Asymmetric copper-catalyzed alkylations and autocatalysis Pellegrini, Tilde
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Pellegrini, T. (2019). Asymmetric copper-catalyzed alkylations and autocatalysis. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Asymmetric copper-catalyzed alkylations and autocatalysis
The work described in this thesis was carried out at Stratingh Intitute for Chemistry, University of Groningen (The Netherlands)
This work was financially supported by Ministry of Education, Culture and Science (Gravitation program 024.001.035) and NWO
Printed by Ridderprint BV, Ridderkerk, The Netherlands Cover picture by Giulia Leonetti
Baracoa (Cuba), 2015
ISBN: 978-94-6375-291-6 (Printed Book) ISBN: 978-94-034-1429-4 (Ebook)
Asymmetric copper-catalyzed
alkylations and autocatalysis
PhD thesis
to obtain the degree of PhD at the University of Groningen
on the authority of the Rector Magnificus prof. E. Sterken
and in accordance with
the decision by the College of Deans. This thesis will be defended in public on
Friday 8 March 2019 at 12.45 hours
by
Tilde Pellegrini
born on 10 August 1989 in Florence, Italy
Supervisors
Prof. S.R. Harutyunyan Prof. W.R. Browne
Assessment Committee
Prof. A.J. Minnaard Prof. M. Pineschi
Table of content
List of abbreviations ... 1
Chapter 1: Introduction ... 3
1.1. Catalysis in asymmetric syntheses ... 4
1.2. Asymmetric metallic catalysis ... 4
1.3. Asymmetric organocatalysis ... 5
1.4. Catalytic (dynamic) kinetic resolution ... 6
1.5. Non-linear effects in asymmetric catalysis ... 8
1.5.1. The model ML2 or (ML)2 ... 9
1.6. Thesis outline... 10
1.7. Bibliography ... 11
Chapter 2: Control of enantioselectivity in the addition of Grignard reagents to symmetric heteroaryl disubstituted olefins ... 13
2.1. Introduction ... 14
2.1.1. Asymmetric addition of organometallic reagents to electron deficient olefins ... 14
2.1.2. Enantioselectivity in the 1,4-addition of nucleophiles to symmetric disubstituted alkenes ... 15
2.1.3. Copper(I)-catalyzed asymmetric addition of Grignard reagents to (N)-containing heteroaryl alkenes ...19
2.2. Aim ... 21
2.3. Results and discussion ... 21
2.3.1. Synthesis of 1,2-disubstituted heteroaryl alkenes ... 21
2.3.2. ACA of Grignard reagents to benzoxazyl alkenes ... 25
2.3.3. ACA of Grignard reagents to symmetric 2-quinoyl alkenes ... 31
2.4. Conclusions ... 32
2.5. Experimental section ... 33
2.5.1. General information ... 33
2.5.2. Synthesis of substrates ... 34
2.5.3. Catalytic asymmetric addition to 16 ... 35
2.5.4. Complexes CuBr·L10 and CuBr·L11 ... 38
2.6. Bibliography ... 39
Chapter 3: Asymmetric conjugate addition of Grignard reagents to symmetric bispyridyl alkenes ... 43
3.1. Introduction ... 44
3.1.1. Importance of pyridines in medicinal chemistry ... 44
3.1.2. Utilization of bispyridyl compounds in chemistry ... 45
3.1.3. Asymmetric conjugate addition to vinyl pyridines ... 48
3.2. Aim ... 51
3.3. Results and discussion ... 51
3.3.1. Asymmetric addition to symmetric 4-pyridyl alkenes ... 51
3.3.2. Asymmetric addition to symmetric 2-pyridyl alkenes ... 58
3.3.3. Selectivity in the addition to 4-pyridyl, 2-pyridyl and 2-benzoxazyl alkenes ...61
3.3.4. Interaction of 1,2-bis(4-pyridyl)ethene (19) with TMSBr ... 65
3.4. Conclusions ... 67
3.5. Experimental section ... 67
3.5.1. General information ... 67
3.5.2. Synthesis of substrates ... 68
3.5.3. General procedure for the synthesis of 2-(benzoxazol-2-yl)-1-(pyridinyl)ethan-1-ol ... 69
3.5.4. General procedure for the synthesis of (E)-2-((pyridinyl)vinyl)benzoxazoles ... 70
3.5.5. General procedure for the asymmetric addition to 19 ... 70
3.5.6. General procedure for the asymmetric addition to 22 ... 73
3.5.7. General procedure for the racemic addition to 22 ... 74
3.5.8. General procedure for the asymmetric addition to 25-27 ... 76
3.5.9. Procedure for the NMR studies about interaction between catalyst, 19 and TMSOTf ...77
3.5.10. Complexes CuBr·L7 and CuBr·L8 ... 78
3.6. Bibliography ... 79
Chapter 4: Autoinductive effects in an asymmetric copper(I)/phosphine catalyzed reaction ... 83
4.1. Introduction ... 84
4.1.1. Asymmetric autoinduction ... 84
4.1.2. Chiral tertiary alcohols ... 89
4.2. Aim ...91
4.3.1. Enantioselectivity as a function of conversion of the starting material in
1,2-additions of Grignard reagents to carbonyls ...91
4.3.2. Autocatalysis or autoinduction? ... 94
4.3.3. Asymmetric autoinductive effects: ketones vs. aldehydes ... 96
4.3.4. Interaction of an alkoxide with a copper/phosphine complex ... 98
4.4. Conclusions ... 102
4.5. Experimental section ... 103
4.5.1. General information ... 103
4.5.2. General procedure for the 1,2-addition of Grignard reagents to ketones (24a,b) ... 103
4.5.3. General procedure for the 1,2-addition of Grignard reagents to aldehydes (37a,b) ... 105
4.5.4. General procedure for monitoring the ee of the product of the reaction . 107 4.5.5. General procedure for the 1,2-addition of Grignard reagents to aldehydes with the use of additives ... 107
4.5.6. General procedure for the reaction carried out with different Grignard reagents ... 107
4.5.7. Procedure for the NMR experiments ... 108
4.6. Bibliography ... 109
Chapter 5: Design of an asymmetric organic autocatalytic reaction: the reduction of ketones and imines with borane ... 83
5.1. Introduction ... 114
5.1.1. Autocatalysis ... 114
5.1.2. Asymmetric autocatalysis ... 118
5.1.3. Corey-Bakshi-Shibata reduction and feasibility of asymmetric autocatalysis……….………122
5.2. Aim ... 124
5.3. Results and discussion ... 124
5.3.1. The Imine Pathway ... 125
5.3.2. The Ketone Pathway ... 130
5.3.3. Reduction of the phenyl-(2pyridyl)-ketone ... 130
5.4. Conclusions ... 133
5.5. Experimental section ... 134
5.5.1. General information ... 134
5.5.3. Synthesis of the imino-ketone 36 ... 135
5.5.4. Procedure for the synthesis of 30 ... 136
5.5.5. Procedure for the reduction of 30 ... 136
5.5.6. Racemic synthesis of the tert-butyl (S)-2-((R)-hydroxy(phenyl)methyl)pyrrolidine-1-carboxylate... 137
5.5.7. Asymmetric synthesis of the tert-butyl-2-((R)-hydroxy(phenyl)methyl)pyrrolidine-1-carboxylate... 140
5.5.8. General procedure for the deprotection of Boc-pyrrolidines ... 140
5.5.9. General procedure for the reduction of 47 ... 140
5.5.10. CBS-Reduction of 49 ... 141
5.6. Bibliography ... 141
Summary ... 146
Samenvatting………150
1
List of abbreviations
ACA: asymmetric conjugate addition API: active pharmaceutical ingredient
BINAP: 2,2'-bis(diphenylphosphino)-1,1'-binaphthyl BINOL: 1,1’-bi-2-naphtol
Boc: tert-butyl carbamate CA: conjugate addition
CBS reduction: Corey-Bakshi-Shibata reduction
CSP-HPLC: chiral solid phase high performance liquid chromatography DMAE: 2-dimethylaminoethanol
ee: enantiomeric excess
EWG: electron withdrawing group
GC/MS: gas chromatography/mass spectroscopy
HMBC: heteronuclear multiple bond correlation spectroscopy HMPA; hexamethylphosphoramide
HRMS: high resolution mass spectroscopy
HSQC: heteronuclear single quantum coherence spectroscopy LA: Lewis Acid
M: metal
MTBE: methyl-tert-butylether NMR: nuclear magnetic resonance o.n.: overnight
rt: room temperature TFA: trifluoroacetic acid THF: tetrahydrofurane
2
TM: transition metal
TMEDA: tetramethylethylendiamine TMSBr: trimethylsilyl bromide TMSCl: trimethylsilyl chloride
Chapter 1
Introduction
Herein, asymmetric catalysis is introduced. After a general explanation of the asymmetric syntheses, metal-catalysis and organocatalysis are explained and examples are given. This introduction touches the topics of kinetic resolution and dynamic kinetic resolution to obtain enantioenriched compounds. Finally, we describe asymmetric amplification as a peculiar case of asymmetric catalysis.
