The Synthesis of Sterically Hindered Amides

252 CHIMIA 2014, 68, Nr. 4 Laureates of the sCs faLL Meeting 2013

doi:10.2533/chimia.2014.252 Chimia 68 (2014) 252–255 © Schweizerische Chemische Gesellschaft The Synthesis of Sterically Hindered Amides

Gabriel Schäfer§ and Jeffrey W. Bode*

§SCS-Metrohm Foundation Award for best oral presentation

Abstract: Amide bond formation is one of the most important reactions due to the ubiquity of the amide functional group in pharmaceuticals and biologically active compounds. However, even the best existing methods reach their limits when it comes to the synthesis of sterically hindered amides. In this article we summarize our research in the formation of sterically hindered amides. We show that the direct coupling of Grignard reagents to isocyanates provides a facile and robust solution to this long-standing challenge and hope that this methodology will find widespread application in the synthesis of pharmaceuticals and materials. Keywords: Amide synthesis · N-Carboxyanhydrides · Grignard reagents · Isocyanates · Steric hindrance

1. Introduction commercially available, allowing the cus- scribed the chemoselective, amide-bond tomer to choose the most suitable reagent forming ligation of hydroxylamines with The synthesis of amide bonds is an for a specific chemical transformation. α-ketoacids (KAHA ligation).[4] This re- integral part of every organic chemistry However, one of the biggest challenges in action proceeds under mild conditions, laboratory and is commonly regarded as amide bond formation chemistry cannot be requires no reagents or catalysts and only the most used chemical reaction in drug solved even with the best coupling reagent produces water and CO as by-products. 2 discovery.[1] The conventional method to in hand: the synthesis of sterically hin- Although the KAHA ligation can be prepare amides is condensation of amines dered amides.[3] The difficulties associated used to form simple amide products, its and carboxylic acids in the presence of a with the formation of this class of amides major application lies in the synthesis of coupling reagent (Scheme 1).[2] stems from the slow nucleophilic attack of larger peptides using a C-terminal peptide The role of the coupling reagent is to the amine onto the activated carboxylate in α-ketoacid and an N-terminal hydroxyl- transform the carboxylic acid moiety into a sterically congested environment. amine (Scheme 2a).[8] Despite the elegance an activated carboxylate, which is attacked During the last 10 years our group of this approach, sterically hindered am- by the amine to generate a new amide bond. has reported several new amide bond ides are difficult to form via KAHA liga- Many different coupling reagents are now formation methods.[4–7] In 2006, we de- tion. In 2012, our group disclosed a new

Scheme 1. Traditional coupling reagent O amide bond forma- O R2 base 2 tion. 1 R H2N solvent R N R1 OH H

via activated carboxylate O

R1 X*

Scheme 2. a) KAHA ligation Chemoselective amide-forming liga- O O tions by Bode and HO R2 –H O, –CO OH 2 2 2 co-workers.[4–7] 1 N R R H DMF (0.2–0.01 M) R1 N O 40 ºC H 1 b) Acyltrifluoroborate O-benzoyl hydroxylamine ligation *Correspondence: Prof. Dr. J. W. Bode ETH Zürich O Laboratorium für Organische Chemie O BzO R2 N R2 Vladimir-Prelog-Weg 1–5 R1 CH-8093 Zürich R1 BF K H H2O/t-BuOH N 3 rt, 30 min H Tel.: +41 44 633 2103 2 E-mail: [email protected] Laureates of the sCs faLL Meeting 2013 CHIMIA 2014, 68, Nr. 4 253

amide-forming ligation using acyltrifluo- Scheme 3. roborates and O-benzoyl hydroxylamines, Identification and quantification of without the need for additional reagents or MgBr catalysts (Scheme 2b).[7] This reaction op- O phenylmagnesium bromide via addition erates under aqueous conditions and gives O Et2O C N to 1-naphthylisocya- N H remarkably rapid amide formation. Many nate, as described by different amides could be prepared in high Gilman et al.[9b] yields, however highly sterically hindered amides could not be formed. In order to ad- dress the challenge of preparing sterically O Ph Ph Ph Ph hindered amides, we sought out to find a RMgBr O C R N Me facile and reliable solution of this long- N Me Et2O(0.25 M) H 0˚Ctort, 30 min standing unmet synthetic need. 3

