observed). For smaller scale, purification can be
performed by preparative TLC using silica plates and (10:1) n-hexane:EtOAc mixture
followed by extraction with EtOAc and evaporation.
1
H NMR (CDCl3, 400 MHz):  = 7.32 (m, 5H), 6.97 (d, 1 H, 3J =
7.8 Hz), 6.82 (m, 3 H), 6.66 (s, 1 H), 6.59 (s, 1 H), 4.91 (m, 1 H),
4.55 (m, 2 H), 4.33 (d, 1 H, 2J = 17.2), 4.06 (d, 1 H, 2J = 17.4),
3.82 (m, 2 H), 3.50 (m, 1 H), 2.82 (d, 1 H, 3J = 5.6 Hz), 2.14 (s,
3H), 2.13 (s, 3H). 13C NMR (CDCl3, 100 MHz): = 48.2, 147.3,
142.9, 136.6, 136.2, 134.4, 134.3, 129.4, 128.9, 128.3 (2C), 128.0, 127.9, 127.8, 127.7, 127.6,
126.5 (2C), 74.8, 67.3, 59.3, 55.9, 55.7, 20.9, 20.8 M.P.:105 °C. IR (neat): 2914 cm-1.
-120-
Chapter 7
7.4
General procedure I: Synthesis of racemic ethano-bridged Tröger bases . Typical
reaction conditions:
In a 5 mL screw-cap vial equipped with a magnetic stirring bar, 100.0 mg of Tröger base 1
(0.399 mmol) were introduced along with 1 ml of dry toluene and 1.76 mg Rh2(OAc)4 (0.004
mmol, 1 or 5 mol%) or 3.8 mg CuTC (0.2 mmol, 5 mol%). To this solution, diazo compound
(0.4 mmol) was added in one portion and the cap was placed at the top (unscrewed). The
reaction mixture was introduced into an already heated oil bath (100 °C). The reaction was
stirred for 1 h at that temperature. Then, 0.4 mmol of the same diazo compound was added in
one portion to the reaction mixture which was stirred for another 15 hours at 100 °C. The
reaction was monitored by ESI-MS. The solution was then allowed to cool to 20 °C and the
solvent was removed under reduced pressure. After NMR-spectroscopic analysis of the crude
reaction mixture, the desired product 73 was purified by column chromatography.
General procedure II: Enantiospecific transformation using CuTC as catalyst. Typical
reaction conditions:
Same as General procedure I using enantiopure (+)-(S,S)-1 (0.1 mmol) as substrate in place of
rac-1. Reactions were kept at 90 °C for 1 h (instead of 16h).
General procedure II: Enantiospecific transformation using Rh2(OAc)4 as catalyst.
Typical reaction conditions:
Same as General procedure I using enantiopure (–)-(R,R)-1 (0.1 mmol) as substrate in place
of rac-1. Reactions were kept at 100 °C for 6 hours (instead of 16h).
General procedure IV: Regioselective transformation
Identical to procedure I; the reaction being carried out under dry conditions and in dinitrogen
atmosphere to afford better yields and avoid the formation of by-products.
-121-
Experimental Part
1,1'-(2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocine-13,13diyl)diethanone (73a)
Following procedure I using CuTC as catalyst, 97.4 mg (70%) of
73a was obtained from 100 mg of rac-1. 1H NMR (500 MHz,
CDCl3)  = 7.25 (d, J = 8.1 Hz, 1H), 6.97 (d, J = 8.0 Hz, 1H), 6.93
(d, J = 8.1 Hz, 1H), 6.86 (d, J = 8.1 Hz, 1H), 6.69 (s, J = 9.0 Hz,
1H), 6.57 (s, J = 12.4 Hz, 1H), 4.94 (d, J = 14.8 Hz, 1H), 4.58 (d, J
= 17.9 Hz, 1H), 4.48 (d, J = 17.9 Hz, 1H), 4.36 (d, J = 17.5 Hz, 1H), 4.23 (d, J = 17.5 Hz,
1H), 3.55 (d, J = 14.8 Hz, 1H), 2.18 (s, 6H), 2.17 (s, 3H), 2.16 (s, 3H), 1.87 (s, 3H).
13
C
NMR (126 MHz, CDCl3)  = 203.3, 201.6, 146.6, 143.7, 136.1, 135.8, 135.0, 134.8, 130.2,
129.5, 128.9, 128.6, 128.3, 128.0, 93.4, 59.3, 57.5, 56.7, 26.8, 26.0, 20.9, 20.8. M.P.: 170173 °C. IR (neat): -1 = 2961, 2920, 1707, 1701, 1492, 1432, 1349, 1219 cm-1.HRMS (ESI+)
calculated for (C22H24N2O2 +1H): 349.191. Found: 349.1907.
Following procedure II, using (+)-(S,S)-2 as substrate at 100 °C, (5R,11S)-73a (70%, ee
64%), [α]D20 = +80 (c 0.02, CH2Cl2).
(5R,11S)-dimethyl
2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocine-
13,13-dicarboxylate (73b)
73b was obtained from 100 mg of rac-1. 1H NMR (400 MHz,
CDCl3)  = 7.21 (d, J = 8.1 Hz, 1H), 7.01 (d, J = 8.0 Hz, 1H), 6.92
– 6.85 (m, 2H), 6.70 (s, 1H), 6.61 (s, 1H), 5.11 (d, J = 18.4 Hz,
1H), 4.84 (d, J = 17.5 Hz, 1H), 4.69 (d, J = 15.3 Hz, 1H), 4.45 (d,
J = 18.4 Hz, 1H), 4.28 (d, J = 17.5 Hz, 1H), 3.76 (s, 3H), 3.72 (d, J = 15.3 Hz, 1H), 3.39 (s,
6H), 2.18 (s, 3H), 2.16 (s, 3H). 13C NMR (100 MHz, CDCl3)  = 169.6, 169.3, 146.9, 143.6,
137.3, 136.9, 135.9, 135.2, 130.6, 128.9, 128.7, 128.5, 128.4, 128.3, 78.6, 59.7, 59.4, 57.3,
53.4, 53.3, 21.2, 21.1. M.P.: 180-183 °C. IR (neat): -1 = 2980, 2925, 1731, 1495, 1441 cm-1.
HRMS (ESI+) calculated for C24H24N2O4 (M+1): 380.4421. Found: 380.4429.
-122-
Chapter 7
Diethyl
2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocine-13,13-
dicarboxylate (73d)
73d was obtained from 100 mg of rac-1. 1H NMR (400 MHz,
CDCl3)  = 7.22 (d, J = 8.0 Hz, 1H), 7.00 (d, J = 8.0 Hz, 1H),
6.88-6.86 (m, 2H), 6.70 (s, 1H), 6.59 (s, 1H), 5.14 (d, J = 18.4 Hz,
1H), 4.86 (d, J = 17.4 Hz, 1H), 4.68 (d, J = 15.2 Hz, 1H), 4.46 (d, J
= 18.4 Hz, 1H), 4.33 – 4.13 (m, 3H), 3.96 (dq, J = 10.7, 7.1 Hz, 1H), 3.84 (dq, J = 10.7, 7.1
Hz, 1H), 3.68 (d, J = 15.2 Hz, 1H), 2.17 (s, 3H), 2.15(s, 3H), 1.22 (t, J = 7.1 Hz, 3H), 0.77 (t,
J = 7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3)  = 169.1, 168.7, 147.0, 143.8, 137.4, 137.0,
135.9, 135.1, 130.8, 128.8, 128.7, 128.3, 128.2, 78.3, 62.3, 62.1, 59.7, 59.3, 57.3, 21.1, 21.08,
14.5, 13.6. M.P.: 190-193 °C. IR (neat): -1 = 2980, 2925, 1735, 1497, 1444 cm-1. HRMS
(ESI+) calculated for C24H29N2O4 (M+1): 409.2121. Found: 409.2116.
Ethyl 14-acetyl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine -14carboxylate (73e)
Following procedure I using CuTC as catalyst, 116.4 mg (80%,
both diastereomers) of 73e was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
chromatography (SiO2, 3 x 30 cm, acetone/pentane 3/97).
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: 99 mg
(68%). 1H NMR (400 MHz, CDCl3)  = 7.24 (d, J = 8.1 Hz, 1H), 6.98 (d, J = 8.0 Hz, 1H),
6.91 (d, J = 8.1 Hz, 1H), 6.86 (d, J = 8.0 Hz, 1H), 6.70 (s, 1H), 6.56 (s, 1H), 4.96 (d, J = 18.1
Hz, 1H), 4.78 (d, J = 15.2 Hz, 1H), 4.49 (d, J = 5.7 Hz, 1H), 4.44 (d, J = 5.0 Hz, 1H), 4.21
(d, J = 17.5 Hz, 1H), 3.72 (s, 3H), 3.55 (d, J = 15.2 Hz, 1H), 2.17 (s, 3H), 2.15 (s, 3H), 1.93
(s, 3H).
13
C NMR (100 MHz, CDCl3)  = 200.5, 169.7, 147.0, 143.3, 136.4, 136.3, 136.1,
135.0, 130.6, 129.5, 128.9, 128.8, 128.4, 128.3, 85.0, 59.6, 57.7, 56.6, 53.3, 26.3, 21.2, 21.1.
IR (neat): -1 = 2921, 1737, 1720, 1495, 1433, 1351, 1233, 1182 cm-1. HRMS (ESI+)
calculated for C22H25N2O3(M+1): 365.1859. Found: 365.1855.
-123-
Experimental Part
Ethyl 14-acetyl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine-14carboxylate (73f)
Following procedure I using CuTC as catalyst, 124 mg (82%, both
diastereomers) of 73f was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: 105.8 mg
Hz, 1H), 4.78 (d, J = 15.2 Hz, 1H), 4.47 (td, J = 18.2 Hz, 2H), 4.27 – 4.08 (m, 3H), 3.55 (d, J
= 15.2 Hz, 1H), 2.17 (s, 3H), 2.16 (d, J = 9.7 Hz, 6H), 2.15 (s, 3H), 1.94 (s, 3H), 1.94 (s,
3H), 1.22 (t, J = 7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3)  = 200.6, 169.2, 147.1, 143.3,
136.4, 136.4, 136.1, 135.0, 130.6, 129.5, 128.9, 128.8, 128.4, 128.3, 84.9, 62.3, 59.6, 57.7,
56.6, 26.2, 21.14, 21.10, 14.4. IR (neat): -1 =2921, 1720, 1496, 1434, 1352, 1217, 1183 cm1
. HRMS (ESI+) calculated for C23H27N2O3(M+1): 379.2016. Found: 379.2015.
Following procedure II, using (+)-(S,S)-1 as substrate at 90 °C, (5S,11R,14R)-73f (67%, ee