4
1.1. Catalysis in asymmetric syntheses
Symmetry is beauty. Symmetric faces result more attractive to people than asymmetric ones.[1] However, synthetic organic chemists are rather attracted to asymmetry in molecules. In fact, nature is an efficient asymmetric selector and homochiral molecules, for example, amino acids and sugars, compose living beings. Consequently, biological systems will interact differently with the two enantiomers and numerous pharmacological substances require to be composed by a single enantiomer to be administered to patients.[2]
Separations of enantiomers from racemic mixtures are commonly used, but their efficiency is limited by the fact than only half of the product can be utilized.[3–5] For this reason, copious methods have been developed to obtain enantiopure compounds. The synthesis with chiral auxiliaries affords enantioenriched products, nevertheless it requires stoichiometric amount of chiral reagents that are often expensive and require to be separated from the product.[6] Due to the convenience of using chiral auxiliaries in substoichiometric amount, enantioselective catalysis was widely developed during last century.[7,8]
1.2. Asymmetric metallic catalysis
Up to date, metallic catalysts are broadly used, both in laboratories and industrial processes.[9] In fact, these transformations usually require low catalyst loadings (less than 10 mol%) and are easily tunable thanks to the possibility to vary the metal and to modify the ligand, carrier of the chiral information.[10,11]
In 2001, the Nobel Prize for chemistry was awarded to Knowles and Noyori (hydrogenations) and Sharpless (oxidations) for their contribution in asymmetric catalysis. The group of Knowles developed a convenient synthesis of (L)-DOPA, an amino acid employed in Parkinson Disease’s treatment via the enantioselective hydrogenation of 1 with a Rh/chiral phosphine complex (Scheme 1a).[12,13] Noyori et
al. extended the scope of this reaction by using an atropoisomeric phosphine ligand
for the rhodium (BINAP, L2, Scheme 1b).[14,15] On the other hand, asymmetric oxidations were established by Sharpless and coworkes: among others, the epoxidation of allylic alcohol by organic peroxides could be performed with Ti(Oi-Pr)4 and diethyl tartrate (L3) as chiral additive. The use of the natural enantiomer (L)-L3 would have allowed the oxidation of the double bond from the enantiotopic face below (as drawn in Scheme 1c), independently from the substitution pattern.[16,17]
5
Scheme 1 Asymmetric reductions and oxidations.
1.3. Asymmetric organocatalysis
Asymmetric metal-catalysis has enabled the creation of stereocenters in many different ways. However, some metals are expensive, air sensitive and/or toxic. Inspired by enzymatic catalysis, organocatalysts are small organic molecules that can catalyze organic transformations via H-bonding and ionic interactions (Non-covalent catalysis) or formation of covalent bonds, like NHC carbenes[18] or amines[19] (covalent catalysis). These catalysts are generally inexpensive as they come from the chiral pool, but often require higher catalytic loadings.
The term organocatalysis was introduced by McMillan in 2000[20], however, the first example dates back to 1912, when Bredig and Fiske reported that chichona alkaloids catalyze the addition of HCN to aldehydes with poor ees.[21] In 1960, O-Acetyl quinine was used by Pracejus as an efficient catalyst for the addition of methanol to ketenes with 74% optical yield at -110°C (Scheme 2a). Interestingly, above -40°C, the reaction afforded the opposite enantiomer of the product.[22] Among the natural amino acids, proline has been widely used as organic catalyst, for reactions occurring via the
6
formation of imines or enamines[19] after the pioneering work of Wiechert et al. concerning proline catalyzed intramolecular aldol reactions (Scheme 2b)[23]. Occasionally, asymmetric organocatalysis is combined with metal-catalysis to enable reactions that inactive organocatalyst would not be able to catalyze alone.[24]
Scheme 2 Asymmetric organocatalytic transformations.
1.4. Catalytic (dynamic) kinetic resolution
Another powerful method to obtain enantiopure compounds using chiral catalysts is through resolution of racemic mixtures, widely used in industrial processes. We refer with the term kinetic resolution to an asymmetric reaction where the conversion of the two enantiomeric substrates to products occurs with different rates (Scheme 3).[25] In this way, after the reaction, one enantiomer is fully converted into the enantiopure product, while the other will be left as enantiopure substrate. Similar to other resolution methods, the upper yield limit for the kinetic resolution of a racemate is 50%. The enantioselectivity will instead depend on the ratio between the kinetic constants of the two asymmetric processes (s = krel = kfast/ kslow).[26]
7 An example is the acylation of benzylic alcohols. (R,R)-Methyl-DUPHOS (17) promotes the selective benzoylation of (R)-14 and (R)-15 is obtained in 81% ee at 25% conversion while the unreacted (S)-14 is recovered with 28% ee at 81% conversion (Scheme 4).[27]
Scheme 4 Kinetic resolution of benzylic alcohol via acylation.[27]
However, if the two enantiomers can interconvert, the product of the resolution can be collected as a single enantiomer with yields higher than 50%. This process is termed
dynamic kinetic resolution (Scheme 5).[28]
Scheme 5 General scheme for dynamic kinetic resolution.[28]
For instance, chiral substituted acetyl acetates (18) can racemize via keto-enol tautomery. In the asymmetric hydrogenantion with Nickel Raney and tartaric acid, the reduction of the (S)-enantiomer is favored and the equilibrium between (S)-18 and (R)-18 is consequently shifted in this direction (Scheme 6).[29]
8
Scheme 6 Dynamic kinetic resolution in the hydrogenation of acetyl acetates.[29]
1.5. Non-linear effects in asymmetric catalysis
In asymmetric catalysis, the ee of the product (eeprod) is linearly correlated to the one of the chiral auxiliary or catalyst (eeaux) by a constant that corresponds to the maximum
ee that can be achieved with an enantiopure catalyst/auxiliary (eemax) (Equation 1;
Curve A, Plot 1 Source: review of Kagan and Girard[30]). However, there are cases where the proportionality between the ee’s of the product and the auxiliary is lost, and we refer to that as non-linear effects (NLE). If the deviation from linearity is positive ((+)-NLE), the curve will resemble Curve B, Plot 1 while negative deviation ((-)-NLE) will generate a curve similar to Curve C, Plot 1.
Equation 1: eeprod= eemax*eeaux
Plot 1 Dependence of the eeprod on the eeaux. A) standard enantioselective reaction; B) (+)-NLE; C)
(-)-NLE. Source: Review of Kagan and Girard.[30]
Kagan and coworkers rationalized the non-linearity of enantioselectivity by different models that involve the formation of diastereomeric complexes or aggregates.[31] Concerning metallic catalysis, the models are abbreviated with MLn or (ML)n where n is the number of ligands or complexes that composes the species involved in the catalysis. The most recurring (and also the simplest) is the model ML2 or (ML)2. Other
9 models will not be treated in detail in this thesis, but an extensive explanation can be found in a fascinating review of Kagan and Girard.[30]
The model ML
2or (ML)
2This model concerns the formation of dimeric complexes and is illustrated in Scheme
7 (Source: H. B. Kagan and T.O. Luukas, Chapter 4, Comprehensive Asymmetric
Catalysis I.[7]). Metal complexes containing two ligands that can have R or S configuration (LR and LS) can be either homochiral (MLRLR and MLSLS) or heterochiral (MLRLS). Each of the homochiral complexes catalyzes the reaction enantioselectively with kinetic constant kRR = kSS, affording the product with a given eemax. MLRLS instead, catalyzes a racemic reaction with kRS. The final relative concentration of the catalyst are x, y and z and K is the equilibrium constant between homo- and heterochiral complexes, while β is the relative amount of heterochiral catalyst compared to the homochiral one.
