Aza-Wittig Rearrangement of N,N-Dipropargylic α-Amino Alkyllithiums: Periselectivity and Steric Course

Abstract

The aza-Wittig rearrangement of enantio-defined N,N-dipropargylic α-amino alkyllithiums, generated by tin-lithium ­exchange, are shown to afford [1,2]- and/or [2,3]-rearrangement products. Both aza-Wittig rearrangements proceed predominantly with inversion of configuration at the Li-bearing carbon terminus.

Key words

asymmetric - amines - carbanions - Wittig rearrangements - periselectivity

References

• Reviews:
• 1a Vogel C. Synthesis  1997,  497 
• 1b Tomooka K. In The Chemistry of Organolithium Compounds   Rappoport Z. Marek I. Wiley; New York: 2004.  Chap. 12.
• For recent examples of synthetic applications of the aza-Wittig rearrangement, see:
• 2a Broka CA. Shen T. J. Am. Chem. Soc.  1989,  111:  2981 
• 2b Coldham I. J. Chem. Soc., Perkin Trans. 1  1993,  1275 
• 2c Åhman J. Somfai P. Tetrahedron Lett.  1996,  37:  2495 
• 2d Anderson JC. Whiting M. J. Org. Chem.  2003,  68:  6160 
• 3 Tomoyasu T. Tomooka K. Nakai T. Tetrahedron Lett.  2003,  44:  1239 
• 5 Gawley’s group has reported the first experimental evidence for the inversion course in the aza-[2,3]-Wittig rearrangement of (S)-N-allyl-2-lithiopyrrolidine, which considerably competes with the [1,2]-Wittig rearrangement process. However, no systematic study has been made on the steric course at the Li-bearing terminus of the aza-[1,2]-Wittig rearrangement. See: Gawley RE. Zhang Q. Campagna S. J. Am. Chem. Soc.  1995,  117:  11817 
• 11 The periselectivity of the Wittig rearrangement of propargylic ethers also depends upon the kind of substituents at the acetylenic terminus, see: Tomooka K. Komine N. Nakai T. Synlett  1997,  1045 
• 16 Retention of stereochemistry in tin-lithium exchange reactions is well-known for the generation of α-hetero alkyllithiums, see: Pearson WH. Lindbeck AC. Kampf JW. J. Am. Chem. Soc.  1993,  115:  2622 ; and references cited therein

4

We have also reported that the rearrangement of N,N-dicrotyl α-amino alkyllithium afforded a mixture of aza-[1,2]- and [2,3]-Wittig rearrangement product, see ref. [3]

6

For the preparation of N,N-dipropargyl alkylstannane (R)-1a, N,N-diisopropylethylamine was utilized as a base instead of NaH.