50%), [α]D20 = +135 (c 0.02, CH2Cl2).
Minor diastereomer {(5R,11S,14R) and (5S,11R,14S)}: white solid 18.1 mg (12%). 1H NMR
(400 MHz, CDCl3)  = 7.19 (d, J = 8.0 Hz, 1H), 7.01 (d, J = 8 Hz, 1H), 6.90 (td, J = 8.0 Hz,
2H), 6.67 (s, 1H), 6.61 (s, 1H), 4.68 (d, J = 15.4 Hz, 1H), 4.60 (d, J = 17.6 Hz, 1H), 4.45 (d,
J = 17.6 Hz, 1H), 4.39 – 4.25 (m, 2H), 4.01 – 3.80 (m, 3H), 2.51 (s, 3H), 2.17 (s, 6H), 0.85 (t,
J = 7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3)  = 206.6, 168.1, 144.0, 135.9, 135.6, 134.9,
130.1, 129.0, 128.7, 128.5, 128.5, 128.1, 85.0, 62.3, 59.5, 58.5, 57.9, 25.3, 21.2, 21.1, 13.7.
M.P.: 162-164 °C. IR (neat): -1 = 2962, 2922, 1736, 1496, 1442, 1352, 1258, 1240, 1184
cm-1. HRMS (ESI+) calculated for C23H27N2O3(M+1): 379.2016. Found: 379.2018.
-124-
Chapter 7
Isopropyl 14-acetyl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine14-carboxylate (73g)
both diastereomers) of 73g was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
6.91 (d, J = 8.0 Hz, 1H), 6.86 (d, J = 8.0 Hz, 1H), 6.71 (s, 1H), 6.56 (s, 1H), 5.08 – 4.96 (m,
2H), 4.76 (d, J = 15.2 Hz, 1H), 4.54 – 4.42 (td, J = 17.4 Hz, 2H), 4.21 (d, J = 17.4 Hz, 1H),
3.52 (d, J = 15.2 Hz, 1H), 2.17 (s, 3H), 2.15 (s, 3H), 1.93 (s, 3H), 1.28 – 1.11 (m, 6H). 13C
NMR (100 MHz, CDCl3)  = 200.6, 168.6, 147.1, 143.4, 136.5, 136.3, 136.3, 134.9, 130.6,
129.5, 128.9, 128.8, 128.5, 128.3, 84.8, 69.9, 59.6, 57.6, 56.6, 26.1, 21.9, 21.6, 21.1, 21.0.
M.P.: 172-175 °C. IR (neat): -1 = 2980, 1720, 1496, 1434, 1370, 1234 cm-1. HRMS (ESI+)
Minor diastereomer {(5R,11S,14R) and (5S,11R,14S)}: 25 mg (16%). 1H NMR (500 MHz,
CDCl3)  = 7.17 (d, J = 8.0 Hz, 1H), 7.00 (d, J = 8.0 Hz, 1H), 6.90 (dd, J = 8.0, 1.6 Hz, 1H),
6.87 (dd, J = 8.0, 1.6 Hz, 1H), 6.66 (d, J = 0.8 Hz, 1H), 6.59 (d, J = 0.8 Hz, 1H), 4.82 – 4.73
(m, 1H), 4.66 (d, J = 15.4 Hz, 1H), 4.60 (d, J = 17.5 Hz, 1H), 4.43 (d, J = 17.6 Hz, 1H), 4.34
– 4.22 (m, 2H), 3.81 (d, J = 15.4 Hz, 1H), 2.51 (s, 3H), 2.16 (s, 3H), 2.16 (s, 3H), 1.09 (d, J
= 6.3 Hz, 3H), 0.62 (d, J = 6.2 Hz, 3H).
13
C NMR (100 MHz, CDCl3)  = 206.9, 167.5,
147.3, 144.1, 136.4, 136.0, 135.6, 134.8, 130.3, 129.0, 128.7, 128.5, 128.4, 128.1, 85.1, 70.1,
59.5, 58.4, 58.0, 25.2, 21.6, 21.12, 21.09, 20.7. IR (neat): -1 = 2922, 1720, 1497, 1433, 1374,
1236, 1218 cm-1. HRMS (ESI+) calculated for C24H29N2O3(M+1): 393.2172. Found:
393.2176.
tert-Butyl 14-acetyl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine14-carboxylate (73h)
both diastereomers) of 73h was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: 112 mg
-125-
Experimental Part
(69%). 1H NMR (400 MHz, CDCl3)  = 7.22 (d, J = 8 Hz, 1H), 6.99 (d, J = 7.6 Hz, 1H),
6.89 (td, J = 8, 7.6 Hz, 2H), 6.71 (s, 1H), 6.56 (s, 1H), 5.02 (d, J = 18.1 Hz, 1H), 4.73 (d, J =
15.2 Hz, 1H), 4.47 (td, J = 18.1, 17.5 Hz, 2H), 4.20 (d, J = 17.5 Hz, 1H), 3.52 (d, J = 15.2
Hz, 1H), 2.18 (s, 3H), 2.13 (d, J = 10.2 Hz, 3H), 1.94 (s, 3H), 1.42 (s, 9H).
13
C NMR (100
MHz, CDCl3)  = 200.8, 168.2, 147.1, 143.5, 136.6, 136.2, 134.9, 130.6, 129.5, 128.8, 128.4,
128.3, 85.1, 83.1, 59.6, 57.5, 56.6, 28.1, 26.1, 21.1, 21.1. M.P.: 179-181 °C. IR (neat): -1 =
2980, 1718, 1495, 1456, 1368, 1245 cm-1. HRMS (ESI+) calculated for C25H31N2O3(M+1):
407.2329. Found: 407.2323.
Minor diastereomer {(5R,11S,14R) and (5S,11R,14S)}: 22.7 mg (14%). 1H NMR (400 MHz,
CDCl3)  = 7.16 (d, J = 8.0 Hz, 1H), 7.00 (d, J = 8.0 Hz, 1H), 6.92 (d, J = 8.1 Hz, 1H), 6.86
(d, J = 8.1 Hz, 1H), 6.66 (s, 1H), 6.61 (s, 1H), 4.65 (td, J = 14.5, 12.6 Hz, 2H), 4.42 (d, J =
17.6 Hz, 1H), 4.29 (td, J = 17.6, 12.6 Hz, 2H), 3.74 (d, J = 14.5 Hz, 1H), 2.51 (s, 3H), 2.17
(s, 3H), 2.16 (s, 3H), 1.08 (s, 9H).
13
C NMR (100 MHz, CDCl3)  = 207.4, 207.1, 166.8,
147.4, 144.3, 136.7, 136.1, 135.6, 134.7, 134.1, 130.4, 129.0, 128.7, 128.3, 128.3, 128.1,
85.7, 82.8, 59.5, 58.1, 57.9, 27.4, 25.2, 21.12, 21.10. IR (neat): -1 = 2922, 1733, 1699, 1497,
1431, 1368, 1258 cm-1. HRMS (ESI+) calculated for C25H31N2O3(M+1): 407.2329. Found:
407.2325.
Ethyl 2,8-dimethyl-14-propionyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine14-carboxylate (73i)
both diastereomers) of 73i was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
(75%). M.P.: 170-173°C. 1H NMR (500 MHz, CDCl3)  = 7.23 (d, J = 8.1 Hz, 1H), 6.98 (d,
J = 8.0 Hz, 1H), 6.90 (dd, J = 8.1, 1.6 Hz, 1H), 6.86 (dd, J = 8.0, 1.6 Hz, 1H), 6.70 (d, J =
0.8 Hz, 1H), 6.56 (d, J = 0.9 Hz, 1H), 5.00 (d, J = 18.1 Hz, 1H), 4.80 (d, J = 15.2 Hz, 1H),
4.48 (t, J = 16.9 Hz, 2H), 4.27 – 4.09 (m, 3H), 3.53 (dd, J = 15.2, 1.1 Hz, 1H), 2.53 – 2.27
(m, 2H), 2.17 (s, 3H), 2.14 (s, 3H), 1.22 (t, J = 7.1 Hz, 3H), 0.57 (t, J = 7.4 Hz, 3H).
13
C
NMR (126 MHz, CDCl3)  = 204.2, 169.3, 147.1, 143.4, 136.5, 136.4, 136.3, 134.9, 130.7,
129.5, 128.9, 128.8, 128.4, 128.3, 84.4, 62.1, 59.7, 57.7, 56.7, 31.7, 21.12, 21.09, 14.4, 8.8.
-126-
Chapter 7
IR (neat): -1 = 2976, 1719, 1496, 1434, 1336, 1256, 1216 cm-1. HRMS (ESI+) calculated for
C24H29N2O3(M+1): 393.2172. Found: 393.2171.
CDCl3)  = 7.17 (d, J = 8.0 Hz, 1H), 6.99 (d, J = 8.0 Hz, 1H), 6.90 (d, J = 8.0 Hz, 1H), 6.86
(d, J = 8.0 Hz, 1H), 6.65 (s, 1H), 6.59 (s, 1H), 4.66 (d, J = 15.4 Hz, 1H), 4.56 (d, J = 14.7
Hz, 1H), 4.41 (d, J = 17.6 Hz, 1H), 4.29 (td, J = 17.6, 15.4 Hz, 2H), 4.00 – 3.81 (m, 3H),
3.08 – 2.88 (m, 2H), 2.16 (s, 6H), 1.08 (t, J = 7.3 Hz, 3H), 0.84 (t, J = 7.1 Hz, 3H).
13
C
NMR (126 MHz, CDCl3)  = 209.1, 168.4, 147.3, 144.2, 136.3, 136.1, 135.5, 134.8, 130.1,
129.0, 128.7, 128.5, 128.4, 128.1, 84.80, 62.25, 59.52, 58.73, 58.00, 30.31, 21.16, 21.09,
13.76, 8.91. IR (neat): -1 = 2922, 1735, 1720, 1497, 1434, 1368, 1259 cm-1. HRMS (ESI+)
Ethyl 14-butyryl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine-14carboxylate (73j)
both diastereomers) of 73j was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
6.89 (td, J = 8.1, 7.8 Hz, 2H), 6.72 (s, 1H), 6.56 (s, 1H), 4.99 (d, J = 18.2 Hz, 1H), 4.82 (d, J
= 15.2 Hz, 1H), 4.49 (td, J = 18.2 Hz, 2H), 4.28 – 4.07 (m, 3H), 3.55 (d, J = 15.2 Hz, 1H),
2.46 – 2.24 (m, 2H), 2.18 (s, 3H), 2.15 (s, 3H), 1.37 – 1.18 (m, 4H), 1.10 – 0.81 (m, 1H), 0.59
(t, J = 7.4 Hz, 3H).
13
C NMR (100 MHz, CDCl3)  = 203.2, 169.3, 147.1, 143.4, 136.5,
136.3, 135.0, 134.3, 130.7, 129.5, 128.8, 128.7, 128.4, 128.3, 84.6, 62.2, 59.6, 57.8, 56.8,
40.2, 21.1, 21.0, 17.9, 14.4, 13.5. M.P.: 168-170 °C. IR (neat): -1 = 2962, 1718, 1496, 1434,
1260, 1216 cm-1. HRMS (ESI+) calculated for C25H31N2O3(M+1): 407.2329. Found:
407.2334.
CDCl3)  = 7.19 (d, J = 8.0 Hz, 1H), 7.00 (b, 1H), 6.89 (td, J = 8.0 Hz, 2H), 6.65 (s, 1H),
6.59 (s, 1H), 4.68 (d, J = 15.4 Hz, 1H), 4.58 (d, J = 17.6 Hz, 1H), 4.49 – 4.23 (m, 3H), 4.02 –
3.80 (m, 3H), 3.03 – 2.80 (m, 2H), 2.17 (s, 6H), 1.7 – 1.5 (m, 5H), 0.91 (t, J = 7.4 Hz, 3H),
0.85 (t, J = 7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3)  = 208.0, 168.3, 144.2, 136.0, 135.5,
-127-
Experimental Part
130.1, 129.0, 128.8, 128.5, 128.0, 62.3, 59.5, 58.7, 57.9, 38.8, 21.2, 21.1, 17.9, 14.0, 13.8.
IR: (neat): -1 = 2925, 1735, 1700, 1497, 1444, 1368, 1258 cm-1. HRMS (ESI+) calculated
for C25H31N2O3(M+1): 407.2329. Found: 407.2317.
Ethyl 14-benzoyl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine-14carboxylate (73k)