Scheme 7 General scheme for the model ML2. Source: H. B. Kagan and T.O. Luukas, Chapter 4,
Comprehensive Asymmetric Catalysis I.[7]
The enantiomeric excess of the product can be expressed as function of the concentration of the catalytic species and of their kinetic constants in Equation 2.
Equation 2: eeprod= eemax*eeaux* 1+𝛽 1+𝑔𝛽
Non-linearity of asymmetric catalysis can be achieved when the following conditions are satisfied:
The rate of formation of the racemic product is slower than the one of the enantioenriched one (g ≤ 1)
The formation of the heterochiral complex is thermodynamically favored (K>>1). In fact, if the racemic complex is not formed (z=0), the value of β is null, and the correlation between ee of the product and the one of auxiliary is linear. In this case, the racemic portion of the catalyst is blocked via the formation of catalytically inactive species. On the other hand, the excess of one of the enantiomers of the chiral auxiliary catalyzes the enantioselective reaction. We can say that, in the model ML2, ligands behave like hands (that are the comparison par excellence for enantiomers).
10
The positive deviation from linearity in asymmetric catalysis is also referred to as
asymmetric amplification.
Scheme 8 Amplification of chirality in organometallic reactions.
Dimeric complexes (ML)2 can also cause non linearity of enantioselectivity according to the same model by forming homo- (MLRMLR and MLSMLS) and heterodimers (MLRMLS).[30] In the addition of dialkyl zinc to aldehydes with L4, reported by Noyori
et al., homochiral dimers easily dissociate to monomeric L4-ZnR, which can add
enantioselectively to aldehydes. Instead, the heterochiral dimers are stable inactive and, in this way, the chirality of the reaction is amplified (Scheme 8a).[32,33] Our group reported a non-linearity effect in the asymmetric copper-catalyzed 1,2-addition of Grignard reagents to enones.[34] Homochiral dimers of CuBr·L5 complex are soluble in MTBE[35] while heterochiral aggregates are not. For this reason, the racemic part of the complex does not participate in the catalysis and asymmetric amplification is therefore achieved (Scheme 8b).
1.6. Thesis outline
Chirality, or rather chiral induction through asymmetric catalysis, constitutes the main thread of this thesis. Different cases of catalysis are discussed and the approach changes through the four Chapters.
11 The first two Chapters address the enantioselective addition of Grignard reagents to symmetric diheteroaryl alkenes, reactive substrates towards the addition of organometallic reagents. Chapter 2 concerns the asymmetric conjugate addition of Grignard reagents to generic heteroaryl alkenes. The reactivity of this system and the consequent issues for enantioselectivity are discussed. Chapter 3 focuses on the addition of organomagnesium reagents to bispyridyl alkenes. The importance of pyridines is highlighted and the different behavior of substrates bearing 2-pyridyl or 4-pyridyl moieties is explained. A comparison between the ability of pyridine and benzoxazole to activate an alkene towards the conjugate addition is made.
The second part of the thesis concerns organic molecules that have a role in their own enantioselective synthesis. Chapter 4 regards autoinductive effects in the asymmetric 1,2-addition of Grignard reagents to enones. The product, an alkoxide, interacts with the copper/phosphine catalyst enabling a faster transmetallation that results in a better enantioselectivity. At the end, Chapter 5 describes the design or an organic asymmetric autocatalytic reaction inspired by Corey-Bakshi-Shibata reduction of ketones and imines with borane. The synthesis of the starting materials is reported and the prospects of asymmetric autocatalysis are discussed.
1.7. Bibliography
[1] D. W. Zaidel, S. M. Aarde, K. Baig, Brain Cogn. 2005, 57, 261–263. [2] L. A. Nguyen, H. He, C. Pham-Huy, Int. J. Biomed. Sci. 2006, 85–100.
[3] S. Ahuja, Chiral Separation Methods for Pharmaceutical and Biotechnological
Products., Wiley, 2013.
[4] F. Toda, Enantiomer Separation: Fundamentals and Practical Methods, Springer, 2004.
[5] M. Todd, Separation of Enantiomers: Synthetic Methods, Wiley, 2014. [6] Y. Gnas, F. Glorius, Synthesis (Stuttg). 2006, 1899–1930.
[7] E. N. Jacobsen, A. Pfaltz, H. Yamamoto, Comprehensive Asymmetric Catalysis
I-III, Springer, 1999.
[8] B. M. Trost, Proc. Natl. Acad. Sci. 2004, 101, 5349–5355.
[9] H. U. Blaser, H.-J. Federsel, Asymmetric Catalysis on Industrial Scale:
Challenges, Approaches and Solutions., Wiley, 2010.
[10] H. Pellissier, J. J. Spivey, Chiral Sulfur Ligands: Asymmetric Catalysis, Royal Society Of Chemistry, 2009.
[11] A. Pfaltz, Chimia (Aarau). 2004, 58, 49–50.
[12] W. S. Knowles, Angew. Chemie - Int. Ed. 2002, 41, 1998–2007. [13] W. S. Knowles, J. Chem. Ed. 1986, 63, 222–225.
[14] R. Noyori, Angew. Chemie - Int. Ed. 2002, 41, 2008.
[15] D. Glynn, J. Shannon, S. Woodward, Chem. - A Eur. J. 2010, 16, 1053–1060. [16] K. B. Sharpless, Angew. Chemie - Int. Ed. 2002, 41, 2024.
12
[17] S. Katsuki, K. B. Sharpless, J. Am. Chem. Soc 1980, 102, 5976–5978. [18] A. Grossmann, D. Enders, Angew. Chemie Int. Ed. 2012, 51, 314–325.
[19] S. Mukherjee, J. W. Yang, S. Hoffmann, B. List, Chem. Rev. 2007, 107, 5471– 5569.
[20] K. A. Ahrendt, C. J. Borths, D. W. C. MacMillan, J. Am. Chem. Soc 2000, 122, 4243–4244.
[21] G. Bredig, W. S. Fiske, Biochem. Z., 1912, 7–23.
[22] H. Pracejus, Justus Liebigs Ann. Chem. 1960, 634, 9–22.
[23] U. Eder, G. Sauer, R. Wiechert, Angew. Chemie - Int. Ed. 1971, 10, 496–497. [24] Z.-Y. Han, D.-F. Chen, Y.-Y. Wang, R. Guo, P.-S. Wang, C. Wang, L.-Z. Gong, J.
Am. Chem. Soc. 2012, 134, 6532–6535.
[25] H. Pellissier, Adv. Synth. Catal. 2011, 353, 1613–1666.
[26] J. M. Keith, J. F. Larrow, E. N. Jacobsen, Adv. Synth. Catal. 2001, 343, 5–26. [27] E. Vedejs, O. Daugulis, S. T. Diver, J. Org. Chem. 1996, 61, 430–431.
[28] R. Noyori, M. Tokunaga, M. Kitamura, Bull. Chem. Soc. Jpn. 1995, 68, 36–56. [29] A. Tai, H. Watanabe, T. Harada, Bull. Chem. Soc. Jpn. 1979, 52, 1468–1472. [30] C. Girard, H. B. Kagan, Angew. Chemie - Int. Ed. 1998, 37, 2922–2959. [31] D. Guillaneux, S.-H. Zhao, O. Samuel, D. Rainford, H. B. Kagan, J. Am. Chem.
Soc 1994, 116, 9430–9439.
[32] E. C. Anthony, M. Kitamura, S. Okada, S. Suga, R. Noyori, J. Am. Chem. Soc.
1989, 111, 4028–4036.
[33] M. Yamakawa ’, R. Noyori, J. Am. Chem. Soc 1995, 117, 6327–6335. [34] F. Caprioli, A. V. R. Madduri, A. J. Minnaard, S. R. Harutyunyan, Chem.
Commun. 2013, 49, 5450.
