One phosphine to rule them all

Buchwald et al: Angew Chem Int Ed Eng 45(39) 6523-7 (2006) DOI 10.1002/anie.200601612

Buchwald group developed a number of biphenyl-based phosphine ligands useful for Pd(0)-catalyzed arylations. Descendants from BINAP, these ligands form 1:1 complexes with Pd(0) that are particularly active catalysts, allowing arylations with both poor electrophiles (unactivated aryl chlorides) and lazy nucleophiles (such as heterocycles with very acidic NH).

The reason why the bulky monodentate electrone rich 2-biphenyl phosphines are good is that they make Pd(0) more nucleophilic – which helps the Pd insertion into Ar-X – while the bulk promotes the fast migration/dissociation rates in the intermeditate complexes and the formation of PdL2 is supressed. [Even a very sterically hindered PdL complex is coordinationally less saturated and hence more reactive than a PdL2 complex with a less bulky ligand. PdL4 is worse still]. The pi-electrons of the second benzene ring donate to Pd and the ring shielding prevents PdL2 from happening.

One of the problems with aryl-based phosphines has been cyclometallation, the palladacycles (such as Hermann catalyst) are good for high temperature Heck reactions but are quite inert below their decomposition temperature. In this case the cyclometallation is prevented by the ortho isopropyl groups.

There are other good phosphine ligands useful for difficult arylations: Xanphos (1), tert-Bu3P (2) and Josiphos (3). The great advantage of Buchwald phosphine ligands is that they are perfectly air stable, highly crystalline solids, easy to make and the produced catalysts are pretty active and robust.

This newest ligand with isopropyls on the ring and tert-butyls on the phosphine is reported to be very good for arylation of poorly reactive N-nucleophiles, with Pd2(dba)3 as a Pd source and NaOtBu in toluene (at 80C) or Cs2CO3 in dioxane (at 100C) . And some functional groups can be even left unprotected, such as primary amide and phenol, without interfering with the arylation process.

I have tried this ligand with my substrates and aryl bromindes and nearly everything that I did worked on the first try with 2-5% of Pd loadings and NaOtBu in toluene at 80C. The only exception was arylation using aryl iodide, as an electrophile – I suspect that the generated iodide anion complexed to Pd and prevented the catalyst from turning over. But all my aryl bromides worked beautifully.

The ligand is available from Aldrich, 638080-5G, and it is not too cheap ($161/5g) – so it would be worth making your own ligand for large-scale experiments. The phosphine can be made in a one-pot reaction from commercial reactants.

Note 1: Cheap, but often needs high temperature and a high-boiling solvent like o-dichlorobenzene. Best to be used 1:1 with Pd(0) since Pd(xanphos)2 is inert
Note 2: Expensive. The free tBu3P is super air-sensitive and should be handled only in glove box – it likes to smolder on air. Use the air-stable tetrafluoroborate salt (tert-Bu3PH) BF4 instead, 2:1 with Pd(0)
Note 3: Josiphos is a chiral ligand, very expensive

2-(2′-bromophenylamino)-thiazole-4-carboxylic acid

 

2-Bromophenyl isothiocyanate 1.0mL (1.593g, 7.44 mmol) was added dropwise into a stirred 7M anh NH3 solution in methanol (40mL), the mixture was stirred for 1 hour and then evaporated. The residue was dissolved in ethanol 25mL, heated to reflux, cooled, filtered from a small amount of cloudy precipitate and the filtrates were evaporated. The residue was induced to crystallization by addition of few drops of water, the solidified mass was dried on highvac. The obtained crude thiourea (1.737 g of a light tan crystalline solid) was dissolved in ethanol 50 mL with a gentle heating.The solution was cooled to RT and solid bromopyruvic acid 1.242g (7.44 mmol) was added, the mixture was stirred for 30 min and then evaporated. The residue was suspended in water 50mL with sonication (for 10 min), anhydrous sodium acetate solid 840mg was added to raise the pH of the mixture and the slurry was sonicated for additional 5 min. The precipitated product was collected by filtration, washed with water, dried by suction and then on highvac.
Y=1.749g (78.5%) of a white crystalline solid.

1H-NMR(d6-DMSO, 400MHz): 12.681 (very br s, 1H), 9.629(very br s, 1H), 8.144(dd, 8.2Hz, 1.6Hz, 1H), 7.718(s, 1H), 7.652(dd, 8.0Hz, 1.5Hz, 1H), 7.394(m, 1H), 7.039(m, 1H); LC-MS(+ESI): 299, 301

Using the same procedure, 1.0mL of 2-chlorophenyl isothiocyanate (1.310g, 7.72mmol) and 40mL of 7M methanolic NH3 provided 1.452g of a crude thiourea that was cyclized with bromopyruvic acid 1.289g (7.72 mmol) and neutralized with AcONa 900mg to provide 1.401g (71%) of 2-(2′-chlorophenylamino)-thiazole-4-carboxylic acid as a white crystalline solid. 1H-NMR(d6-DMSO, 400MHz): 12.705 (very br s, 1H), 9.779(very br s, 1H), 8.358(dd, 8.3Hz, 1.5Hz, 1H), 7.748(s, 1H), 7.477(dd, 8.0Hz, 1.4Hz, 1H), 7.350(m, 1H), 7.070(m, 1H); LC-MS(+ESI): 255, (257)

So many carbonyls, so little use

Lately, I have been enjoying the pleasures of 3,5-diketoesters. The effort turned out to be a complete waste of time. But the joy of cyclization – priceless.