both diastereomers) of 73k was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
(64%). 1H NMR (500 MHz, CDCl3)  = 8.17-8.14 (m, 2H), 7.46 – 7.34 (m, 1H), 7.28 – 7.23
(m, 2H), 7.02 (d, J = 8.0 Hz, 1H), 6.87 (d, J = 8 Hz, 1H), 6.75 – 6.67 (m, 2H), 6.56 – 6.43
(m, 2H), 5.54 (t, J = 18.7 Hz, 1H), 5.03 (d, J = 15.1 Hz, 1H), 4.71 (d, J = 17.4 Hz, 1H), 4.47
(d, J = 18.3 Hz, 1H), 4.24 (d, J = 17.4 Hz, 1H), 4.18 – 4.02 (m, 2H), 3.61 (d, J = 15.1, 1H),
2.18 (s, 3H), 1.99 (s, 3H), 1.05 (t, J = 7.1 Hz, 3H). 13C NMR (126 MHz, CDCl3)  = 191.50,
169.99, 147.24, 143.07, 137.07, 136.73, 136.10, 135.64, 134.95, 132.99, 131.30, 129.95,
128.81, 128.64, 128.35, 128.21, 128.19, 128.05, 82.35, 62.02, 59.84, 58.91, 57.39, 21.09,
21.01, 14.26. M.P.: 143-145 °C. IR (neat): -1 = 2922, 1737, 1686, 1495, 1446, 1242, 1211
Minor diastereomer {(5R,11S,14R) and (5S,11R,14S)}: white solid 24.6 mg (14%). 1H NMR
(500 MHz, CDCl3)  = 8.59 (d, J = 8.4 Hz, 2H), 7.65 (t, J = 6.7 Hz, 1H), 7.54 (t, J = 7.0 Hz,
2H), 7.28 (d, J = 7.6 Hz, 1H), 7.10 – 7.01 (m, 2H), 6.91 (d, J = 7.6 Hz, 1H), 6.71 (s, 1H),
6.59 (s, 1H), 4.82 (d, J = 15.5 Hz, 1H), 4.65 (d, J = 17.3 Hz, 1H), 4.41 (d, J = 17.7 Hz, 1H),
4.33 – 4.12 (m, 3H), 4.08 – 3.86 (m, 2H), 2.25 (s, 3H), 2.17 (s, 3H), 0.94 – 0.87 (m, 3H). 13C
NMR (126 MHz, CDCl3)  = 197.48, 168.78, 147.10, 144.11, 136.83, 136.26, 135.73,
135.59, 134.82, 133.71, 133.42, 131.03, 130.85, 130.46, 129.11, 129.01, 128.77, 128.69,
128.64, 128.38, 128.34, 128.07, 84.39, 62.37, 61.45, 59.59, 58.16, 21.23, 21.07, 13.78. M.P.:
139-140 °C. IR (neat): -1 = 2920, 2851, 1736, 1671, 1496, 1446, 1256, 1222 cm-1. HRMS
(ESI+) calculated for C28H29N2O3(M+1): 441.2172. Found: 441.2174.
-128-
Chapter 7
Ethyl 14-isobutyryl-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocine14-carboxylate (73l)
both diastereomers) of 73l was obtained from 100 mg of rac-1.
Separation
of
the
diastereomers
was
done
by
column
Hz, 1H), 4.84 (d, J = 15.2 Hz, 1H), 4.51 – 4.39 (m, 2H), 4.27 – 4.16 (m, Hz, 2H), 4.08 (dq, J
= 10.8, 7.1 Hz, 2H), 3.49 (d, J = 16.1 Hz, 1H), 3.43 (dt, J = 13.4, 6.7 Hz, 2H), 2.17 (s, 3H),
2.14 (s, 3H), 1.23 (t, J = 7.1 Hz, 3H), 0.85 (d, J = 6.8 Hz, 3H), 0.03 (d, J = 6.6 Hz, 3H). 13C
NMR (100 MHz, CDCl3)  = 208.8, 169.3, 147.1, 143.6, 136.4, 136.3, 134.9, 131.0, 129.5,
129.0, 128.8, 128.4, 128.3, 84.2, 62.1, 59.5, 57.8, 57.1, 36.5, 23.1, 21.1, 21.1, 18.0, 14.5. IR
(neat): -1 = 2962, 1718, 1607, 1495, 1434, 1260, 1216 cm-1. HRMS (ESI+) calculated for
C25H31N2O3(M+1): 407.2329. Found: 407.2334.
1-(2,8-dimethyl-14-phenyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14-yl)
ethanone (73m)
Following procedure I using CuTC as catalyst, 122.2 mg (80%, both
diastereomers) of 73m was obtained from 100 mg of rac-1. Separation
of the diastereomers was done by column chromatography (SiO2, 3 x
30 cm, acetone/pentane 5/95).
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White solid
111.5 mg (73%). 1H NMR (400 MHz, CDCl3)  = 7.61 (b, 2H), 7.38 (t, J = 7.8 Hz, 2H),
7.34 – 7.26 (m, 2H), 7.00 (d, J = 8.0 Hz, 1H), 6.96 (d, J = 8.0 Hz, 1H), 6.85 (d, J = 8.0 Hz,
1H), 6.64 (s, 2H), 5.28 (d, J = 14.6 Hz, 1H), 4.73 (d, J = 17.6 Hz, 1H), 4.60 (d, J = 17.4 Hz,
1H), 4.41 (d, J = 17.6 Hz, 1H), 4.31 (d, J = 17.4 Hz, 1H), 3.52 (d, J = 14.6 Hz, 1H), 2.18 (s,
3H), 2.14 (s, 3H), 1.66 (s, 3H). 13C NMR (100 MHz, CDCl3)  = 205.5, 147.5, 144.7, 141.5,
136.2, 135.9, 135.8, 134.4, 130.5, 129.8, 129.0, 128.8, 128.7, 128.3, 128.1, 127.8, 127.4,
82.33, 59.62, 59.43, 56.37, 25.27, 21.15, 21.07. M.P.: 215-217 °C. IR (neat): -1 = 2923,
1706, 1495, 1260, 1220 cm-1. HRMS (ESI+) calculated for C26H27N2O (M+1): 383.2117.
Found: 383.2129.
-129-
Experimental Part
Following procedure II, using (+)-(S,S)-1 as substrate, (5S,11R,14R)-73m: major
diastereomer (73%, ee 90%), [α]D20 = +198 (c 0.02, CH2Cl2).
Minor diastereomer {(5R,11S,14R) and (5S,11R,14S)}: yellow gel 5.3 mg, (6%). 1H NMR
(400 MHz, CDCl3)  = 7.53 (d, J = 6.6 Hz, 2H), 7.28 (m, 1H), 7.13 (d, J = 7.4 Hz, 2H), 6.93
(td, J = 8.5 Hz, 2H), 6.85 (d, J = 8.5 Hz, 1H), 6.69 (s, 1H), 6.34 (s, 1H), 4.75 (d, J = 15.8 Hz,
1H), 4.60 (d, J = 17.4 Hz, 1H), 4.37 (d, J = 17.4 Hz, 1H), 4.21 (d, J = 15.8 Hz, 1H), 4.10 (d,
J = 16.9 Hz, 1H), 3.97 (d, J = 16.9 Hz, 1H), 2.29 (s, 3H), 2.17 (s, 3H), 2.08 (s, 3H).
13
C
NMR (100 MHz, CDCl3)  = 210.7, 147.8, 143.7, 137.4, 137.1, 136.1, 134.9, 134.4, 129.8,
128.9, 128.6, 128.3, 128.1, 127.5, 127.4, 81.6, 59.7, 59.6, 59.2, 23.0, 21.12, 21.10. IR (neat):
-1 = 2917, 1701, 1494, 1447, 1349, 1260 cm-1. HRMS (ESI+) calculated for
C26H27N2O(M+1): 383.2117. Found: 383.2121.
1-(2,8-dimethyl-14-phenyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14-yl)
propan-1-one (73ma)
Following procedure I using CuTC as catalyst, 131.5 mg (83%, major
diastereomer) of 73ma was obtained from 100 mg of rac-1.
Separation of the diastereomers was done by column chromatography
(SiO2, 3 x 30 cm, acetone/pentane 5/95).
131.5 mg (60%). 1H NMR (400 MHz, CDCl3)  = 7.58 (b, 2H), 7.35 (t, J = 7.7 Hz, 2H), 7.32
– 7.27 (m, 12H), 6.99 (d, J = 8.0 Hz, 1H), 6.95 (d, J = 7.7 Hz, 1H), 6.84 (d, J = 8.0 Hz, 1H),
6.63 (s, 1H), 6.62 (s, 1H), 5.28 (d, J = 14.6 Hz, 1H), 4.73 (d, J = 17.6 Hz, 1H), 4.60 (d, J =
17.5 Hz, 1H), 4.41 (d, J = 17.6 Hz, 1H), 4.29 (d, J = 17.4 Hz, 1H), 3.49 (d, J = 14.6 Hz, 1H),
2.25 – 1.96 (m, 8H), 0.26 (t, J = 7.4 Hz, 3H).
13
C NMR (125 MHz, CDCl3)  = 209.37,
147.66, 144.74, 141.61, 136.28, 136.21, 135.90, 134.47, 130.67, 129.78, 129.03, 128.78,
128.76, 128.28, 128.09, 127.72, 127.33, 82.13, 59.70, 59.35, 56.41, 30.68, 21.16, 21.09, 9.47.
M.P.: 196-198 °C. IR (neat): -1 = 2964, 1706, 1495, 1259, 1015 cm-1. HRMS (ESI+)
calculated for C27H29N2O(M+1): 397.2274. Found: 397.2293.
Following procedure II, using (+)-(S,S)-1 as substrate, (5S,11R,14R)-73ma: major
-130-
Chapter 7
(2,8-dimethyl-14-phenyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]diazocin-14-yl)
(phenyl)methanone (73mb)
Following procedure I using CuTC as catalyst, 152 mg (85%, major
diastereomer) of 73mb was obtained from 100 mg of rac-1.
Separation of the diastereomers was done by precipitation in
acetone/hexane solution.
152 mg (85%). 1H NMR (400 MHz, CDCl3)  = 7.72 (d, J = 7.5 Hz, 2H), 7.68 – 7.61 (m,
2H), 7.39 (t, J = 7.8 Hz, 2H), 7.22 (t, J = 7.4 Hz, 1H), 7.15 (t, J = 7.4 Hz, 1H), 7.06 – 6.95
(m, 3H), 6.84 (d, J = 7.9 Hz, 1H), 6.71 (d, J = 8.0 Hz, 1H), 6.66 (s, 1H), 6.48 (s, 1H), 6.44 (d,
J = 8.1 Hz, 1H), 5.41 (d, J = 14.7 Hz, 1H), 5.12 (d, J = 17.4 Hz, 1H), 4.70 (d, J = 17.4 Hz,
1H), 4.50 (d, J = 17.5 Hz, 1H), 4.27 (d, J = 17.5 Hz, 1H), 3.48 (d, J = 13.9 Hz, 1H), 2.14 (s,
3H), 2.00 (s, 3H). 13C NMR (100 MHz, CDCl3)  = 198.4, 147.7, 144.7, 141.0, 138.0, 136.5,
135.7, 135.2, 134.4, 131.5, 131.3, 129.4, 129.2, 129.0, 128.7, 128.4, 128.2, 127.8, 127.6,
127.4, 127.3, 82.5, 62.2, 59.7, 56.9, 21.1, 21.0. M.P.: 255-257 °C. IR (neat): -1 = 2921,
1671, 1596, 1495, 1259 cm-1. HRMS (ESI+) calculated for C31H29N2O(M+1): 445.2274.
Found: 445.2272.
Following procedure II, using (+)-(S,S)-1 as substrate, (5S,11R,14R)-73mb: major
diastereomer before crystallization (85%, ee 90%) and after crystallization (80%, ee 95%),
[α]D20 = +245 (c 0.02, CH2Cl2).
1-(2,8-dimethyl-14-(p-tolyl)-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocin-14yl)ethanone (73mc)