[35] F. Caprioli, M. Lutz, A. Meetsma, A. Minnaard, S. Harutyunyan, Synlett 2013,
Chapter 2
Control of enantioselectivity in the addition of Grignard
reagents to symmetric heteroaryl disubstituted olefins
Symmetric olefins bearing two electron withdrawing substituents are very reactive towards the addition of hard organometallic reagents. For this reason, the asymmetric addition of organolithium, organomagnesium and organozinc reagents, the background reaction competes with the catalytic pathway and high enantioselectivities can hardly be achieved. In this Chapter, we describe conjugate addition of Grignard reagents to bisheteroaryl olefins promoted by a copper/phosphine catalyst. The use of a Lewis acid allows selective acceleration of 1,4-addition pathway over side products formation, but is deleterious for the enantioselectivity. The formation of side products becomes more prominent as the length of the alkyl chain of Grignard reagents increases. The addition of methylmagnesium bromide proceeds with excellent enantioselectivity and good yield.
14
2.1. Introduction
2.1.1. Asymmetric addition of organometallic reagents to electron deficient
olefins
Electron deficient olefins, such as α,β-unsaturated ketones, (thio)esters, enamides, nitroalkenes, cyanoalkenes, alkenyl phosphates and sulfones, are important substrates for the formation of stereodefined C-C bonds. The presence of an electron withdrawing substituent activates the sp2-carbon in β-position towards the conjugate addition of nucleophiles (CA, Scheme 1) leading to the formation of a chiral sp3-carbon. Organic (semi)metallic reagents such as organoboron[1,2], -zinc[3–5], -zirconium[6] –aluminum[7– 11] and organomagnesium[12] reagents are commonly used for the asymmetric Michael addition in combination with a chiral metallic catalyst (copper(I) or rhodium(I) for arylboronic acids[2]). This reaction proceeds through transmetallation of the organometallic reagent on the chiral catalyst, binding to the electron deficient olefin substrate and subsequent stereoselective addition to the alkene. The order of reactivity of these reagents is RB(OH)2 << RZrX ≈ R2Zn < R3Al < RMgBr.
Scheme 9 General scheme for the asymmetric conjugate addition of organometallic reagents to
alkenes substituted with electron withdrawing substituents.
The addition of organoboron reagents to electron poor alkenes can occur only if mediated by a catalyst. Therefore, the chemo- and enantioselectivity of the reaction can be easily controlled through the choice of the metal salt and the ligand. Instead, the addition of a highly reactive Grignard reagents to electron deficient olefins, for instance an enone, can lead to four products (Scheme 2)[13]:
1. The 1,4-addition enolate product, that after aqueous workup affords the ketone
1.
2. The 1,2-addition product, alkoxide 2. The amount of this product increases with the hardness of the nucleophile.
3. The reduction product 3 caused by a β-hydride transfer from an organometallic reagent.
4. The enolization product 4, regenerating the starting material upon aqueous workup.
15
Scheme 10 Chemo- and regioselectivity in the addition of organometallic reagents to enones.
Moreover, the addition of a hard nucleophile to the substrate without mediation of the chiral catalyst is possible and occurs in a racemic fashion (background reaction). This decreases the overall enantioselectivity of the reaction. Dialkylzinc reagents display lower reactivity compared to their organomagnesium counterparts and leading to a slower background reaction. Grignard reagents, on the other hand, are cheaper, easily available and more atom economic when comparing to dialkylzinc reagents.
2.1.2. Enantioselectivity in the 1,4-addition of nucleophiles to symmetric
disubstituted alkenes
Enantioselective Micheal additions have been extensively studied and constitute a useful tool in asymmetric synthesis because of the broad choice of acceptors and donors. However, the stereoselective conjugate addition of nucleophiles to symmetric disubstituted electron poor alkenes is still underexplored. Symmetric 1,2-disubstituted alkenes with the E configuration have a C2h symmetry, and a C2v symmetry when Z (Scheme 3). The higher symmetry does not represent a challenge for the enantiocontrol of the reaction. Nevertheless, symmetric alkenes are hardly found among the plethora of reported methodologies regarding asymmetric conjugate addition (ACA).
Scheme 11 Symmetry of E and Z symmetric disubstituted alkenes.
However, in biological systems, the ACA these symmetric alkenes occurs with high enantioselectivity leading to the synthesis of chiral biomolecules. An example is the biosynthesis of (L)-aspartic acid from fumaric acid and ammonia catalyzed by the enzyme aspartase[14]. Modification of this enzymatic reaction allows the asymmetric addition of hydroxylamines and hydrazines with ee’s 97-99% (Scheme 4a).[15] Non-enzymatic ACA were also achieved, by using unreactive organometallic species like organoboron reagents: the addition of alkylmalonates (5) to different symmetric
16
alkenes (6) mediated by the chiral BINOLate 7 affords the product with ees up to 98% (Scheme 4b).[16]
Scheme 12 Asymmetric addition of nucleophiles to 1,2-diactivated alkenes.
The first contribution using metal catalysis concerns the rhodium(I) catalyzed addition of arylboronic acids to fumaric esters and maleimide reported by Hayashi and coworkers.[17] In this case, phosphine ligands, commonly used in rhodium catalysis[2], afforded the addition reaction with good to excellent yields (94-96%) but poor enantioselectivities (3-21% ee). Instead, the use of a norbornadiene ligand (L1) resulted in a decrease of the reactivity (78% yield with L1, 90% with L2) but an improvement in the enantioselectivity (90% ee, Scheme 5a). In 2012, Wu and coworkers reported, that among dienes as chiral ligands for the rhodium, L3, can be efficient in the addition of boronic acids to fumaric esters (Scheme 5b).[18] Bicyclic[2,2,2]octadienyl ligands were also employed in the construction of C-N chiral axes via the addition of arylboronic acids to N-substituted maleimide[19] and the enantioselective cyclopropanation of fumaric esters[20]. Likewise, Rh(I)/diphosphine catalyst (L4) provided excellent enantioselectivities in the addition of arylboronic acids to unprotected or N-alkyl maleimides at low temperatures (-0 - 0 oC, Scheme 5c)[21].
17
Scheme 13 Asymmetric rhodium(I)-catalyzed addition of arylboronic acids to disubstituted alkenes.
Excellent results were obtained in the ACA to deactivated alkenes when using catalytic systems based on Rh/chiral ligand and organoboron reagents. However, these methodologies are limited to arylations; a protocol that allows the alkylation and incudes the use of readily available organometallic reagents, such as organomagnesium reagents, is therefore desirable.
The group of Feringa reported the copper(I)-catalyzed addition of Grignard reagents to a variety of fumaric acid derivatives (Scheme 6).[22] The addition of MeMgBr to the alkene 9a catalyzed by CuBr/L5, proceeds with excellent regio- and enantioselectivity, thanks to significantly lower reactivity of this Grignard reagent, when comparing with EtMgBr.[23] Instead, the addition of EtMgBr to alkenes 9a and 9c affords a mixture of products with moderate enantioselectivities. When diethyl fumarate (9b) is used as
18
Michael-acceptor, the product is obtained with moderate to good ee’s (46% and 65% using respectively L5 and L6 as ligands for the copper), but the addition of a Grignard reagent to the olefin 9d is not stereoselective.[22]
Scheme 14 Asymmetric addition of Grignard reagents to various 1,2-disubstituted alkenes.[22] In the same thesis work, the addition of dialkylzinc reagents to substrates 9a-d is also described and racemic products were obtained in most cases (10a’ with 21% ee).[22] From these data, we can evince that the control of both regio- and enantioselectivity becomes an issue when both Michael-acceptors and donors are particularly reactive. This fact can be due to two factors:
Due to the higher reactivity, the catalytic process undergoes with lower enantioselectivity.
The presence of uncatalyzed addition of the Grignard reagent to the alkene as a racemic background reaction.
19 The interest of this chapter concerns the control of the enantioselectivity in the ACA of Grignard reagents to symmetric alkenes.
2.1.3. Copper(I)-catalyzed asymmetric addition of Grignard reagents to
(N)-containing heteroaryl alkenes
Heteroarenes are present in drug candidates, with an average of two heteroaryl rings per candidate drug.[24,25] A feature of around 50% of all Active Pharmaceutical Ingredients (APIs) is chirality: as the configuration of the drug is often determining its biological activity, a method to obtain it as a single enantiomer is fundamental.