7

All the compounds were characterized by ¹H NMR and ¹³C NMR. Data for selected products are as follows. (R)-1a: ¹H NMR (CDCl₃): δ = 0.86-0.94 (m, 15 H), 1.24-1.53 (m, 12 H), 1.80-1.93 (m, 1 H), 2.07-2.20 (m, 1 H), 2.23 (t, J = 2.3 Hz, 2 H), 2.58-2.82 (m, 2 H), 3.17 (dd, J = 9.2, 5.4 Hz, 1 H), 3.26 (dt, J = 6.5, 2.3 Hz, 2 H), 3.55 (dt, J = 6.5, 2.3 Hz, 2 H), 7.17-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 11.0, 13.6, 27.6, 29.3, 34.2, 34.7, 42.6, 55.9, 73.1, 80.3, 125.6, 128.3, 128.5, 142.4. [α]_D -45.6 (c 0.25, CHCl₃). (R)-1b: ¹H NMR (CDCl₃): δ = 0.86-0.92 (m, 15 H), 1.23-1.50 (m, 12 H), 1.82 (t, J = 2.3 Hz, 6 H), 1.82-1.92 (m, 1 H), 2.02-2.20 (m, 1 H), 2.60-2.87 (m, 2 H), 3.08 (dd, J = 16.2, 2.3 Hz, 2 H), 3.23 (dd, J = 4.5, 10.4 Hz, 1 H), 3.50 (dd, J = 16.2, 2.3 Hz, 2 H), 7.16-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 3.8, 10.9, 13.7, 27.6, 29.4, 31.6, 34.6, 43.2, 55.9, 75.6, 80.6, 125.6, 128.2, 128.6, 142.7. [α]_D -63.6 (c 0.25, CHCl₃). (R)-1c: ¹H NMR (CDCl₃): δ = 0.86-0.93 (m, 15 H), 1.13 (t, J = 7.4 Hz, 6 H), 1.24-1.53 (m, 12 H), 1.80-1.93 (m, 1 H), 2.07-2.28 (m, 1 H), 2.19 (tq, J = 7.4, 2.1 Hz, 4 H), 2.58-2.70 (m, 1 H), 2.75-2.86 (m, 1 H), 3.07 (dt, J = 16.2, 2.1 Hz, 2 H), 3.31 (dd, J = 9.4, 5.4 Hz, 1 H), 3.50 (dt, J = 16.2, 2.1 Hz, 2 H), 7.16-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 10.8, 12.6, 13.6, 13.9, 27.6, 29.4, 34.6, 34.8, 43.3, 55.9, 75.8, 86.5, 125.5, 128.2, 128.5, 142.7. [α]_D -65.2 (c 0.25, CHCl₃).

8

General Procedure for aza-Wittig Rearrangement: To a THF (8 mL) solution of (R)-1c (200 mg, 0.36 mmol) was added n-BuLi (1.13 mL, 1.60 M in hexane, 1.80 mmol) dropwise at -78 °C. After the addition, the solution was stirred for 15 min at -78 °C, and the temperature was allowed to rise to 0 °C over a period of 4 h. The reaction was quenched with sat. aq NH₄Cl and the product was extracted with Et₂O. The combined organic phase was dried over Na₂SO₄, filtered and the solvent was removed in vacuo. The residue was purified by silica gel column chromatography (hexane-Et₂O = 5:1) to give the [1,2]-product 3c (10 mg, 10%), [2,3]-product 4c (35 mg, 36%) and allene 6c (14 mg, 15%).