selected examples: F O Ph Ph 2. Addition of Grignard Reagents to Me O Ph Ph iPr O Ph Ph O Ph Ph F Isocyanates N Me N Me N Me H N Me H H H F F In our effort to find a practical method Me Me iPr iPr for the preparation of sterically hindered F amides we reexamined two reports from 87% yield 79% yield 73% yield 75% yield Gilman on the addition of Grignard re- Scheme 4. Addition of sterically hindered Grignard reagents to isocyanate 3. agents to isocyanates to form amide bonds (Scheme 3).[9] Two factors were intrigu- ing to us: i) Gilman described a C–C bond Me Me O MgBr O forming approach for amide synthesis, C R R N Et2O(0.25 M) N which is in contrast to the standard cou- 0˚Ctort, 30 min H pling-reagent based C–N bond formation Me Me Me Me 4 process; and ii) the synthetic community, including ourselves, has nearly forgotten selectedexamples: this elegant approach of Gilman as only a Cl F Me O Me Me few applications of this reaction have ap- Me O Me O Me O N Me peared in the literature in the last 85 years H N N N H H H Me Me Cl CF3 − none of them targeted sterically hindered Me Me Me Me Me Me [10] amides. We reasoned that this was due 83% yield 87% yield 83% yield 87% yield to the fact that Gilman and coworkers Boc themselves were not interested in the am- Me N Me O Me O Me Me Me ide products as such, but rather in finding Me O Me O N N CO2Me a way to accurately identify and quantify H H N N CF3 CO Me their Grignard reagents. H Me Me 2 Me Me H Me Me O We were eager to see if the addition of Me Me Grignard reagents to isocyanates would 83% yield 70% yield 85% yield 79% yield allow the synthesis of sterically hindered Scheme 5. Addition of mesitylmagnesium bromide (4) to various isocyanates. amides. We began our investigation by us- ing the sterically hindered isocyanate 3 as Scheme 6. Synthesis a model substrate. Several bulky Grignard iPr iPr iPr iPr O of exceptionally reagents could be added with ease to iso- MgBr O C N hindered amides 5 cyanate 3 and provided the corresponding N H iPr and 6. products in high yields (Scheme 4).[11] iPr iPr iPr iPr iPr 5,78% yield The reaction proved to be simple and Et2O(0.25 M) 0˚Ctort, 30 min user-friendly: A solution of the Grignard reagent was added to a solution of isocya- MgBr O O N nate in ethereal solvent at 0 ºC, followed C H by warming to room temperature. The N analytically pure amide was obtained after 6,75% yield aqueous workup and washing of the crude material with hexanes. To explore the versatility of our meth- ides could be accessed using our condi- partner, setting a benchmark in the area of odology, we examined several other iso- tions. The chemoselective addition of the steric hindrance in amide bond synthesis cyanates in the reaction with mesityl- Grignard reagent to the isocyanate moiety (Scheme 6). magnesium bromide (4, Scheme 5). A over ester or ketone functional groups is wide range of other sterically hindered noteworthy – the corresponding amides isocyanates could be successfully used, in- were obtained in good yields. 3. Synthesis of Isocyanates cluding tert-butyl, adamantyl, 2,6-dichlo- The highlight of our work was the prep- rophenyl or 2-fluoro-6-trifluoromethyl aration of two extremely hindered amides A criticism of our work is the use of isocyanate. Furthermore, sterically hin- 5 and 6 via the combination of the most isocyanates, a class of compounds that is dered α-trifluoromethyl-substituted am- hindered substrates from each reaction biased with a poor reputation since the 254 CHIMIA 2014, 68, Nr. 4 Laureates of the sCs faLL Meeting 2013