In 2009, the pioneering work of Lam et al. disclosed the ability of N-containing heteroarenes to activate olefins towards the enantioselective copper-catalyzed conjugate addition of hydrides.[26] In the following year, the addition of organoboronic acids to these substrates, promoted by a chiral rhodium(I) catalyst and microwave irradiation, was also achieved by the same group. They obtained a cornucopia of chiral heteroarenes derivatives with excellent yields and enantioselectivities (12, Scheme
7).[27] Instead, the group of Lautens prepared aza-dihydrobenzoezepines with high ees, via a domino reaction involving the rhodium-catalyzed addition of borates and palladium-catalyzed C-O coupling.[28] These examples constitute two profitable methods for the asymmetric arylation of vinyl heteroarenes; nevertheless, procedures for the alkylation are still underexplored in literature.
Scheme 15 Asymmetric rhodium(I)-catalyzed addition of boronic acids to vinyl heteroarenes.
Our group hypothesized that by using highly reactive Grignard reagents, in the presence of a chiral copper catalyst, might allow the alkylation of alkenyl heteroarenes. Early efforts towards the asymmetric addition of phenylmagnesium bromide to vinyl pyridines, promoted by a nickel(I)-catalyst, were made, but they resulted in poor enantioselectivities (0-15% ee).[29] In 2016, our group developed the asymmetric copper(I)-catalyzed addition of Grignard reagents to different alkenyl heteroarenes.[30] This methodology affords a multitude of chiral heteroaryl derivatives with yields between 46 and 96% and excellent ee’s (86-99%, Scheme 8). The key to the success
20
of this reaction is the use of a Lewis acid (BF3.OEt2) that activates the substrate towards nucleophilic attack at β-position (Scheme 9).
Scheme 16 Enantioselective copper(I)-catalyzed addition of Grignard reagents to alkenyl
heteroarenes[30].
It was proposed that the catalytic cycle starts with the π-Cu(I)-complexation (14) between the activated alkenyl heteroarene (11a, Scheme 9) and the transmetallated complex (13). Next, the oxidative addition results in a σ-Cu(III)-adduct (15) that releases the product after reductive elimination (12, Scheme 9).[30,31]
The reactivity of this system can be modified by tuning not only the catalyst, but also the activator for the substrate. This characteristic represents an opportunity to achieve high enantioselectivity on the conjugate addition Grignard reagents to highly reactive olefins.
21
Scheme 17 Hypothetical catalytic cycle[30]
2.2. Aim
A scarce number of examples of enantioselective alkylations of symmetric disubstituted alkenes has been reported. Due to our findings concerning the addition of Grignard reagents to alkenyl heteroarenes, we have decided to explore the asymmetric addition of organomagnesium reagents to symmetric disubstituted alkenes. This Chapter describes how we tackled the presence of the prominent background reaction and how we tuned the reactivity of the catalytic system to obtain enantioenriched 1,2-bis(heteroaryl)substituted alkenes (Scheme 10), valuable chiral heteroaryl compounds.
Scheme 18 Copper catalyzed addition of organomagnesium reagents to bisheteroaryl olefins.
2.3. Results and discussion
2.3.1.
Synthesis of 1,2-disubstituted heteroaryl alkenes
First, we prepared the substrates which were not commercially available. We chose to prepare alkenes having benzoxazyl, quinolyl and pyrimidyl moieties (16, 17 and 18
22
respectively, Figure 1), to gain a view over the reactivity of this class of olefins. The alkenes 16 and 17 were synthesized using modifications of the literature procedures.[32,33] The synthesis of alkene 18 proved to be challenging because of the low availability of starting materials containing a 2-pyrimidine moiety, and the olefin was not succesfully synthesized. However, the attempts for its synthesis will be discussed.
Figure 1 Symmetric heteroaryl disubstituted alkenes 16-18
Condensation reactions are not commonly used to synthesize 2-benzoxazyl alkenes due to the high cost of 2-benzoxazyl carbaldehyde.[34] The described synthesis of 16 consists of the formation of the benzoxazole from 2-hydroxilaninile and fumaric acid using harsh conditions (polyphosphoric acid, PPA, as a solvent).[35] However, because of the technical difficulties in handling PPA (viscosity and use of high temperatures), we opted for a microwave assisted synthesis of 16 (Scheme 11)[32]. When fumaric acid (19) was used as a starting material, 21 was found as a byproduct (Scheme 11a). To favor the ring closure of the benzoxazyl ring over the conjugate addition, we decided to use malic acid (22) as precursor.[36] No significant change in the yield was observed (Scheme 11b). Instead, an increase from 18 to 27% was observed when the concentration of the reaction was reduced to half, probably due to the better solubility and stirring.
23
Scheme 19 Synthesis of 1,2-bis(2-benzoxazyl)-ethene (17).
Concerning 17, this substrate was prepared by condensation of quinaldine (23) and 2-quinoline carbaldehyde (24). The procedures for the condensation of 23 with an aldehyde require high temperatures and long reaction times (more than 24h).[37,38] Unfortunately, we observed degradation of 17 after prolonged reaction times. Gladly, we could obtain the desired product via condensation reaction in acetic anhydride at 140 °C (Scheme 12).[33] The preheating of the bath is crucial to prevent the degradation of the product and the reaction time of 20 min proved to be optimal.
Scheme 20 Synthesis of 1,2-bis(2-quinoyl)-ethene (17).
For the synthesis of compound 18, procedures involving the use of 2-pyrimidine carbaldehyde were avoided due to the cost of this starting material. At first, we attempted the synthesis of 18 via olefin metathesis (Scheme 13a).[39] The 2-vinyl pyrimidine 27 was prepared by Suzuki coupling with a yield of 29%: probably part of the product was lost due to its volatility.[40] The metathesis approach did not result in the formation of the product, neither at reflux of CH2Cl2 using the second generation Grubbs catalyst (28), nor at 100 oC in toluene with the catalyst M2 (29).
In the next attempt, we decided to submit the already synthesized 27 to Heck reaction (Scheme 13b).[41] Also in this case, no formation of the desired product was detected.
24
It is feasible that in both metathesis and Heck reactions, the poisoning of the catalyst by the pyrimidine occurs faster than the reaction.
The last strategy was to perform a Suzuki coupling on the vinyl borate 32. Despite the numerous efforts to convert 2-alkynylpyrimidine (30) in 32 via zirconium-catalyzed hydroboration, 32 was not detected among the products (Scheme 13c). A possible explanation for the unsuccessful reaction is the hindrance and electron with drawing effect of the pyrimidine as substituent for the alkyne. Consequently, we decided to abandon the synthesis of compound 18.
Scheme 21 Attempts towards the synthesis of 1,2-bis(2-pyridimyl)-ethene (18).
Having successfully synthesized olefins 16 and 17, we moved towards testing the asymmetric catalytic additions to these substrates.
25
2.3.2. ACA of Grignard reagents to benzoxazyl alkenes
We started from 16, as benzoxazole substrate as it was found to be a good model for addition reactions in the previous work of our group.[30] The addition of ethylmagnesium bromide on 16 was tested using the optimized methodology,[30] that involves the use of BF3.OEt2 as Lewis acid and 5 mol% of a catalyst formed by complexation of CuBr·SMe2 and L8 (Table 1). Unfortunately, substrate 16 was poorly soluble in commonly used solvents like diethyl ether, MTBE, toluene at room temperature and -78 oC. Instead, 16 is soluble in CH2Cl2 at room temperature, but not at -78oC. When ethereal solvents were used as a solvent, racemic product (33a,
entries 1 and 3, Table 1). No conversion to the product was observed in toluene, as
well as in diethyl ether in the absence of BF3.OEt2 (entries 2 and 4). On the contrary, in CH2Cl2 the reaction reached full conversion and 16 was recovered with 48% yield and 52% ee in the absence of a Lewis acid (entry 5). This observation denotes an enhanced reactivity of this double substituted alkene with respect to the monosubstituted ones.[30] With BF3.OEt2 it was possible to enhance the yield of the product at the expense of the ee (entries 6-8) that dropped to 27% when 1.5 equiv. of BF3.OEt2 was used. The strength of the Lewis acid affects these two parameters in a similar way. In fact, TMSBr (1.5 equiv.) was found to be an efficient Lewis acid affording the product with 86% yield and 54% ee (entry 9) but the yield was only 36% when 0.5 equiv. of TMSBr was used. The highest enantioselectivities were achieved when using TMSCl, but this LA did not allow the reaction to give full conversion (entry
26
Table 1 Screening of Lewis acids in the addition.