Following procedure I using CuTC as catalyst, 181 mg (80%,
major diastereomer) of 73mc was obtained from 100 mg of
rac-1. Separation of the diastereomers was done by column
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White
solid 181 mg (80%). 1H NMR (400 MHz, CDCl3)  = 7.47 (b, 2H), 7.28 (d, J = 8.0 Hz, 1H),
7.17 (d, J = 8.3 Hz, 2H), 6.98 (d, J = 8.0 Hz, 1H), 6.94 (d, J = 8.1 Hz, 1H), 6.84 (d, J = 8.8
Hz, 1H), 6.62 (s, 2H), 5.24 (d, J = 14.6 Hz, 1H), 4.72 (d, J = 17.5 Hz, 1H), 4.58 (d, J = 17.5
(s, 3H), 2.17 (s, 3H), 2.13 (s, 3H), 1.64 (s, 3H).
13
C NMR (125 MHz, CDCl3)  = 205.59,
147.63, 144.88, 137.57, 136.31, 135.89, 135.88, 134.43, 130.56, 129.82, 129.78, 128.80,
-131-
Experimental Part
128.72, 128.24, 128.07, 127.32, 82.12, 59.64, 59.46, 56.33, 25.23, 21.37, 21.16, 21.09. M.P.:
174-175 °C. IR (neat): -1 = 2922, 1705, 1679, 1496, 1432, 1260, 1220 cm-1. HRMS (ESI+)
calculated for C27H29N2O (M+1): 397.2274. Found: 397.2271.
Following procedure II, using (+)-(S,S)-1 as substrate, (5S,11R,14R)-73mc: major
1-(2,8-dimethyl-14-(m-tolyl)-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocin-14-yl)
ethanone (73md)
major diastereomer) of 73md was obtained from 100 mg of rac1. Separation of the diastereomers was done by column
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White
solid 187.8 mg (83%). 1H NMR (400 MHz, CDCl3)  = 7.45 (b, 1H), 7.36 – 7.22 (m, 3H),
7.07 (d, J = 7.5 Hz, 1H), 7.01 – 6.92 (m, 2H), 6.83 (d, J = 8.1 Hz, 1H), 6.62 (s, 2H), 5.24 (d, J
= 14.6 Hz, 1H), 4.73 (d, J = 17.6 Hz, 1H), 4.58 (d, J = 17.4 Hz, 1H), 4.39 (d, J = 17.6 Hz,
1H), 4.28 (d, J = 17.4 Hz, 1H), 3.50 (d, J = 14.6 Hz, 1H), 2.36 (s, 4H), 2.17 (s, 3H), 2.14 (s,
3H), 1.64 (s, 3H). 13C NMR (100 MHz, CDCl3)  = 205.6, 147.6, 144.8, 141.5, 138.7, 136.3,
135.9, 135.8, 134.4, 130.6, 129.8, 128.9, 128.8, 128.7, 128.5, 128.2, 128.1, 128.0, 124.5,
82.3, 59.6, 59.5, 56.4, 25.3, 22.0, 21.2, 21.1. M.P.: 174-175 °C. IR (neat): -1 = 2922, 1705,
1679, 1496, 1432, 1260, 1220 cm-1. HRMS (ESI+) calculated for C27H29N2O(M+1):
397.2274. Found: 397.2263.
Following procedure II, using (+)-(S,S)-1 as substrate, (5S,11R,14R)-73md: major
diastereomer (83%, ee 84 %), [α]D20 = +153 (c 0.02, CH2Cl2).
1-(14-(4-chlorophenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo[b,f] [1,5]diazocin14-yl)ethanone (73me)
Following procedure I using CuTC as catalyst, 138.3 mg
(83%, both diastereomers) of 73me was obtained from 100
mg of rac-1. Separation of the diastereomers was performed
by
column
chromatography
(SiO2,
3
x
30
cm,
acetone/pentane 5/95).
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White solid 138.3 mg (83%). 1H
NMR (400 MHz, CDCl3)  = 7.53 (b, 2H), 7.34 (d, J = 8.9 Hz, 2H), 7.27 (d, J = 8.1 Hz, 1H),
-132-
Chapter 7
7.02 – 6.93 (m, 2H), 6.85 (d, J = 8.0 Hz, 1H), 6.63 (s, 1H), 6.62 (s, 1H), 5.23 (d, J = 14.5 Hz,
1H), 4.65 (d, J = 17.5 Hz, 1H), 4.55 (d, J = 17.5 Hz, 1H), 4.39 (d, J = 17.6 Hz, 1H), 4.28 (d, J
= 17.5 Hz, 1H), 3.46 (d, J = 14.2 Hz, 1H), 2.17 (s, 3H), 2.14 (s, 3H), 1.63 (s, 3H). 13C NMR
(100 MHz, CDCl3)  = 205.01, 147.44, 144.47, 140.02, 136.15, 135.93, 135.72, 134.61,
133.86, 130.48, 129.86, 129.27, 128.85, 128.80, 128.39, 128.09, 82.03, 59.58, 59.32, 56.29,
25.22, 21.17, 21.08. M.P.: 168-170 °C. IR (neat): -1 = 2922, 1707, 1493, 1349, 1220 cm-1.
HRMS (ESI+) calculated for C26H26N2OCl (M+1): 417.1728. Found: 417.1730.
Following procedure II, using (+)-(S,S)-1 as substrate, (5S,11R,14R)-73me: major
diastereomer before crystallization (83%, ee 93%) and after crystallization (75%, ee 99%),
[α]D20 = +226 (c 0.02, CH2Cl2).
1-(2,8-dimethyl-14-(4-nitrophenyl)-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocin14-yl)ethanone (73mf):
(85 %, major diastereomer) of 73mf was obtained from 100
mg of rac-1. Separation of the diastereomers was done by
column chromatography (SiO2, 3 x 30 cm, acetone/pentane
5/95).
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White solid 145.2 mg (85%). 1H
NMR (400 MHz, CDCl3)  = 8.24 (d, J = 9.1 Hz, 2H), 7.92 – 7.64 (m, 2H), 7.30 (d, J = 8.0
Hz, 1H), 7.01 – 6.95 (m, 2H), 6.86 (d, J = 8.0 Hz, 1H), 6.64 (s, 2H), 5.30 (d, J = 14.5 Hz,
1H), 4.59 (d, J = 17.5 Hz, 1H), 4.54 (d, J = 17.5 Hz, 1H), 4.44 (d, J = 17.6 Hz, 1H), 4.30 (d, J
= 17.6 Hz, 1H), 3.48 (d, J = 14.5 Hz, 1H), 2.19 (s, 3H), 2.15 (s, 3H), 1.64 (s, 3H). 13C NMR
(125 MHz, CDCl3)  = 204.0, 148.8, 147.5, 147.2, 144.0, 136.5, 135.5, 134.9, 130.4, 130.0,
129.1, 128.8, 128.6, 128.3(2C), 128.1, 124.3(2C), 82.93, 59.57, 59.40, 56.47, 25.43, 21.18,
21.09. M.P.: 174-175 °C. IR (neat): -1 = 2916, 1711, 1601, 1520, 1496,1346 cm-1. HRMS
(ESI+) calculated for C26H26N3O3(M+1): 428.1968. Found: 428.1951.
Following procedure II, using (+)-(S,S)-2 as substrate, (5S,11R,14R)-73mf: major
-133-
Experimental Part
1-(2,8-dimethoxy-14-phenyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5] diazocin-14-yl)
ethanone(73mg)
(80%, major diastereomer) of 73mg was obtained from 100
mg of rac-1. Separation of the diastereomers was done by
column chromatography (SiO2, 3 x 30 cm, acetone/pentane
5/95).
Major diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White solid 132.5 mg (80%).. 1H
NMR (400 MHz, CDCl3)  = 7.60 (b, 2H), 7.43 – 7.32 (m, 3H), 7.31 – 7.24 (m, 1H), 7.04 (d,
J = 8.7 Hz, 1H), 6.73 (dd, J = 8.7, 2.9 Hz, 1H), 6.61 (dd, J = 8.7, 2.9 Hz, 1H), 6.37 (d, J = 2.9
(d, J = 17.4 Hz, 1H), 4.40 (d, J = 17.6 Hz, 1H), 4.29 (d, J = 17.4 Hz, 1H), 3.69 (s, 3H), 3.66
(s, 3H), 3.51 (d, J = 14.5 Hz, 1H), 1.67 (s, 3H).
13
C NMR (100 MHz, CDCl3)  = 205.4,
157.8, 156.8, 143.1, 141.4, 140.1, 137.8, 137.5, 131.9, 129.3, 129.1, 127.8, 127.4, 114.5,
113.1, 112.8, 112.6, 82.5, 59.9, 59.6, 56.6, 55.6, 55.5, 25.3. M.P.: 213-215 °C. IR (neat): -1
= 2928, 1705, 1608, 1493, 1462 cm-1. HRMS (ESI+) calculated for C26H27N2O3(M+1):
415.2016. Found: 415.2014.
Following procedure II, using (+)-(S,S) enantiomer of dimethoxy Tröger base as substrate,
(5S,11R,14R)-73mg: major diastereomer (80%, ee 92%), [α]D20 = +210 (c 0.02, CH2Cl2).
(5R,11S,14S)-ethyl2,8-dimethyl-14-phenyl-6,12-dihydro-5,11-ethanodibenzo[b,f][1,5]
diazocine-14-carboxylate (73n)
Following procedure I using Rh2(OAc)4 as catalyst, 82.5 mg
(50%) of 73n was obtained from 100 mg of rac-1; the product
being obtained as a single diastereomer in the crude reaction
mixture which was further purified by column chromatography
(SiO2, 3 x 30 cm, acetone/pentane 7/93). 1H NMR (400 MHz,
CDCl3): 7.71 (d, J = 7.4 Hz, 2H), 7.36 (t, J = 7.6 Hz, 2H),
7.31 – 7.22 (m, 2H), 7.01 (d, J = 8.0 Hz, 1H), 6.92 (d, J = 8.0, 1H), 6.84 (d, J = 8.0 Hz, 1H),
17.7 Hz, 1H), 4.33 (dd, J = 17.7, 13.3 Hz, 1H), 3.84 – 3.72 (m, 2H), 3.60 (dq, J = 10.7, 7.1
Hz, 1H), 2.18 (s, 3H), 2.13 (s, 3H), 0.63 (t, J = 7.1 Hz, 3H). 13C NMR (100 MHz, CDCl3) =
172.8, 147.4, 145.6, 143.0, 136.7, 136.3, 134.9, 134.4, 130.4, 128.9, 128.6, 128.6(2C), 128.2,
-134-
Chapter 7
128.0, 127.8, 127.7, 127.1(2C), 75.7, 62.2, 61.3, 59.2, 56.2, 21.0, 20.9, 13.4. M.P.: 210212 °C. IR (neat): -1 = 2923, 2854, 1723, 1495, 1447, 1376, 1243 cm-1. HRMS (ESI+)
calculated for C33H33N2O2(M+1): 413.2224 Found: 413.2223.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73n (50%, ee 93%), [α]D20 = –335 (c 0.02, CH3CN).
