Dibenzylacrylamide (synthesised in house, TOXIC!)
tributyltin hydride (Lancaster, TOXIC!)
neat triethylborane (Aldrich, TOXIC, explosive, flammable!)
t-butyl iodide (toxic!)
dichloromethane (distilled from CaH₂)
potassium fluoride (for work-up)
Silica

Dissolve up dibenzylacrylamide (200mg, 0.796mmol, 1eq) in DCM (21mL) with stirring and degass for about 10 min by bubbling nitrogen through the solution. Add t-butyl iodide (3.18mmol, 4eq, 379uL) and tributyltin hydride (4.1eq, 878uL, 3.26mmol) under nitrogen with stirring. Add triethylborane (46 uL, 0.318 mmol, 0.4 eq) and then using a syringe bubble 10mL air through the solution. Monitor by TLC (KMnO₄), reaction is normally complete within 2h, but may be left overnight without formation of by-products. Once complete, remove the solvent in vacuo and add KF (4eq, 200mg), water (4eq, 60uL) and ethyl acetate (ca. 3mL), stir overnight. Filter off the KF/Bu₃SnF solid through Celite with a little MgSO4 added to absorb the water. Vacc off the EtOAc and then repeat the work-up procedure with stirring for 2 h, filter as above. Purification by gradient column chromatography (petrol to petrol:ethyl acetate 8:1 v :v) afforded the product (4,4-Dimethyl-pentanoic acid dibenzylamide) in 60-66 % yield as a colourless oil.

This procedure is a reliable method to gain these kinds of products in around 60% yield. The reactions have been repeated about 10 times on scales of about 150mg to 500mg of the starting acrylamide. Yields are not improved by performing the reaction at lower temperatures, but yields do improve slightly on scale up (probably less as a proportion is lost in the work-up). Triethylborane as a 1.0M solution in hexanes or THF and can also be used without loss in yield. This is good because the neat stuff is extremely dangerous since it reacts explosively with oxygen. The work up is a bit tedious but the final products may be isolated in high purity after chromatography. If a large excess of tributyltin hydride is evident before the work-up then this can be converted to tributyltin iodide by the dropwise addition of 3% iodine solution in DCM until the solution no longer decolourises. The KF workup will then remove the Bu₃SnI. Alternatively, most of the tributyltin hydride should be removable by flushing the column with petrol at the start. When performing the filtration through Celite, don't swill out the flask excessively as you will just dissolve up the Bu₃SnF left behind and eventually it will dissolve and come through the Celite. Just use enough to be sure you've dissolved up the crude product. The success of other carbon centred radical additions depends on the nucleophilicity of the radical and decreases in the order 3^o>2^o>1^o as expected. Finally it's worth remembering that acrylamides, tin compounds, triethylborane and alkyl iodides are all extremely toxic (especially as solutions in DCM) so wear blue nitrile gloves and take care when handling these reagents (in a fume hood).

R_f (product, SiO₂): 0.37 (4:1 petrol:EtOAc). ¹H NMR (300 MHz, CDCl₃)delta (ppm): 7.15-7.40 (10 H, m, Ar-H), 4.61 (2 H, s, rotamer, -CH₂-) 4.45 (2 H, s, rotamer, -CH₂-), 2.38 (2 H, app. quintet, -COCH₂-), 1.63 (2 H, app. quintet, -CH₂-C(Me)₃), 0.86 (9 H, s, C(CH₃)₃).

For examples of enantioselective radical additions using Et₃B/O₂ system with Lewis acids see: 1. Urabe et al., J. Org. Chem., 1995, (60), 3576-3577. 2. Sibi & Ji, J. Org. Chem., 1997, (62), 3800-3801.

For removal of tin residues using the method above see: Caddick et al., Synlett, 1993, 231-232.