Entry LA (equiv.) Solvent Conversion (%)a Yield (%) b eec
1 BF3.OEt2 (1.5) Et2O 78 n.d. 6
2 None Et2O 0 - -
3 BF3.OEt2 (1.5) MTBE ~50 n.d. racemic
4 BF3.OEt2 (1.5) Toluene 0 - - 5 None CH2Cl2 >95 48 52 6 BF3.OEt2 (1.5) CH2Cl2 >95 58 27 7 BF3.OEt2 (1.0) CH2Cl2 >95 56 44 8 BF3.OEt2 (0.5) CH2Cl2 >95 42 53 9 TMSBr (1.5) CH2Cl2 >95 86 54 10 TMSBr (0.5) CH2Cl2 >95 36 53 11 TMSCl (1.0) CH2Cl2 >95 53 58
a Conversion to 33a determined by 1H-NMR b Isolated yield. c Determined via CSP-HPLC.
Analysis of the data presented in Table 1, shows that only moderate enantioselectivities can be achieved in this transformation. This can be explained by a fast competing background reaction, in both the presence and the absence of Lewis acid (Scheme 14). It was observed that Lewis acid accelerates the addition of EtMgBr to the olefin, therefore improve the regio and chemoselectivity of the reaction, in both the catalyzed and the background reactions.
Scheme 22 Non-catalyzed addition of EtMgBr to 18.
Aiming to improve the enantioselectivity by tuning the structure of the chiral catalyst, different classes of chiral diphosphines L5-L16 were screened (Table 2) using either
27 BF3.OEt2, TMSCl, BCl3 or no Lewis acid. Ligands L5, L9 and L15 afforded the addition product with low ees (2-18%). Using ligands L11, L12, L14 and L16, the addition proceeded with moderate to good yields and moderate enantioselectivities.
28
a Conversion to 33a was determined by 1H-NMR b Isolated yield. c Determined via CSP-HPLC. d 10 mol% of the catalyst used e 1.0 equiv. of EtMgBr was used f THF used as solvent g EtMgBr diluted in CH2Cl2 to 0.6 mL and added over 2h h EtMgBr diluted in CH2Cl2 and added over 4h
Our interest was caught by L10: with 0.5 equiv. of BF3.OEt2, we could obtain 33a in 42% yield and 69% of ee (entry 5, Table 2). Curiously, the use of 10 mol% of the catalyst and 1.0 equiv. of the organometallic reagent, caused a decrease of the enantioselectivity (entries 6 and 7, respective ees 64% and 45%). The change of reaction solvent from CH2Cl2 to THF resulted in a complete loss of enantiocontrol (entry 8). BCl3 has a positive effect on the yield and negative on ee (yield 61%, ee 53%). With L16 we obtained 33a in 81% yield and 59% ee (opposite enantiomer, entry 12). To our delight, in the absence of BF3.OEt2, the ee was enhanced to 68% which could be further improved to 76% ee by slow addition of the Grignard reagent over 2h (entries
13 and 14). Again, the yield of the product dropped when the catalyst loading was
doubled (entry 15). Unfortunately, the results of entry 14 could not be repeated. To guarantee the reproducibility of the reaction, we decided to add the EtMgBr in 4h, as
Entry Ligand LA (equiv.) Conversion (%)a Yield (%) b eec 1 L5 TMSCl (1.0) >95 24 6 2 L5 BF3.OEt2 (0.5) ~90 40 4 3 L9 TMSCl (1.0) >95 37 18 4 L9 BF3.OEt2 (0.5) >95 58 2 5 L10 BF3.OEt2 (0.5) >95 42 69 6d L10 BF3.OEt2 (0.5) 81 47 64 7e L10 BF3.OEt2 (0.5) >95 48 45 8f L10 BF3.OEt2 (0.5) 40 n.d. racemic 9 L10 BCl3 (0.5) >95 61 53 10 L11 BF3.OEt2 (0.5) >95 n.d. 42 11 L12 BF3.OEt2 (0.5) >95 58 26 12 L13 BF3.OEt2 (0.5) >95 81 -59 13 L13 None >95 57 -68 14g L13 None >95 65 -76 15d,g L13 None >95 34 -68 16h L13 None >95 79 -71 17 L14 None >95 n.d. -26 18g L14 None >95 n.d. -38 19 L15 BF3.OEt2 (0.5) >95 n.d. 7 20g L16 None >95 39 40
29 reported in entry 16 (yield and ee, 79 and 71% respectively) and take the conditions of this entry as the optimized reaction conditions.
To identify the reason for reproducibility issues with respect to the enantioselectivity, we considered the possibility of the enolization of the product, caused by the free Grignard in solution in the moment of the quench (Scheme 15a). We tried to reproduce those conditions by submitting enantioenriched 33a to 2.0 equiv. of MeMgBr alone or with 1.5 equiv. of BF3.OEt2: no significant change of ee (76% before, 74% and 73% after, Scheme 15b) was observed, thus enolization can be excluded as possible racemization pathway.
Scheme 23 Possible racemization of the product after the quench upon deprotonation by
organomagnesium reagents.
The scope of the Grignard reagents was evaluated in the optimized conditions (Table
3). The addition of n-HexMgBr affords 33c with similar enantioselectivity to 33a but
with poor yield of 7% (entry 2, Table 3). In fact, the major product derived from 1,2-addition to the benzoxazole ring. The use of 0.5 equiv. of BF3.OEt2 was beneficial for the yield for all of the organometallic reagents used (entries 4-7). With 1.0 equiv. the yield further improved, albeit at the cost of enantioselectivity (entry 7). The trend observed for linear Grignard reagents shows longer chains lead to lower yields (33a, yield = 81%, 33b, yield = 65%, 33b, yield = 29%). The enantioselectivity is always moderate under these conditions. The use of CypMgBr afforded 33d as a racemate in good yield (entry 8). In the previous screening, we found that 1.5 equiv. of TMSBr as an additive improved yield and ee, for the reaction (entry 9, Table 1), however in this reaction it resulted in poor conversion of 14% (entry 11, Table 3). As expected, MeMgBr was less reactive than the other organomagnesium reagents (entries 9 and
30
10) and in the presence of BF3.OEt2 (1.5 equiv.) the reaction proceeded with 89% conversion, 58% yield and moderate ee of 39% (entry 10).[23]
Table 3 Grignard reagents scope.
Entry RMgBr a LA (equiv.) Conversion (%)b Yield (%)c eed
1 EtMgBr None >95 79 71 2 n-HexMgBr None 80 7 70 3 i-BuMgBr None 0 - - 4e EtMgBr BF3.OEt2 (0.5) >95 81 59 5 n-BuMgBr BF3.OEt2 (0.5) >95 65 40 6 n-HexMgBr BF3.OEt2 (0.5) >95 29 45 7 n-HexMgBr BF3.OEt2 (1.5) >95 56 23 8 CypMgBr BF3.OEt2 (0.5) >95 71 6 9 MeMgBr BF3.OEt2 (0.5) 51 n.d. n.d. 10 MeMgBr BF3.OEt2 (1.5) 89 58 39 11 n-HexMgBr TMSBr (1.5) 14 n.d. n.d.
a RMgBr diluted in CH2Cl2 to 0.6 mL and added over 4h b Conversion to 33 was determined by 1H-NMR. c Isolated yield. d Determined via CSP-HPLC. e Fast addition of the concentrated RMgBr.
Hoping that higher ees can be achieved with MeMgBr, thanks to its reactivity, we optimized the reaction conditions further. Changing to ligand L10 did not improve the enantioselectivity (entry 1, Table 4), and the reaction did not proceed in the absence of BF3.OEt2 (entry 2). To our delight, the use of L16 allowed the formation of 33e in 59% yield and 97% ee (entry 4) even if a non-catalyzed reaction can occur in the presence of BF3.OEt2.