Bhopal accident with methyl isocyanate. Scheme 7. Synthesis It is important to note that isocyanates a) From carboxylic acids of isocyanates. with a higher molecular mass, especially O sterically hindered examples, are easily Ph Ph -N N+ P NEt Ph Ph handled materials and can be stored over HO N O 3 O Me O C months without any sign of decomposi- toluene N Me O 115ºC, 45 min tion. Additionally, isocyanates are highly appreciated starting materials in polymer DPPA 3,85% yield chemistry and many of them are commer- cially available[12] or can be readily pre- b) From amines pared via well-established methods, start- Boc Boc ing from either carboxylic acids or amines. N Cl O Cl N The sterically hindered isocyanates pre- Cl Cl Cl O O Cl sat. aq. NaHCO3/ O CH Cl (1:1) C pared via these methods can be purified HCl H N CF 2 2 N CF 2 3 0ºC, 30 min 3 by column chromatography or distillation triphosgene (Scheme 7).[10] 65% yield

4. N-Carboxyanhydrides as α-Isocyanatocarboxylic Acid Scheme 8. Surrogates Traditional:NCAs in ring-opening polymerizations N-Carboxyanhydrides (NCAs) in polymer initiator O H chemistry and in the Despite the broad substrate scope of (base or nucleophile) N R H N synthesis of sterically the isocyanate-Grignard methodology, - n CO2 R1 R2 n H hindered, N-acylated one substrate class could not be directly O O amino acids. accessed: sterically hindered, N-acylated HN amino acids. The reason for this was the O R1 R2 lack of a suitable starting material, namely N-carboxyanhydride O α-isocyanatocarboxylic acids. We sought (NCA) 3 O H R MgBr R3 N N to find a solution for this limitation and C O OH O 1 2 were intrigued by the extensive literature R1 R2 O R R on N-carboxyanhydrides (NCA) – a sub- α-isocyanatocarboxylate sterically hindered intermediate N-acylated amino acids strate class commonly used as starting ma- Our work:Addition of Grignard reagents to NCAs terials in the preparation of polypeptides via ring-opening polymerization.[13] We speculated that NCAs could serve as α-isocyanatocarboxylic acid surrogates Scheme 9. Addition in combination with two equivalents of a O Me Me of mesitylmagnesium Me Me O O H bromide (4) to NCA 7. Grignard reagent. The first equivalent of HN N O OH Grignard reagent serves as a base and de- MgBr THF (0.2 M) Me O protonates the N–H proton, upon which an Me -78 ˚C,15min then rt, 1h α-isocyanatocarboxylate is released. This 4 (2.1 equiv) 7 93% yield reactive intermediate is subsequently at- tacked by the second equivalent of organo- metallic reagent to form the desired amide bond (Scheme 8). O When 3-oxa-1-azaspiro[4.5]decane- O O H 2,4-dione (7) was treated with mesitylmag- R3 R3 N MgBr HN O THF (0.2 M) OH nesium bromide (4) at –78 ºC and subse- O R1 R2 R1 R2 -78 ˚C,15min quently warmed to room temperature, the then rt, 1h (2.1 equiv) corresponding gem-disubstituted, N-acyl amino acid was isolated in nearly quanti- selected examples: tative yield (Scheme 9).[14] F t-Bu t-Bu CN F F O O Me O The substrate scope of the Grignard re- O H H Me H H N N N agent and the NCA was found to be very N OH OH Me OH F OH broad and several highly sterically hin- t-Bu O O O F O dered, N-acylated amino acids could be obtained in good to high yield, including 73% yield 81% yield 68% yield 90% yield products with sensitive functional groups Me Me Me Me O Me Me (Scheme 10). Organolithium reagents were O Me H O H O N H N H OH N also viable coupling partners in this reac- OH N OH Me OH Me O tion and several interesting heterocyclic Me O Me O H O Ph Ph N products could be obtained (Scheme 11). MeS Ph The existence of an α-isocyanato- Boc 84% yield 75% yield 80% yield 75% yield carboxylate intermediate was successfully ee over 99:1 established using in situ IR spectroscopy: When two equivalents of mesitylmagne- Scheme 10. Addition of Grignard reagents to N-carboxyanhydrides. Laureates of the sCs faLL Meeting 2013 CHIMIA 2014, 68, Nr. 4 255

