ECHET96 Search CD [Molecules: 5] [Related articles/posters: 082 032 114 109 067 ]
email.gif - 0.3 KECHET96 Article 081 Iain Coldham
email.gif - 0.3 KRe: ECHET96 Article 081 Bob Gawley

Alkylations of N-allyl-2-lithiopyrrolidines. Several analogies to reactions of N-methyl compounds and one surprise

Robert E. Gawley* and Silvio Campagna

Department of Chemistry, University of Miami, Coral Gables, Florida 33124-0431, USA

Introduction and background

The alkylation of a-amino organolithium compounds has been a topic of interest to synthetic chemists for well over two decades [1-4]. Some recent developments relevant to the present work are Beak's discovery that N-BOC-pyrrolidine can be enantioselectively deprotonated and alkylated (eqn. 1), [5,6] and our finding that 2-lithio-N-methyl-pyrrolidines and -piperidines possess exceptional configurational stability and can be alkylated with a variety of electrophiles (eqn. 2) [7-9].

Secondary organolithiums having a heteroatoms such as nitrogen or oxygen are tetrahedral, and therefore stereogenic. One of the fascinating aspects of the chemistry of compounds such as these is the stereochemical course of their reactions with electrophiles. The transition states for SE2 substitutions giving retention and inversion are not very far apart in energy [10]. Whether the electrophilic substitution (of lithium by an electophile) occurs with retention or inversion of configuration at the metalated carbon (or with racemization), sometimes depends on the electrophile. Early work from the Still group indicated that a-alkoxy organolithiums react with retention of configuration [11]. However, subsequent studies by Hoppe showed that the lithiated carbamate illustrated in eqn. (3) affords products of inversion with acid chlorides, tin chlorides, and the protonating agents triphenylmethane and acetic acid; reaction with methanol, alkyl halides, esters and anhydrides, on the other hand, gave products of retention!

The lithiated N-BOC-pyrrolidines in eqn. (1), which are configurationally stable at low temperature, give products of retention of configuration with all electrophiles reported to date [5,6]. In contrast, 2-lithio-N-methyl-pyrrolidines and -piperidines afford products with inversion, retention or racemization, depending on the electrophile (Scheme 1) [7,9]. For those electrophiles that do not afford racemic products, the carbonyl-containing electrophiles react with essentially 100% retention for both piperidines and pyrrolidines. With alkyl halides, piperidines are highly selective, yielding products of inversion with nearly 100% selectivity. Pyrrolidines are somewhat less selective, affording only about 78% inversion of configuration.

One drawback to synthetic applications of some a-hetero-organolithiums is the fact that they do not always react efficiently with electrophiles. For example, we have been unsuccessful in achieving high yields of alkylation products with N-BOC-2-lithio-pyrrolidines and activated (e.g. allylic) alkyl halides, and unactivated primary alkyl halides are even worse. In contrast, and for reasons we do not understand, 2-lithio-N-methyl-pyrrolidines and -piperidines are excellent nucleophiles with a broad spectrum of electrophiles, affording good yields of products with all electrophiles tested, regardless of the stereochemistry of the reaction (eqn. 2, Scheme 1).

In spite of this broad spectrum of reactivity, the processes illustrated in Scheme 1 have been (so far) limited to N-methyl compounds. Because we are ultimately interested in applying these reactions to synthetic problems, we decided to explore the possibility of extending these processes to heterocycles where there is a removable substituent on nitrogen. To that end, we tested N-allyl pyrrolidines, and report our preliminary findings herein. In order to evaluate both chemical reactivity and stereochemical features, we used educt of high ee (enantiomeric excess). Parenthetically, these are the same compounds that we used to define the stereochemical features of [2,3]-anionic and ylide rearrangements of a-aminoorganolithiums [12].

Off with the BOC, on with the allyl

The procedure for preparing (S)-N-allyl-2-(tributylstannyl)pyrrolidine is shown in Scheme 2. Beak's (S)-N-BOC-2-tributylstannylpyrrolidine (92-94% ee, 92:4 - 97:3 er) is treated with B-bromocatechol borane to remove the BOC group, and immediately alkylated with allyl bromide to afford (S)-N-allyl-2-(tributylstannyl)pyrrolidine, in 50-55% overall yield after flash chromatography on neutralized silica. This compound is thermally labile, but can be stored in a freezer under an inert atmosphere for several days.

Transmetallations...

In a fashion analogous to the procedures for the N-methyl compounds, the N-allylpyrrolidinyl stannane can be transmetalated using butyllithium in THF/TMEDA at low temperature in a few minutes. Although this organolithium is prone to Wittig rearrangement near room temperature [12], it is stable at low temperature. Quenching the organolithium with a carbon dioxide stream and reducing the crude N-allylproline with LAH afforded N-allylprolinol in 57% yield (after flash chromatography) as a colourless oil. An authentic sample of (S)-N-allylprolinol was obtained by alkylation of (S)-prolinol. Comparison of the two specific rotations indicates that the alkylation product is (R), and of 87% ee (93.5:6.5 er). Thus, the reaction of the lithiated N-allylpyrrolidine occurred with at least 96% retention of configuration. The reaction probably goes with complete retention, since the differences in specific rotations (of the beginning stannane as well as the two products) are probably within experimental error, and because preliminary data reported elsewhere indicate that N-allyl-2-lithiopyrrolidine is even more resistant to racemization than its N-methyl analogue [12], which may be due to additional stabilization by the double bond. These results are summarized in Scheme 3.