(5R,11S,14S)- or (5S,11R,14R)-methyl 2,8-dimethyl-14-phenyl-6,12-dihydro-5,11ethanodibenzo [b,f] [1,5] diazocine-14-carboxylate (73o)
Following procedure I using Rh2(OAc)4 as catalyst, 120 mg
(75 %, both diastereomers) of 73o was obtained from 100 mg of
rac-1. Separation of the diastereomers was performed by column
Major Diastereomer {(5R,11S,14S) and (5S,11R,14R)}: White
solid 113 mg (71%) 1H NMR (400 MHz, CDCl3) = 7.74 (d, J =
7.8 Hz, 2H), 7.44 – 7.35 (m, 2H), 7.29 (ddd, J = 7.8, 4.5, 2.8 Hz, 2H), 7.03 (d, J = 8.0 Hz,
1H), 6.96 (d, J = 8.0 Hz, 1H), 6.87 (d, J = 8.0 Hz, 1H), 6.67 (s, 1H), 6.62 (s, 1H), 5.23 (d, J =
14.7 Hz, 1H), 4.83 (d, J = 17.5 Hz, 1H), 4.57 (d, J = 17.8 Hz, 1H), 4.41-4.31 (m, J = 15.8 Hz,
2H), 3.82 (d, J = 14.7 Hz, 1H), 3.21 (s, 3H), 2.20 (s, 3H), 2.15 (s, 3H). 13C NMR (100 MHz,
CDCl3): = 173.6, 147.5, 145.5, 142.9, 136.8, 136.4, 135.0, 134.5, 130.3, 129.0, 128.8(2C),
128.7, 128.4, 128.1, 127.9, 127.8, 127.2(2C), 76.0, 62.3, 59.3, 56.2, 52.5, 21.2, 21.1. M.P.:
213-215 °C. IR (neat): -1 = 2921, 1728, 1495, 1446, 1244 cm-1. HRMS (ESI+) calculated
for C26H27N2O2(M+1): 399.2067. Found: 399.2066.
Minor diastereomer: yellow gel 6.3 mg (4%). 1H NMR (400 MHz, CDCl3):  = 7.73 (d, J =
7.9 Hz, 2H), 7.28 (t, J = 7.8 Hz, 1H), 7.19 – 7.06 (m, 3H), 6.97 (d, J = 8.0 Hz, 1H), 6.91 –
6.81 (m, 2H), 6.73 (s, 1H), 6.30 (s, 1H), 4.90 (d, J = 16.0 Hz, 1H), 4.75 (d, J = 18.0 Hz, 1H),
4.62 (d, J = 18.0 Hz, 1H), 4.24 (d, J = 17.2 Hz, 1H), 4.15 (td, J = 16, 17.2 Hz, 2H), 3.61 (s,
3H), 2.19 (s, 3H), 2.04 (s, 3H). 13C NMR (100 MHz, CDCl3): = 174.7, 147.7, 144.0, 138.1,
137.0, 136.7, 134.7, 134.6, 130.5, 128.9, 128.6(2C), 128.4, 128.2, 127.9(2C), 127.6, 127.5,
76.1, 59.6, 58.7, 58.1, 52.9, 21.12, 21.06. IR (neat): -1 = 2923, 1741, 1496, 1434, 1260, 1217
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73o (71%, ee 99%), [α]D20 = –340 (c 0.05, CH2Cl2).
-135-
Experimental Part
Following procedure III, using (+)-(S,S)-1 as substrate: major diastereomer (71%, ee 99%),
[α]D20 = +345 (c 0.05, CH2Cl2). Its absolute configuration (5SN,11RN,14RC) was assigned by
VCD. See the manuscript.
(5R,11S,14S)-methyl 14-(4-methoxyphenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethano
dibenzo [b,f][1,5]diazocine-14-carboxylate (73oa)
(83 %, white solid) of the major diastereomer of 73oa was
obtained from 100 mg of rac-1, after purification by
chromatography (SiO2, 3 x 30 cm, acetone/pentane 3/97). 1H
NMR (400 MHz, CDCl3):  =7.61 (d, J = 8.3 Hz, 2H), 7.267.22 (m, 1H), 6.99 (d, J = 7.5 Hz, 1H), 6.96 – 6.79 (m, 4H),
18.0 Hz, 1H), 4.33 (td, J = 17.8, 18.0 Hz, 2H), 3.78 (s, 3H), 3.76(d, J = 15.0 Hz, 1H), 3.19 (s,
3H), 2.17 (s, 3H), 2.13 (s, 3H). 13C NMR (100 MHz, CDCl3):  = 173.8, 159.3, 147.6, 145.6,
136.9, 136.4, 135.0, 134.9, 134.5, 130.3, 129.0, 128.8, 128.44(2C), 128.40, 128.1, 127.96,
114.1(2C), 75.4, 62.4, 59.3, 56.0, 55.6, 52.5, 21.2, 21.1. M.P.: 220-222 °C. IR (neat): -1 =
2923, 1726, 1604, 1497, 1433, 1246, 1164 cm-1. HRMS (ESI+) calculated for C27H29N2O3
[(M+1)+]: 429.2172, Found: 429.2172.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73oa (83%, ee 98%), [α]D20 = –298.4 (c 0.03, CH3CN).
(5R,11S,14S)-methyl
2,8-dimethyl-14-(naphthalen-2-yl)-6,12-dihydro-5,11-ethano
dibenzo[b,f][1,5] diazocine-14-carboxylate (73ob)
(80 %, white solid) of the major diastereomer of 2g was
NMR (400 MHz, CDCl3): δ = 8.08 – 7.95 (m, 2H), 7.92 –
7.79 (m, 3H), 7.53 – 7.45 (m, 2H), 7.35 (d, J = 8.0 Hz, 1H),
7.04 (d, J = 8.0 Hz, 1H), 6.97 (d, J = 8.0 Hz, 1H), 6.85 (d, J = 8.0 Hz, 1H), 6.68 (s, 1H), 6.59
(s, 1H), 5.34 (d, J = 14.6 Hz, 1H), 4.86 (d, J = 17.4 Hz, 1H), 4.59 (d, J = 17.8 Hz, 1H), 4.39
(td, J = 17.4, 17.8 Hz, 2H), 3.95 (d, J = 14.6 Hz, 1H), 3.20 (s, 3H), 2.20 (s, 3H), 2.13 (s, 3H).
-136-
Chapter 7
13
C NMR (100 MHz, CDCl3):  173.5, 147.6, 145.5, 140.5, 136.7, 136.4, 135.1, 134.6,
133.5, 133.0, 130.4, 129.0, 128.7, 128.6, 128.5(2C), 128.1, 127.9, 127.8, 126.5, 126.4, 126.3,
125.2, 76.10, 62.56, 59.38, 56.27, 52.59, 21.20, 21.07. M.P.: 270-272 °C. IR (neat): -1 =
2920, 1726, 1496, 1433, 1242 cm-1. HRMS (ESI+) calculated for C30H29N2O2 [(M+1)]:
449.2223. Found: 449.2227.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73ob (80%, ee 98%), [α]D20 = –298 (c 0.02, CH2Cl2).
(5R,11S,14S)-methyl
2,8-dimethyl-14-(p-tolyl)-6,12-dihydro-5,11-ethanodibenzo[b,f]
[1,5] diazocine-14-carboxylate (73oc)
(72 %, white solid) of the major diastereomer of 73oc was
NMR (400 MHz, CDCl3): δ = 7.61 (d, J = 8.1 Hz, 2H), 7.28 (d,
J = 6.3 Hz, 1H), 7.19 (d, J = 8.1 Hz, 2H), 7.02 (d, J = 8.0 Hz, 1H), 6.94 (d, J = 8.0 Hz, 1H),
6.85 (d, J = 8.0 Hz, 1H), 6.65 (s, 1H), 6.60 (s, 1H), 5.20 (d, J = 14.7 Hz, 1H), 4.81 (d, J =
17.4 Hz, 1H), 4.56 (d, J = 17.7 Hz, 1H), 4.34 (td, J = 17.4, 17.7 Hz, 2H), 3.80 (d, J = 14.7
Hz, 1H), 3.21 (s, 3H), 2.36 (s, 3H), 2.19 (s, 3H), 2.15 (s, 3H). 13C NMR (100 MHz, CDCl3) 
= 173.7, 147.5, 145.5, 139.8, 137.5, 136.7, 136.3, 134.8, 134.4, 130.2, 129.4(2C), 128.9,
128.6, 128.3, 128.0, 127.8, 127.0(2C), 75.6, 62.3, 59.2, 56.0, 52.4, 21.3, 21.1, 21.0. M.P.:
195-197 °C. IR (neat): -1 = 2922, 1727, 1496, 1432, 1337, 1244 cm-1. HRMS (ESI+)
calculated for C27H29N2O2[(M+1)]: 413.2223, Found: 413.2228.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73oc (72%, ee 98%), [α]D20 = –276 (c 0.05, CH3CN).
(5R,11S,14S)-methyl
2,8-dimethyl-14-(m-tolyl)-6,12-dihydro-5,11-ethanodibenzo[b,f]
[1,5]diazocine-14-carboxylate (73od)
Following procedure I using Rh2(OAc)4 as catalyst, 115 mg (70 %,
white solid) of the major diastereomer of 73od was obtained from
100 mg of rac-1, after purification by chromatography (SiO2, 3 x
30 cm, acetone/pentane 3/97). 1H NMR (400 MHz, CDCl3):  =
-137-
Experimental Part
7.59 (d, J = 7.7 Hz, 1H), 7.53 (s, 1H), 7.37 – 7.28 (m, 2H), 7.14 (d, J = 7.7 Hz, 1H), 7.06 (d, J
= 8.0 Hz, 1H), 6.99 (d, J = 8.0 Hz, 1H), 6.90 (d, J = 8.0 Hz, 1H), 6.70 (s, 1H), 6.65 (s, 1H),
5.24 (d, J = 14.7 Hz, 1H), 4.85 (d, J = 17.5 Hz, 1H), 4.61 (d, J = 17.7 Hz, 1H), 4.39 (td, J =
17.5, 17.7 Hz, 1H), 3.85 (d, J = 14.7 Hz, 1H), 3.25 (s, 3H), 2.44 (s, 3H), 2.24 (s, 3H), 2.19 (s,
3H).
13
C NMR (100 MHz, CDCl3):  = 173.7, 147.6, 145.6, 142.8, 138.3, 136.8, 136.4,
134.9, 134.5, 130.3, 129.0, 128.7, 128.6, 128.59, 128.4, 128.1, 127.9, 127.7, 124.3, 75.9,
62.4, 59.3, 56.2, 52.5, 22.0, 21.2, 21.1. M.P.: 170-172 °C. IR (neat): -1 = 2923, 1727, 1496,
1432, 1247 cm-1. HRMS (ESI+) calculated for C27H29N2O2[(M+1)]: 413.2223. Found:
413.2220.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73od (70%, ee 98%), [α]D20 = –284 (c 0.05, CH3CN).
(5R,11S,14S)-methyl 2,8-dimethyl-14-(4-nitrophenyl)-6,12-dihydro-5,11-ethanodibenzo
[b,f] [1,5]diazocine-14-carboxylate (73of)
(82 %) of the major diastereomer of 73of was obtained from
100 mg of rac-1, after purification by chromatography (SiO2,
3 x 30 cm, acetone/pentane 5/95. 1H NMR (400 MHz,
CDCl3):  = 8.23 (d, J = 8.3 Hz, 2H), 7.89 (d, J = 8.3 Hz, 2H),
7.26 (d, J = 8.0 Hz, 1H), 7.01 (d, J = 8.0 Hz, 1H), 6.95 (d, J = 8.0 Hz, 1H), 6.86 (d, J = 8.0
Hz, 1H), 6.65 (s, 1H), 6.60 (s, 1H), 5.22 (d, J = 14.6 Hz, 1H), 4.78 (d, J = 17.5 Hz, 1H), 4.51
– 4.23 (m, 3H), 3.76 (d, J = 14.6 Hz, 1H), 3.20 (s, 3H), 2.18 (s, 3H), 2.14 (s, 3H). 13C NMR
(100 MHz, CDCl3): = 172.3, 150.1, 147.6, 147.2, 144.6, 136.3, 136.1, 135.6, 134.9, 130.3,