Scheme 11. Addition [1] S. D. Roughley, A. M. Jordan, J. Med. Chem. O of organolithium re- O O 2011, 54, 3451. H agents to NCA 7. [2] E. Valeur, M. Bradley, Chem. Soc. Rev. 2009, HN O R N RLi THF (0.2 M) OH 38, 606. -78 ˚C,15min O [3] V. R. Pattabiraman, J. W. Bode, Nature 2011, then rt, 1h 480, 471. (2.2 equiv) 7 [4] J. W. Bode, R. M. Fox, K. D. Baucom, Angew. Chem. Int. Ed. 2006, 45, 1248. selected examples: [5] J. W. Bode, S. S. Sohn, J. Am. Chem. Soc. 2007, Me 129, 13798. O N O O [6] V. R. Pattabiraman, A. O. Ogunkoya, J. W. H H H N N Bode, Angew. Chem. Int. Ed. 2012, 51, 5114. O OH N OH N N OH [7] A. M. Dumas, G. A. Molander, J. W. Bode, Me O O O Angew. Chem. Int. Ed. 2012, 51, 5683. [8] a) T. Fukuzumi, L. Ju, J. W. Bode, Org. Biomol. 83% yield 73% yield 57% yield Chem. 2012, 10, 5837; b) Y.-L. Huang, R. Frey, M. E. Juarez-Garcia, J. W. Bode, Heterocycles 2012, 84, 1179; c) J. Wu, J. Ruiz-Rodríguez, J. M. Comstock, J. Z. Dong, J. W. Bode, Chem. sium bromide (4) were added to a solution via traditional coupling reagent-based ap- Sci. 2011, 2, 1976. of 3-oxa-1-azaspiro[4.5]decane-2,4-dione proaches. The Grignard-isocyanate pro- [9] a) H. Gilman, C. R. Kinney, J. Am. Chem. Soc. –1 1924, 46, 493; b) H. Gilman, M. Furry, J. Am. (7) an intense absorption at 2250 cm tocol does not need an excess amount of Chem. Soc. 1928, 50, 1214. appeared immediately, consistent with the reagents, the reactions are complete with- [10] a) H. M. Singleton, W. R. Edwards, Jr., J. Am. typical valence vibration of isocyanates. in minutes at room temperature and the Soc. Chem. 1938, 60, 540; b) L. Field, J. E. products are easily purified. Furthermore, Lawson, J. W. McFarland, J. Am. Chem. Soc. the addition of organometallic species to 1956, 78, 4389; c) J. W. McFarland, R. L. Harris, J. Org. Chem. 1967, 32, 1273; d) K. 5. Conclusions NCAs opens up a new way for the syn- A. Parker, E. G. Gibbons, Tetrahedron Lett. thesis of sterically hindered, N-acylated 1975, 16, 981; e) Y. Zhang, J. Jiang, Y. Chen, The synthesis of sterically hindered amino acids. The reaction proceeds via Tetrahedron Lett. 1987, 28, 3815; f) I. Stefanuti, amides is considered as an unsolved chal- an α-isocyanatocarboxylate intermediate, S. A. Smith, R. J. K. Taylor, Tetrahedron Lett. 2000, 41, 3735; g) A. Padwa, K. R. Crawford, P. lenge in the area of amide bond forma- whose existence was demonstrated by re- Rashatasakhon, M. Rose, J. Org. Chem. 2003, tion. The addition of Grignard reagents to act-IR.[13] We hope that the direct coupling 68, 2609; h) M. I. Antczak, J. M. Ready, Chem. isocyanates provides a facile and robust of isocyanates with Grignard reagents will Sci. 2012, 3, 1450. solution to this problem and a wide range find a widespread use in the preparation of [11] G. Schäfer, C. Matthey, J. W. Bode, Angew. of sterically hindered amides can be syn- challenging amides, both in academic and Chem. Int. Ed. 2012, 51, 9173. [12] E. Delebecq, J.-P. Pascault, B. Boutevin, F. [10] thesized with ease and in high yields. industrial laboratories. Ganachaud. Chem. Rev. 2013, 113, 80. Most of the described amide products in [13] H. R. Kricheldorf, Angew. Chem. Int. Ed. 2006, this report would be difficult to synthesize Received: January 30, 2014 45, 5752. [14] G. Schäfer, J. W. Bode, Org. Lett. 2014, 16, 1526.