9

Data for selected products are as follows. Compound 3a: ¹H NMR (CDCl₃) δ = 1.80-1.89 (m, 2 H), 2.02 (t, J = 2.6 Hz, 1 H), 2.21 (t, J = 2.6 Hz, 1 H), 2.41 (ddq, J = 16.8, 5.3, 2.6 Hz, 2 H), 2.69 (t, J = 7.9 Hz, 2 H), 2.91-2.98 (m, 1 H), 3.47 (t, J = 2.6 Hz, 2 H), 7.19-7.31 (m, 5 H). ¹³C NMR (CDCl₃):
δ = 23.2, 32.0, 35.3, 35.7, 54.0, 70.5, 71.4, 80.9, 81.9, 125.7, 128.2, 128.2, 141.7. HRMS (EI): m/z calcd for C₁₅H₁₇N: 211.1361. Found: 211.1351. Compound 3c: ¹H NMR (CDCl₃): δ = 1.12 (t, J = 7.4 Hz, 3 H), 1.13 (t, J = 7.4 Hz, 3 H), 1.77-1.86 (m, 2 H), 2.18 (q, J = 7.4 Hz, 2 H), 2.19 (q, J = 7.4 Hz, 2 H), 2.20-2.46 (m, 2 H), 2.68 (t, J = 7.9 Hz, 2 H), 2.80-2.90 (m, 1 H), 3.42 (t, J = 2.4 Hz, 2 H), 7.15-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 12.4, 12.5, 14.0, 14.3, 23.6, 32.0, 35.4, 36.2, 54.7, 70.8, 75.9, 80.0, 84.0, 125.8, 128.3, 128.4, 142.3. HRMS [(S)-methylbenzyl urea derivative 7, EI]: m/z calcd for C₂₈H₃₄N₂O: 414.2671. Found: 414.2657. Compound 4b: ¹H NMR (CDCl₃): δ = 1.64 (t, J = 3.1 Hz, 3 H), 1.75-1.87 (m, 2 H), 1.80 (s, 3 H), 2.64 (t, J = 8.1 Hz, 2 H), 3.23 (t, J = 7.1 Hz, 1 H), 3.30 (m, 2 H), 4.68 (m, 2 H), 7.14-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 3.5, 13.8, 32.3, 35.3, 36.1, 59.7, 74.4, 74.6, 77.3, 98.3, 125.7, 128.3, 128.4, 142.2, 206.8. HRMS [(S)-methylbenzyl urea derivative 9, EI]: m/z calcd for C₂₆H₃₀N₂O: 386.2358. Found: 386.2350. Compound 4c: ¹H NMR (CDCl₃): δ = 1.04 (t, J = 7.4 Hz, 3 H), 1.11 (t, J = 7.4 Hz, 3 H), 1.77-1.98 (m, 4 H), 2.13-2.22 (m, 2 H), 2.65 (t, J = 7.9 Hz, 2 H), 3.24 (t, J = 6.4 Hz, 1 H), 3.33 (tq, J = 2.3, 16.2 Hz, 2 H), 4.81 (dt, J = 1.3, 3.8 Hz, 2 H), 7.14-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 12.2, 12.4, 14.1, 20.6, 32.4, 35.8, 36.2, 59.3, 77.4, 77.5, 84.8, 105.9, 125.7, 128.3, 128.4, 142.3, 205.8. HRMS [(S)-methylbenzyl urea derivative, EI]: m/z calcd for C₂₈H₃₄N₂O: 414.2671. Found: 414.2654. Compound 6c (major diastereomer): ¹H NMR (CDCl₃): δ = 1.05 (t, J = 7.4 Hz, 3 H), 1.11 (t, J = 7.4 Hz, 3 H), 1.75-1.85 (m, 2 H), 1.98-2.12 (m, 2 H), 2.13-2.24 (m, 2 H), 2.61-2.74 (m, 2 H), 3.22-3.52 (m, 3 H), 4.95-5.00 (m, 1 H), 5.27 (dd, J = 1.3, 6.2 Hz, 1 H), 7.18-7.30 (m, 5 H). ¹³C NMR (CDCl₃): δ = 12.4, 13.4, 14.1, 22.1, 32.3, 36.3, 37.9, 56.9, 77.4, 84.9, 86.7, 93.9, 125.7, 128.3, 128.5, 142.2, 203.8.

10

The formation of 6c can be explained by two possible pathways. The first is the conversion of the alkynyl group into an allenyl group by deprotonation at the propargylic position of 2c, followed by [1,2]-allenyl migration (path A). The second is the deprotonation of [1,2]-shifted product (path B). However, the exact mechanism is unclear at present.

12

Attempts to separate the enantiomers of rearrangement products 3 and 4 at this stage using chiral HPLC (OD-H, AD) failed. Also, attempts to determine the diastereomer ratio of 7 and 9 using achiral HPLC analysis (ODS) failed.

13

Non-stereospecificity of the rearrangement of 2a is understandable if the electronic repulsion between the radical pairs prevents radical recombination. In that case, the radical recombination proceeds at a slow rate; thus, the chiral migrating radical can be racemized during the reaction.

14

HPLC analysis was carried out on a Chiralcel OD-H (0.46 × 25 cm) using heptane:i-PrOH:MeOH = 99.8:0.15:0.05 v/v (0.5 mL/min) as the mobile phase.

15

Low enantiopurity of (R,S)-8 is attributed to the racemization of chiral α-oxy lithium which generated via Sn-Li exchange.