129.1, 128.7, 128.6, 128.4, 128.3(2C), 128.0, 124.1(2C), 76.4, 62.0, 59.3, 56.2, 52.9, 21.2,
21.1. ). M.P: 218-220 °C IR (neat): -1 = 2923, 1731, 1604, 1519, 1497, 1433, 1347, 1244
cm-1. HRMS (ESI+) calculated for C26H26N3O4[(M+1)]: 444.1917. Found: 444.1915.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73of (50%, ee 64%), [α]D20 = –123 (c 0.02, CH3CN).
-138-
Chapter 7
(5R,11S,14S)-methyl 14-(4-chlorophenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo
[b,f] [1,5] diazocine-14-carboxylate (73oh)
(78 %, white solid) of the major diastereomer of 73oh was
NMR (400 MHz, CDCl3):  = 7.65 (d, J = 8.4 Hz, 2H), 7.34
(d, J = 8.8 Hz, 2H), 7.24 (d, J = 8 Hz, 1H), 7.01 (d, J = 8.0 Hz, 1H), 6.93 (d, J = 8.0 Hz, 1H),
6.85 (d, J = 8.0 Hz, 1H), 6.64 (s, 1H), 6.60 (s, 1H), 5.17 (d, J = 14.6 Hz, 1H), 4.78 (d, J =
17.5 Hz, 1H), 4.48 (d, J = 17.7 Hz, 1H), 4.32 (td, J = 17.5, 17.7 Hz, 2H), 3.75 (d, J = 14.6
Hz, 1H), 3.20 (s, 3H), 2.18 (s, 3H), 2.14 (s, 3H).
13
C NMR (100 MHz, CDCl3): = 173.2,
147.4, 145.2, 141.4, 136.5, 136.3, 135.2, 134.7, 133.8, 130.3, 129.0, 129.0(2C), 128.7,
128.7(2C), 128.5, 128.2, 128.0, 75.69, 62.15, 59.28, 56.08, 52.65, 21.17, 21.07. M.P.: 203205 °C. IR (neat): -1 = 2922, 1728, 1489, 1433, 1399, 1245 cm-1. HRMS (ESI+) calculated
for C26H26N2O2Cl[(M+1)]: 433.1677. Found: 433.1681.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73oh (78%, ee 97%), [α]D20 = –255.1 (c 0.02, CH3CN).
(5R,11S,14S)-methyl14-(3-chlorophenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo
[b,f][1,5]diazocine-14-carboxylate (73oi)
Following procedure I using Rh2(OAc)4 as catalyst, 130 mg (75 %,
white solid) of the major diastereomer of 73oi was obtained from
100 mg of rac-1, after purification by chromatography (SiO2, 3 x
30 cm, acetone/pentane 5/95). 1H NMR (400 MHz, CDCl3): =
7.65 (d, J = 8.6 Hz, 2H), 7.34 (d, J = 8.6 Hz, 2H), 7.25 (s, 1H),
7.01 (d, J = 8.0 Hz, 1H), 6.94 (d, J = 8.0 Hz, 1H), 6.86 (d, J = 8.0 Hz, 1H), 6.65 (s, 1H), 6.60
(s, 1H), 5.17 (d, J = 14.6 Hz, 1H), 4.78 (d, J = 17.5 Hz, 1H), 4.49 (d, J = 17.7 Hz, 1H), 4.33
(td, J = 17.5, 17.7 Hz, 2H), 3.76 (d, J = 14.6 Hz, 1H), 3.20 (s, 3H), 2.18 (s, 3H), 2.14 (s, 3H).
13
C NMR (100 MHz, CDCl3)  = 173.2, 147.4, 145.2, 141.4, 136.5, 136.3, 135.2, 134.7,
133.8, 130.3, 129.0, 128.9(2C), 128.7, 128.6(2C), 128.5, 128.2, 128.0, 75.68, 62.14, 59.27,
56.07, 52.64, 21.16, 21.06. M.P: 200-202 °C. IR (neat): -1 = 2923, 1728, 1496, 1433, 1399,
-139-
Experimental Part
1244, 1219 cm-1. HRMS (ESI+) calculated for C26H26N2O2Cl[(M+1)]: 433.1677, found:
433.1674.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73oi (75%, ee 97%), [α]D20 = –262.6 (c 0.03, CH3CN).
(5R,11S,14S)-methyl
14-(4-bromophenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethano
dibenzo[b,f][1,5]diazocine-14-carboxylate (73oj)
(75 %, white solid) of the major diastereomer of 73oj was
NMR (400 MHz, CDCl3):  = 7.59 (d, J = 8.4 Hz, 2H), 7.49
(d, J = 8.8 Hz, 2H), 7.24 (d, J = 8.0 Hz, 1H), 7.00 (d, J = 8.0 Hz, 1H), 6.93 (d, J = 8.0 Hz,
1H), 6.85 (d, J = 8.0 Hz, 1H), 6.64 (s, 1H), 6.60 (s, 1H), 5.16 (d, J = 14.6 Hz, 1H), 4.77 (d, J
= 17.5 Hz, 1H), 4.48 (d, J = 17.7 Hz, 1H), 4.32 (td, J = 17.5, 17.7 Hz, 1H), 3.74 (d, J = 14.6
Hz, 1H), 3.20 (s, 3H), 2.18 (s, 3H), 2.14 (s, 3H).
13
C NMR (100 MHz, CDCl3):  = 173.1,
147.4, 145.1, 142.0, 136.5, 136.3, 135.2, 134.7, 131.9(2C), 130.3, 129.06(2C), 129.02, 128.7,
128.5, 128.2, 128.0, 122.1, 75.76, 62.10, 59.28, 56.08, 52.67, 21.18, 21.07. M.P: 190-192 °C.
IR (neat): -1 = 2922, 1728, 1587, 1495, 1485, 1433, 1395, 1244, 1217 cm-1. HRMS (ESI+)
calculated for C26H26N2O2Br[(M+1)]: 477.1172. Found: 477.1161.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73oj (75%, ee 97%), [α]D20 = –220.9 (c 0.02, CH3CN).
(5R,11S,14S)-methyl
2,8-dimethyl-14-(4-(trifluoromethyl)phenyl)-6,12-dihydro-5,11-
ethanodibenzo[b,f][1,5]diazocine-14-carboxylate (73ok)
(70 %) of the major diastereomer of 73ok was obtained from
100 mg of rac-1, after purification by chromatography (SiO2,
3 x 30 cm, acetone/pentane 5/95). 1H NMR (400 MHz,
CDCl3):  = 7.85 (d, J = 8.4 Hz, 2H), 7.66 (dd, J = 8.4 Hz,
2H), 7.26 (d, J = 8.0 Hz, 1H), 7.02 (d, J = 8.0 Hz, 1H), 6.95 (d, J = 8.0 Hz, 1H), 6.86 (d, J =
8.0 Hz, 1H), 6.66 (s, 1H), 6.61 (s, 1H), 5.22 (d, J = 14.6 Hz, 1H), 4.79 (d, J = 17.5 Hz, 1H),
4.46 (d, J = 17.7 Hz, 1H), 4.35 (d, J = 17.5, 17.7 Hz, 1H), 3.77 (d, J = 14.6 Hz, 1H), 3.21 (s,
-140-
Chapter 7
3H), 2.19 (s, 3H), 2.15 (s, 3H). 13C NMR (100 MHz, CDCl3) = 172.9, 147.4, 146.9, 145.0,
136.4, 136.3, 135.3, 134.8, 130.3, 129.1, 128.7, 128.5, 128.3, 128.0, 127.6(2C), 125.8, 125.7,
125.7, 76.13, 62.17, 59.31, 56.19, 52.78, 21.19, 21.07.
19
F NMR (282 MHz, CDCl3)  = -
62.55. IR (neat): -1 = 2925, 1731, 1616, 1497, 1409, 1324 cm-1. HRMS (ESI+) calculated
for C27H26N2O2F3[(M+1)]: 467.1940, Found: 467.1950.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73ok (70%, ee 97%), [α]D20 = –298 (c 0.02, CH3CN).
(5R,11S,14S)-methyl 14-(4-fluorophenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethanodibenzo
[b,f][1,5]diazocine-14-carboxylate (73ol)
(50%) of the major diastereomer of 73ol was obtained from 100
mg of rac-1, after purification by chromatography (SiO2, 3 x 30
cm, acetone/pentane 5/95). 1H NMR (400 MHz, CDCl3) δ =
7.69 (dd, J = 8.4, 5.5 Hz, 2H), 7.23 (d, J = 10.2 Hz, 1H), 7.11 –
6.97 (m, 3H), 6.93 (d, J = 8.0 Hz, 1H), 6.86 (d, J = 8 Hz, 1H), 6.65 (s, 1H), 6.61 (s, 1H), 5.18
(d, J = 14.6 Hz, 1H), 4.78 (d, J = 17.5 Hz, 1H), 4.50 (d, J = 17.7 Hz, 1H), 4.32 (td, J = 17.5,
17.7 Hz, 2H), 3.77 (d, J = 14.6 Hz, 1H), 3.21 (s, 3H), 2.19 (s, 3H), 2.14 (s, 3H). 13C NMR
(125 MHz, CDCl3) δ = 173.5, 163.7, 161.3, 147.5, 145.3, 138.6, 136.6, 136.3, 135.2, 134.7,
130.3, 129.1, 129.0, 128.9, 128.7, 128.5, 128.2, 128.0, 115.8, 115.6, 75.6, 62.3, 59.3, 56.0,
52.6, 21.2, 21.1. 19F NMR (282 MHz, CDCl3)  = -114.89. IR (neat): -1 = 2923, 1729, 1497,
1433, 1258, 1157 cm-1. HRMS (ESI+) calculated for C26H26N2O2F[(M+1)]: 417.1972.
Found: 417.1964.
Following procedure III, using (–)-(R,R)-1 as substrate: major diastereomer (5R,11S,14S)73ol (50%, ee 97%), [α]D20 = –283 (c 0.02, CH3CN).
(5R,11S,14S)-methyl
2,8-dimethoxy-14-(4-nitrophenyl)-6,12-dihydro-5,11-ethano
dibenzo [b,f][1,5]diazocine-14-carboxylate (75)
Following procedure I using Rh2(OAc)4 as catalyst, using
dimethoxy tröger base (113 mg, 0.4 mmol), the major
diastereomer of 75 (153 mg, 89%) was obtained after
purification
by
chromatography
-141-
(SiO2,
3
x
30
cm,
Experimental Part
acetone/pentane 10/90). 1H NMR (400 MHz, CDCl3):  = 7.69 (d, J = 7.7 Hz, 2H), 7.36 (t, J
= 7.6 Hz, 2H), 7.33 – 7.26 (m, 2H), 7.05 (d, J = 8.7 Hz, 1H), 6.70 (dd, J = 8.7, 2.8 Hz, 1H),
6.60 (dd, J = 8.7, 2.8 Hz, 1H), 6.34 (td, J = 8.7, 2.8 Hz, 2H), 5.20 (d, J = 14.6 Hz, 1H), 4.80
(d, J = 17.5 Hz, 1H), 4.55 (d, J = 17.8 Hz, 1H), 4.31 (td, J = 17.5, 17.8 Hz, 2H), 3.78 (d, J =
14.6 Hz, 1H), 3.69 (s, 3H), 3.64 (s, 3H), 3.20 (s, 3H).
13
C NMR (100 MHz, CDCl3):  =
173.6, 157.2, 156.9, 143.1, 142.8, 140.9, 138.3, 138.1, 131.8, 129.2, 128.8(2C), 127.9,
127.2(2C), 113.9, 113.1, 112.6, 112.1, 76.2, 62.6, 59.6, 56.4, 55.5, 55.5, 52.6. M.P: 231233 °C. IR (neat): -1 = 2923, 1726, 1607, 1495, 1446, 1258, 1235 cm-1. HRMS (ESI+)