Range of electrophiles

Similarly, a number of other electrophiles were tested, and the results are shown in Scheme 4. In all cases, the products of alkylation were formed in excellent yields. The absolute configuration of the products were confirmed by independent synthesis (E = carbon dioxide, benzyl bromide [13]), or by order of elution on a Pirkle column after de-allylation (see below) and acylation with naphthoyl chloride [14] (E = propyl bromide, heptyl bromide, 3-phenylpropyl bromide). Other configurations are assigned by analogy (E = PhCHO, naphthoyl chloride). Comparison of the absolute configuration and enantiomer purity of the various products with that of the stannane reveal that the trends observed in the N-methyl series (Scheme 1) are largely followed: electrophiles having a carbonyl add with virtually complete retention of configuration, while primary alkyl halides react with ca. 90% inversion of configuration. The 90% stereoselectivity for reactions with alkyl halides is significantly higher than that observed with 2-lithio-N-methylpyrrolidines, which were only ca. 78% selective [9] (also showing inversion).

Never assume you know what will happen....

The surprise is benzyl bromide. In the N-methyl series (Scheme 1), the product of benzylation was racemic, but here a 76:24 ratio of enantiomers was found. Even more surprising is that this alkyl halide reacted with ca. 78% retention of configuration, making it the only alkyl halide (so far found) that reacts with these lithioheterocycles with retention!

Allyl group removal

To complete the reaction sequence, we used Wilkinson's catalyst to remove the allyl group[15] for selected examples, as shown in Scheme 5. There are some limitations to this protocol: one is that the allyl group removal fails for N-allylpyrrolidines having a hydroxyl group in the 2-substituent (i.e. benzaldehyde addition product and prolinol) and for 1,2-diallylpyrrolidine. The other is that removal of the allyl group for the compound with the hex-5'-enyl substituent is accompanied by some double bond migration in the side chain. This latter experiment has only been done once, however, and efforts to optimize the conditions are in progress.

Summary and conclusions

Replacement of the BOC group in N-BOC-2-tributylstannyl pyrrolidines with an allyl group, followed by tin-lithium exchange, affords an a-aminoorganolithium species that reacts with either retention or inversion of configuration, depending on the electrophile. This two-step sequence may find use in applications where direct alkylation of the BOC-pyrrolidine is inefficient, or when the opposite enantiomer is desired.

References

  1. P. Beak; D. B. Reitz Chem. Rev. 1978, 78, 275.
  2. P. Beak; W. J. Zajdel; D. B. Reitz Chem. Rev. 1984, 84, 471.
  3. R. E. Gawley; K. Rein In Comprehensive Organic Synthesis. Selectivity, Strategy, and Efficiency in Modern Organic Chemistry; ed. B. M. Trost and I. Fleming, Pergamon, Oxford, 1991; Vol. 3, pp. 65-83.
  4. T. K. Highsmith; A. I. Meyers In Advances in Heterocyclic Natural Product Synthesis; ed. W. H. Pearson, JAI, Greenwich, CT, 1991; Vol. 1, pp. 95-135.
  5. S. T. Kerrick; P. Beak J. Am. Chem. Soc. 1991, 113, 9708.
  6. P. Beak; S. T. Kerrick; S. Wu; J. Chu J. Am. Chem. Soc. 1994, 116, 3231.
  7. R. E. Gawley; Q. Zhang J. Am. Chem. Soc. 1993, 115, 7515.
  8. R. E. Gawley; Q. Zhang Tetrahedron 1994, 50, 6077.
  9. R. E. Gawley; Q. Zhang J. Org. Chem. 1995, 60, 5763.
  10. E. D. Jemmis; J. Chandrasekhar; P. v. R. Schleyer J. Am. Chem. Soc. 1979, 101, 527.
  11. W. C. Still; C. Sreekumar J. Am. Chem. Soc. 1980, 102, 1201.
  12. R. E. Gawley; Q. Zhang; S. Campagna J. Am. Chem. Soc. 1995, 117, 11817.
  13. J. E. Nordlander; M. J. Payne; F. G. Njoroge; M. A. Balk; G. D. Laikos; V. M. Vishwanath J. Org. Chem. 1984, 49, 4107.
  14. W. H. Pirkle; C. J. Welch; G. S. Mahler; A. I. Meyers; L. M. Fuentes; M. Boes J. Org. Chem. 1984, 49, 2504.
  15. C. Bennett; B. Leguzza; B. Ganem Tetrahedron Lett. 1981, 22, 1483.