31
Table 4 Asymmetric addition of MeMgBr.
Entry Ligand LA (equiv.) Conversion (%)a Yield (%)b eec 1 L10 BF3.OEt2 (1.5) >95 79 -35 2 L10 None 0 - - 3 d (R)-L16 BF3.OEt2 (1.5) >95 35 -95 4d (S)-L16 BF3.OEt2 (1.5) >95 59 97 5 No catalyst BF3.OEt2 (1.5) >95 - -
a Conversion to 33e was determined by 1H-NMR b Isolated yield. c Determined via CSP-HPLC. d MeMgBr diluted in CH2Cl2 to 0.6 mL and added over 2h.
In the catalytic addition of Grignard reagents to 16, good enantioselectivities were achieved only at expense of yield. In fact, the use of a Lewis acid favors the non-catalyzed addition of the nucleophile over the enantioselective pathway, but is necessary to prevent the formation of side products. At this point, we were eager to study the behavior of 17 in this reaction.
2.3.3. ACA of Grignard reagents to symmetric 2-quinoyl alkenes
Next, the reactivity of the alkene 17 was investigated in our reaction. The highest conversion (58%) was observed when L8 was used as ligand for the copper in the presence of 1.5 equiv. of BF3.OEt2 (entry 1, Table 5). However, no product 34 could be isolated by column chromatography. An increase of the temperature to -50oC or an increase in the amount of Lewis acid entails lower conversion of the starting material to the product (entries 2 and 3). Likewise, the addition does not occur in the presence of TMSBr and has a poor conversion with TMSOTf (entries 4 and 5). When L13 is used the NMR-conversion reaches 45%. In order to obtain a racemic product, we tested racemic BINAP as ligand but we obtained low or no conversion using BF3.OEt2 or AlCl3 (entries 7 and 8) respectively. Additionally, without copper catalyst, 34 could not be formed. We presume that this is due to the steric bulk of the TMS group.
32
Table 5 Reactivity of 17 towards the copper(I)-catalyzed addition of EtMgBr.
Entry Ligand LA (equiv.) Conversion (%)a
1 L8 BF3.OEt2 (1.5) 58 2b L8 BF3.OEt2 (1.5) 0 3 L8 BF3.OEt2 (3.0) 26 4 L8 TMSBr (1.5) 0 5 L8 TMSOTf (1.5) 19 6 L13 BF3.OEt2 (1.5) ~45 7 rac-BINAPc BF3.OEt2 (1.5) 30 8 rac-BINAPc AlCl3 0 9 No catalyst TMSOTf (1.5) 0
a Conversion to 34 was determined by 1H-NMR b Reaction performed at -50oC c10 mol% of catalyst were used.
We suppose that the low reactivity of 17 in this reaction is caused by the excessive steric hindrance close to the alkene. In fact, once a Lewis acid is coordinated to the nitrogen atom, it is in proximity of both sides of the alkene. The reactivity of 17 is comparable to the one of 35 (Figure 2), which is poorly reactive towards the addition of Grignard reagents[42]. This suggests that 2-(hetero)aryl-substituted vinyl quinolines are probably too hindered for this type of reaction.
Figure 2 Structure of 2-styrylquinoline.
2.4. Conclusions
In this Chapter, the syntheses and applications of two symmetric disubstituted olefins, substrates in the reactions of ACA of Grignard reagents are described. These compounds can be obtained via ring closure and subsequent formation of the heteroarene or via condensation. In both cases, the yields are moderate due to degradation of the product in the reaction conditions.
The behavior in the asymmetric addition differs for the two substrates. The alkene bearing two 2-benzoxazyl moieties is very reactive towards the addition of
33 organomagnesium reagents. This reactivity has implications on the enantioselectivity, because the non-catalyzed addition of the organometallic reagents competes with the catalytic pathway. The use of Lewis acid improves the yield of the reaction, as it promotes the conjugate addition rather than the formation of byproducts, but it is deleterious for the ee. The length of the chain of the Grignard reagent influences greatly the reaction outcome: a longer chain is leading to lower yields. Again, Lewis acids can impede side reactions, at expense of enantioselectivity. However, the addition of methylmagnesium bromide proceeds with excellent enantioselectivity and in good yield, because of the mild reactivity of this organometallic reagent, allowing the catalytic reaction to outcompete non catalyzed reaction.
Finally, we found that the symmetric bisquinoyl olefin hardly undergoes conjugate addition with organomagnesium reagents, probably due to the steric hindrance of the Lewis acid close to the alkene.
2.5. Experimental section
2.5.1. General information
All reactions using oxygen- and/or moisture-sensitive materials were carried out with anhydrous solvents (vide infra) under a nitrogen atmosphere using oven dried glassware and standard Schlenk techniques. Reactions were monitored by 1H NMR. Purification of the products, when necessary, was performed by flash-column chromatography using Merck 60 Å 230-400 mesh silica gel, Merck 90 active neutral or VWR AnalaR NORMAPUR aluminum oxide basic. NMR data was collected on Bruker Avance NEO 600 (1H at 600.0 MHz; 13C at 150.87MHz), equipped with a Prodigy Cryo-probe and Varian VXR400 (1H at 400.0 MHz; 13C at 100.58 MHz), equipped with a 5 mm z-gradient broadband probe. Chemical shifts are reported in parts per million (ppm) relative to residual solvent peak (CDCl3, 1H: 7.26 ppm; 13C: 77.16 ppm). Coupling constants are reported in Hertz. Multiplicity is reported with the usual abbreviations (s: singlet, bs: broad singlet, d: doublet, dd: doublet of doublets, ddd: doublet of doublet of doublets, t: triplet, td: triplet of doublets, q: quartet, m: multiplet). Exact mass spectra were recorded on a LTQ Orbitrap XL apparatus with ESI ionization. Enantiomeric excesses (ees) were determined by chiral HPLC analysis using a Shimadzu LC-10ADVP HPLC equipped with a Shimadzu SPD-M10AVP diode array detector and by Waters Acquity UPC2 system with PDA detector and QDA mass detector.
Unless otherwise indicated, reagents and substrates were purchased from commercial sources and used as received. Solvents not required to be dry were purchased as technical grade and used as received. Dry solvents were freshly collected from a dry solvent purification system prior to use. Inert atmosphere experiments were
34
performed with standard Schlenk techniques with dried (P2O5) nitrogen gas. Grignard reagents were purchased from Sigma Aldrich and used as received (EtMgBr (3.0M in Et2O), n-HexMgBr, i-ButMgBr (2.0 M in Et2O), CypMgBr (1.8M in Et2O). All other Grignard reagents were prepared from the corresponding alkyl bromides and Mg activated with I2 in Et2O and concentration was determine by NMR titration method[43]. Chiral ligands (L5, L8, L9 and L12 - L16) were purchased from Sigma Aldrich and Solvias. Chiral ligands L10 and L11 were prepared according to literature method.[44] All reported compounds were characterized by 1H and 13C NMR and compared with literature data. All new compounds were fully characterized by 1H and 13C NMR and HRMS techniques.
2.5.2. Synthesis of substrates
(E)-1,2-bis(2-benzoxazyl)-ethene (16)
In a 35 mL microwave vial, 0.14 g of malic acid (1.35 mmol, 1 equiv.) and 0.38 g of 2-hydroxyaniline (3.33 mmol, 2.5 equiv.) and 10 mL of toluene were added. The suspension was stirred at 70 oC for 1 h. 10 mg of H3BO3 (0.16 mmol, 0.12 equiv.) and 10 mg of p-toluensulfonic acid (0.06 mmol, 0.04 equiv.) were added. The vial was heated in the microwave at 170 oC (300 W, high stirring) for 1 h. The solvent was removed under reduced pressure and the crude was purified with a short flash-column chromatography (Al2O3, CH2Cl2). Compound 16 was obtained as bright orange crystal (0.128 g, 27% yield).