calculated for C26H27N2O4[(M+1)]: 431.1965, Found: 431.1977.
Following procedure III, using (–)-(R,R)-dimethoxy Tröger base (ee 98%) as substrate:
major diastereomer (5R,11S,14S)-75 (89%, ee 97%), [α]D20 = –335 (c 0.02, CH3CN).
(5R,11S,14S)-methyl 2-methoxy-8-nitro-14-phenyl-6,12-dihydro-5,11-ethanodibenzo[b,f]
[1,5]diazocine-14-carboxylate (80):
Following procedure IV, using (25 mg, 0.084 mmol), the major
isomer 80 (27 mg, 71%), was obtained after purification by
chromatography (SiO2, 3 x 30 cm, acetone/pentane 15/85 1H
NMR (400 MHz, CDCl3):  = 7.96 (d, J = 8.8 Hz, 1H), 7.75 (s,
1H), 7.49 (d, J = 8.8 Hz, 1H), 7.04 (d, J = 8.7 Hz, 1H), 6.64 (d, J
= 8.7 Hz, 1H), 6.40 (s, 1H), 5.16 (d, J = 18.5 Hz, 1H), 4.89 (d, J = 17.9 Hz, 1H), 4.70 (d, J =
15.3 Hz, 1H), 4.50 (d, J = 18.5 Hz, 1H), 4.35 (d, J = 17.9 Hz, 1H), 4.30 – 4.16 (m, 2H), 4.03
– 3.83 (m, 2H), 3.76 (d, J = 15.3 Hz, 1H), 3.68 (s, 3H), 1.25 (t, J = 7.1 Hz, 3H), 0.82 (t, J =
7.1 Hz, 3H).
13
C NMR (100 MHz, CDCl3):  = 168.3, 168.1, 157.5, 153.7, 145.8, 141.7,
139.2, 137.2, 131.7, 129.7, 124.4, 122.9, 113.8, 112.4, 78.3, 62.8, 62.7, 59.5, 59.3, 57.5, 55.7,
14.4, 13.8. M.P: 210-212 °C. IR (neat): -1 = 2962, 2931, 1735, 1579, 1499, 1342, 1256,
1218 cm-1.HRMS (ESI+) calculated for C39H38N2O2Na [(M+Na)+]: 478.1585. Found:
478.1585.
-142-
Chapter 7
7.5
Synthesis of Chiral Selector TB-CSP:
(5S,11R,14R)-methyl
14-(4-(allyloxy)phenyl)-2,8-dimethyl-6,12-dihydro-5,11-ethano
dibenzo[b,f][1,5]diazocine-14-carboxylate (88)
Following procedure I using Rh2(OAc)4 as catalyst, using (+)1 (500 mg, 0.4 mmol), the major diastereomer of 88 (635.5
mg, 70%) was obtained after purification by chromatography
(SiO2, 3 x 30 cm, acetone/pentane 10/90). 1H NMR (400
MHz, CDCl3):  = 7.60 (d, J = 8.2 Hz, 2H), 7.24 (d, J = 8.2
Hz, 1H), 7.09 – 6.78 (m, 5H), 6.61 (d, J = 8.2 Hz, 2H), 6.16 – 6.04 (m, 1H), 5.41 (d, J = 17.2
– 4.46 (m, 2H), 4.42 – 4.20 (m, 2H), 3.78 (d, J = 14.3 Hz, 1H), 3.19 (s, 3H), 2.17 (s, 3H),
2.13 (s, 3H). 13C NMR (100 MHz, CDCl3):  = 173.8, 158.4, 147.6, 145.6, 136.9, 136.4,
135.0, 134.6, 133.6, 130.6, 130.3, 129.0, 128.8, 128.4, 128.1, 127.9, 118.1, 115.2, 114.9,
69.2, 62.4, 59.3, 56.0, 52.5, 21.2, 21.1. M.P: 231-233 °C. IR (neat): -1 = 2923, 1720, 1609,
1515, 1444, 1251, 1205 cm-1. HRMS (ESI+) calculated for C29H30N2O3[(M+1)]: 454.5615,
Found: 454.5671.
Following procedure III, using (+)-(S,S)-1 as substrate: major diastereomer (5S,11R,14R)-88
(70%, ee 98%), [α]D20 = +244.1 (c 0.02, CH3CN).
Synthesis of TB-CSP
A 250-mL round bottom flask was charged
with 3 grams mercaptoproyl silica (3 µm)
(1.6 wt% of Sulfur) (hexamethyldisilazan
(11 R)
MeOHHPLC
reflux
AIBN
(5 S)
(14 R)
endcapped) and equipped with a mechanical
stirrer. 30 ml MeOHHPLC was added and the
suspension was stirred for 10 min under a
constant N2 flow. 0.5 g (5S,11R,14R)-88 and 15 mg AIBN were dissolved in 10 ml
MeOHHPLC.and added under constant N2 counter stream using a Pasteur pipette. The reaction
was stirred under reflux and nitrogen for 12 h. The modified silica was transferred to a glass
filter and washed 3 times with methanol, transferred again into the round bottom flask and
-143-
Experimental Part
refluxed for 15 min. After filtration and a final wash with 3 times methanol and 3 times
DCMHPLC (CH2Cl2) the silica was dried for 12 h at 60°C using a vacuum dry box.
Finally, the packing material was slurry packed at 600 bar into a 250 x 3 mm stainless steel
column.
CHNS analysis was performed to estimate the selector coverage.
wt% of Carbon
1578
1574
wt% of Hydrogen
2.41
2.29
wt% of Nitrogen
0.856
0.913
Amount of Sulfur = 514 µmol S/g silica.
Amount of 88 = 316 µmol TB/g silica.
Net immobilization of 88 on silica: 61% coverage.
-144-
wt% of Sulfur
1.60
1.69
Chapter 7
7.6
Characterization of 80 by NMR Analysis:
COSY Experiment (CDCl3, 500 MHz)
9
7 10
4
3
1
12
6 14 12 6
-145-
14
Experimental Part
NOESY Experiment (CDCl3, 500 MHz)
7
10
4
3
1
12
9
7 10
4
3
1
12
6 14 12 6
14
9
-146-
6
14
12 6
14
Chapter 7
HSQC Experiment (CDCl3, 500 MHz)
-147-
Experimental Part
HSQC Experiment (CDCl3, 500 MHz)
(Expended parts)
12
6
14
12
6
14
12
14
6
9
7
10
4
1
3
9
7
4
10
-148-
3
1
Chapter 7
HMBC Experiment (CDCl3, 500 MHz)
9
7
10
4
3
1
12
6 14 12 6
14
2J(C13−H14)
13
3J(C13−H12)
-149-
Experimental Part
7.7
Characterization of Salt 85
1
H Experiment (CDCl3, 500 MHz)
7
1
COSY Experiment (CDCl3, 500 MHz)
7
1
1
7
-150-
Chapter 7
HMBC Experiment (CDCl3, 500 MHz)
7
1
N+-Me
12
-151-
Experimental Part
7.8
CSP-HPLC Data
Resolution of Tröger bases-1:
S. Satishkumar and M. Periasamy,
Tetrahedron: Asymmetry 2006, 17, 1116
CSP-HPLC: whelk O1 column, n-Hexane/i-PrOH 90/10%, 1 ml/min, 23 oC, λ = 254 nm
Resolution of dimethoxyTröger bases:
S. Satishkumar and M. Periasamy
Tetrahedron: Asymmetry 2009, 20, 2257
CSP-HPLC: HPLC: OD-H column, n-Hexane/i-PrOH 90/10%, 1 ml/min, 23 oC, λ = 254 nm
-152-
Chapter 7
Compound 73a
73a
CSP-HPLC: IC column, n-Hexane/i-PrOH 80/20%, 1 ml/min, 23 oC, λ = 254 nm
Compound 73f
73f
-153-
Experimental Part
Compound 73m
73m
CSP-HPLC: Whelk O1 column, n-Hexane/i-PrOH 99/1%, 0.7 ml/min, 23 oC, λ = 254 nm
Compound 73ma
73ma
-154-
Chapter 7
Compound 73mb
73mb
Compound 73mc
73mc
-155-
Experimental Part
Compound 73md
73md
Compound 73me
73me
-156-
Chapter 7
Compound 73mf
73mf
Compound 73mg
73mg
-157-
Experimental Part
Compound 73o
73o
CSP-HPLC: whelk O1 column, n-Hexane/i-PrOH 99.5/0.5%, 0.5 ml/min, 23 oC, λ = 254 nm
Compound 73oa
73oa
CSP-HPLC: Whelk O1 column, n-Hexane/i-PrOH 99/1%, 1 ml/min, 23 oC, λ = 254 nm
-158-
Chapter 7
Compound 73ob
73ob
CSP-HPLC: Whelk O1 column, n-Hexane/i-PrOH 99/1%, 1 ml/min, 23 oC, λ = 254 nm
Compound 73oc
73oc
CSP-HPLC: Whelk O1 column, n-Hexane/i-PrOH 99.3/0.7%, 1 ml/min, 23 oC, λ = 254 nm
-159-
Experimental Part
Compound 73od
73od
CSP-HPLC: Whelk O1 column, n-Hexane/i-PrOH 99.5/0.5%, 1 ml/min, 23 oC, λ = 254 nm
Compound 73of
73of
CSP-HPLC: OD-H column, n-Hexane/i-PrOH 95/5%, 1 ml/min, 23 oC, λ = 254 nm
-160-
Chapter 7
Compound 73oh
73oh
CSP-HPLC:Whelk O1 column, n-Hexane/i-PrOH 99.5/0.5%, 0.5 ml/min, 23 oC, λ = 254 nm
Compound 73oi
73oi
CSP-HPLC:Whelk O1 column, n-Hexane/i-PrOH 99.5/0.5%, 0.5 ml/min, 23 oC, λ = 254 nm
-161-
Experimental Part
Compound 73oj
73oi
CSP-HPLC: OD-H column, n-Hexane/i-PrOH 95/5%, 1 ml/min, 23 oC, λ = 254 nm
Compound 73ok
73ok
CSP-HPLC: IC column, n-Hexane/i-PrOH 99/1%, 0.5 ml/min, 23 oC, λ = 254 nm
-162-
Chapter 7
Compound 73ol
73ol
Compound 75
75
CSP-HPLC: OD-H column, n-Hexane/i-PrOH 99/1%, 0.7 ml/min, 23 oC, λ = 254 nm
-163-
Experimental Part
7.9
Crystallographic Data
Summary of crystal data, intensity measurement and structure refinement for:
7.8.1
Compound rac-64
Compound
formula
64
C25H24N2O
mol. wt
cryst. Syst.
Space group
a (Å)
b (Å)
c (Å)
(deg)
(deg)
(deg)
V (Å3)
Z
color
crystal dim. (mm)
Dcalc (gcm-3)
F000
(mm-1)
trans. Min.
T (K)
and max
scan
mode
hkl limits
 limits (deg)
num. of data meas.
num. of data with I > 4 (I)
num. of var.
R
Rw
GOF
-3
Largest peak in final diff. (eÅ )
368.5
tetragonal
P-421c
23.9934(10)
23.9934(10)
6.8262(2)
90
90
90
3929.7(4)
8
yellow
0.28×0.25×0.18
1.246
0.076
150
'phi scans'
-29,29/-29,29/-8,8
7.4/51.6
22803
1718
6.8
0.030
0.028
1.00(2)
-0.25 , 0.23
-164-
Chapter 7
7.8.2




Compound 67
Crystal Data
Formula
Triclinic, Pī
a (Å)
b (Å)
c (Å)
(deg)
(deg)
(deg)
V (Å3)
Z
color
crystal dim. (mm)
Dcalc (Mgm-3)
F000
(mm-1)
T (K)
scan mode
hkl limits
Independent reflections
num. of data meas.
num. of data with I > 4 (I)
num. of var.
R
Rw
GOF
Largest peak in final diff.
-3
(eÅ )
C25H25ClN2
8.5252 (7)
9.2990 (8)
13.3113 (11)
89.316 (6)
81.513 (6)
81.019 (7)
1030.85 (15)
2
Cube, orange
0.5×0.5×0.5
1.253
412
0.2
293

-11,11/-13,13/-18,18
5980
0.046
0.065
-165-

Experimental Part
7.8.3
Compound 73e
-166-
Chapter 7
7.8.4
Compound 73g
-167-
Experimental Part
7.8.5
Compound 73m
Crystal data and structure refinement for 73m
Identification code
Empirical formula
Formula moiety
shelxl
C26 H26 N2O
C26H26 N2O
Formula weight
Temperature
Wavelength
Crystal system
Space group
Unit cell dimensions
Volume
382.49
180(2) K
1.54184 Å
Monoclinic
C 2/c
a = 17.0172(3) Å, = 90°.
b = 8.60512(13) Å, = 91.4359(17)°.
c = 28.3901(5) Å,  = 90°
4156.01(12) Å3
Z
Density (calculated)
Absorption coefficient
F(000)
Crystal size
Theta range for data collection
Index ranges
Reflections collected
8
1.223 Mg/m3
0.578 mm-1
1632
0.2416 x 0.1364 x 0.1012 mm3
3.11 to 74.26°.
-17<=h<=21, -10<=k<=10, -35<=l<=35
22460
Completeness to theta = 74.26°
Absorption correction
Max. and min. transmission
Refinement method
Data / restraints / parameters
Goodness-of-fit on F2
Final R indices [I>2sigma(I)]
R indices (all data)
4161 [R(int) = 0.0242]
98.3 %
Analytical
0.952 and 0.901
Full-matrix least-squares on F2
4161 / 0 / 264
1.025
R1 = 0.0406, wR2 = 0.1118
R1 = 0.0457, wR2 = 0.1174
0.197 and -0.183 e.Å-3
Largest diff. peak and hole
-168-
Chapter 7
7.8.6
Compound 73me
Crystal data and structure refinement for 73me
Identification code
Empirical formula
Formula weight
Temperature
as1033
C26 H25 Cl N2 O
416.93
180(2) K
Wavelength
Crystal system
Space group
1.54184 Å
Monoclinic
P 21
a = 8.6155(2) Å, = 90°
b = 27.9341(4) Å, = 115.704(3)°
c = 9.8713(2) Å,  = 90°
2140.59(8) Å3
4
1.294 Mg/m3
1.727 mm-1
Volume
Z
F(000)
Crystal size
Index ranges
Refinement method
Absolute structure parameter
880
0.6873 x 0.3184 x 0.0836 mm3
3.16 to 73.42°.
-10<=h<=10, -34<=k<=33, -12<=l<=11
12251
7210 [R(int) = 0.0192]
94.4 %
Analytical
0.867 and 0.535
7210 / 1 / 548
1.112
R1 = 0.0525, wR2 = 0.1549
R1 = 0.0526, wR2 = 0.1550
0.061(18)
0.379 and -0.274 e.Å-3
-169-
Experimental Part
7.8.7
Compound 73o
Crystal data and structure refinement for 73o
Identification code
Empirical formula
Formula weight
AS721
C26 H26 N2 O2
398.49
Temperature
Wavelength
Crystal system
Space group
180(2) K
0.71073 Å
Monoclinic
P 21
a = 8.4867(7) Å, = 90°
b = 29.108(2) Å, = 115.153(6)°
c = 9.4608(8) Å,  = 90
2115.5(3) Å3
4
1.251 Mg/m3
0.079 mm-1
Volume
Z
F(000)
Crystal size
Index ranges
Refinement method
848
0.45 x 0.4 x 0.3 mm3
2.38 to 26.74°.
-10<=h<=10, -36<=k<=35, -11<=l<=11
21210
8812 [R(int) = 0.0338]
99.4 %
Numerical
0.9758 and 0.9755
8812 / 1 / 658
0.965
R1 = 0.0333, wR2 = 0.0786
R1 = 0.0397, wR2 = 0.0812
0.2(7)
0.133 and -0.165 e.Å-3
-170-
Chapter 7
7.8.8
Compound 73o
Crystal data and structure refinement for 73o (for 2 molecules of 73o)
Identification code
Empirical formula
Formula weight
Temperature
as758bis
C52 H50 Cl2 N4 O4
865.86
180(2) K
Wavelength
Crystal system
Space group
0.71073 Å
Monoclinic
P 21
a = 8.6604(6) Å
b = 27.9874(16) Å
c = 10.0032(7) Å
2190.9(2) Å3
2
1.313 Mg/m3
0.200 mm-1
Volume
Z
F(000)
Crystal size
Index ranges
Refinement method
Extinction coefficient
912
0.30 x 0.26 x 0.14 mm3
1.46 to 25.62°.
-10<=h<=10, -33<=k<=33, -12<=l<=12
19130
8232 [R(int) = 0.0627]
99.6 %
Numerical
0.9773 and 0.9422
8232 / 1 / 567
1.077
R1 = 0.0922, wR2 = 0.2671
R1 = 0.0947, wR2 = 0.2685
0.03(14)
0.030(3)
0.670 and -0.557 e.Å-3
-171-

Thesis Reference - Archive ouverte UNIGE

Transcription

Similar documents

our product brochure

DuraSIL-SL - Best Materials

Receptor radio pentru acces autovehicule Rosslare AY-L23C

RFS Brochure_5044 Series_New Address_06-12-13.fm

A Better Sunscreen: Structural Effects on Spectral Properties

Service Manual

Reaction product analysis by HPLC-SPE

TTMAV-700C Tower Mounted Amplifier Twin Variable Gain

Supporting Information Metal-Ion Catalysis in Alkaline Ethanolysis of