1H-NMR (400 MHz, CDCl3), δ 7.80 (d, J = 9.4 Hz, 2H), 7.70 (s, 2H), 7.60 (d, J = 7.7
Hz, 2H), 7.47 – 7.33 (m, 4H).
13C NMR (101 MHz, Chloroform-d) δ 160.9, 150.8, 142.3, 126.6, 125.1, 123.8, 120.8,
110.9.
HRMS (ESI+): m/z calcd. for C16H10N2O2 ([M+H+]) 263.08150, found 263.08178.
(E)-1,2-bis(2-quinolyl)-ethene (17)
In a Schlenk under dry and inert atmosphere, equipped with reflux condenser 0.79 g of 2-quinoline carboxaldehyde (5 mmol, 1 equiv.) and 2-methylquinoline (5 mmol, 1 equiv.) were dissolved. The Schlenk was placed in a pre-heated bath at 140 oC and stirred for 20 min. The reaction mixture was allowed to cool to rt and then poured into ice. A saturated solution of NaHCO3 was added until pH 9 (gas formed at this stage). The reaction mixture was extracted with toluene (3x20 mL) and dried with MgSO4. The solvent was removed under reduced pressure. The crude was purified by flash-column chromatography (Al2O3, pentane:AcOEt, 4:1) and crystallization from EtOH. Compound 17 was obtained as bright yellow crystals (0.114 g, 20% yield).
35 1H NMR (400 MHz, Chloroform-d) δ 8.20 (d, J = 8.6 Hz, 2H), 8.12 (d, J = 8.5 Hz, 2H), 7.96 (s, 2H), 7.83 (m, 4H), 7.74 (td, J = 8.5, 1.4 Hz, 2H), 7.54 (td, J = 8.2, 1.2 Hz, 2H). 13C NMR (101 MHz, Chloroform-d) δ 156.0, 148.8, 135.2, 130.4, 130.0, 128.2, 128.1, 127.2, 120.1.
HRMS (ESI+): m/z calcd. for C20H14N2 ([M+H+]) 283.12298, found 283.12319.
2-vinylpyrimidine (27)
A solution of 2-chloropyrimidine (1.27 g, 8.00 mmol), potassium vinyltrifluoroborate (1.29 g, 9.60 mmol), PdCl2(dppf)·CH2Cl2(131 mg, 0.16 mmol), and Et3N (1.12 mL, 8.00 mmol) in i-PrOH (125 mL) was heated to
reflux for 16 h. The mixture was cooled to rt and partitioned between CH2Cl2 (100 mL) and H2O (40 mL). The aqueous layer was separated and extracted
with CH2Cl2 (2 x 50 mL). The combined organic layers were washed with a saturated
solution of NaCl (100 mL), dried over MgSO4, filtered, and the solvent was removed under reduced pressure. The pure compound 27 was obtained after flash chromatography (SiO2, pentane:Et2O, 9:1) as a pale yellow oil (0.247 g, 2.3 mmol, 29% yield)
1H NMR (400 MHz, Chloroform-d) δ 8.70 (d, J = 4.9 Hz, 2H), 7.13 (t, J = 4.9 Hz, 1H),
6.88 (dd, J = 17.3, 10.6 Hz, 1H), 6.62 (dd, J = 17.4, 1.6 Hz, 1H), 5.73 (dd, J = 10.6, 1.6 Hz, 1H).
13C NMR (101 MHz, Chloroform-d) δ 157.0, 144.0, 136.5, 123.8, 119.1.
HRMS (ESI+): m/z calcd. for C13H13N2 ([M+]) 107,06037, found 107.06005.
2.5.3. Catalytic asymmetric addition to 16
General procedure
In a heat dried Schlenk tube equipped with septum and magnetic stirring bar, the CuBr ·SMe2 (5 mol%), and (S,Sp)-L13 (6 mol%) were dissolved in CH2Cl2 (1mL/0.1mmol of substrate) and stirred under nitrogen atmosphere for 15 min. The substrate (0.1 mmol, 1 equiv.) was added at once. After stirring for 5 min at rt the reaction mixture was cooled to -78 °C and BF3·OEt2 (0-1.5 equiv.) was added. RMgBr (2.0 equiv) was diluted in CH2Cl2 (0.6 ml total volume) and added over 4 hours. After stirring at -78 °C for 16h, the reaction was quenched with MeOH (0.5 mL) followed by saturated aqueous NH4Cl solution and warmed to rt. The reaction mixture was extracted with CH2Cl2 (3 × 10 mL). Combined organic phases were dried over MgSO4, filtered and solvents were evaporated under reduced pressure. The oily crude was purified by column chromatography on neutral Al2O2 using a mixture of pentane and EtOAc (9:1) as eluent. The configuration of the products was not assigned.
36
2,2'-(butane-1,2-diyl)bis(benzoxazole) (33a)
The reaction was performed with 16 (26 mg, 0.1 mmol, 1.0 equiv.), EtMgBr (0.2 mmol, 3.0 M in Et2O) diluited in CH2Cl2 (0.6 mL total volume), CuBr·SMe2 (2.0 mg, 0.010 mmol, 5 mol%), (S,Sp)-L13 (4.2 mg, 0.012 mmol, 6 mol%), in 1 mL CH2Cl2. Product 33a was obtained as yellow oil (46.2 mg, 1.6 mmol, yield 79%, ee 71%). The absolute configuration of 33a was not assigned. 1H NMR (400 MHz, Chloroform-d) δ 7.80 – 7.59 (m, 2H), 7.53 – 7.39 (m, 2H), 7.34 – 7.23 (m, 4H), 3.75 (qd, J = 7.5, 5.7 Hz, 1H), 3.60 (dd, J = 15.6, 7.5 Hz, 1H), 3.40 (dd, J = 15.6, 7.2 Hz, 1H), 2.08 – 1.84 (m, 2H), 0.99 (t, J = 7.4 Hz, 3H). 13C NMR (101 MHz, Chloroform-d) δ 167.8, 164.5, 150.8, 150.7, 144.0, 141.2, 141.1, 124.7, 124.2 (2C), 119.8, 119.7, 110.5, 110.4, 39.2, 31.6, 26.3, 11.2.
HRMS (ESI+): m/z calcd. for C18H16N2O2 ([M+H+]) 293.1285, found 293.1288.
CSP-HPLC: (254nm, Chiralcel OB-H, n-heptane/i-PrOH = 95:5, 40 °C, 0.5 ml/min.),
tR = 13.83 min (major), tR = 12.38 min (minor).
2,2'-(hexane-1,2-diyl)bis(benzoxazole) (33b)
The reaction was performed with 16 (26 mg, 0.1 mmol, 1.0 equiv.), n-BuMgBr (0.2 mmol, 1.8 M in Et2O) diluted in CH2Cl2 (0.6 mL total volume), CuBr·SMe2 (1.0 mg, 0.005 mmol, 5 mol%), (S,Sp)-L13 (4.1 mg, 0.006 mmol, 6 mol%), BF3·OEt2 (0.006 mL, 0.05 mmol, 0.5 equiv.) in 1 mL CH2Cl2. Product 33b was obtained as colorless oil (20.7 mg, 0.065 mmol, 65% yield, 40% ee). The absolute configuration of 33b was not assigned.
1H NMR (400 MHz, Chloroform-d) δ 7.72 – 7.61 (m, 2H), 7.54 – 7.40 (m, 2H), 7.34 –
7.21 (m, 4H), 3.84 – 3.72 (m, 1H), 3.59 (dd, J = 15.6, 7.7 Hz, 1H), 3.39 (dd, J = 15.6, 7.0 Hz, 1H), 2.06 – 1.82 (m, 2H), 1.39 – 1.22 (m, 4H), 0.91 – 0.76 (m, 3H).
13C NMR (151 MHz, Chloroform-d) δ 168.2, 164.7, 151.0, 150.9, 141.4, 141.3, 124.82,
124.80, 124.3, 120.0, 119.9, 110.7, 110.6, 38.0, 33.3, 32.3, 29.1, 22.6, 14.0.
HRMS (ESI+): m/z calcd. for C20H20N2O2 ([M+H+]) 321,1598, found 321.1602.
CSP-HPLC: (254nm, Chiralcel OZ-H, n-heptane/i-PrOH = 95:5, 40 °C, 0.5
