Using palladium acetate in Suzuki and Heck couplings

Palladium(II) acetate is not itself the catalyst. In the reaction mixture it is reduced to a palladium(0) species, usually by the phosphine ligand, the boronic acid, an amine base or the solvent, and that palladium(0) enters the catalytic cycle: oxidative addition of the aryl halide, transmetalation from the boronic acid (Suzuki) or migratory insertion of the alkene (Heck), and reductive elimination or β-hydride elimination to release the product. Because the reduction happens in situ, anything that changes how fast palladium(0) forms changes the reaction.

That is why batch quality matters. Commercial palladium acetate can contain the nitrite-bridged impurity Pd₃(OAc)₅(NO₂), left from manufacture with nitric acid, and an insoluble polymeric form. Both alter the induction period and the amount of active catalyst, and are the usual explanation for a coupling that worked at 1 mol% last month and fails at 1 mol% today. Our palladium(II) acetate is the pure trimer, Pd₃(OAc)₆, and its ¹H NMR spectrum in CDCl₃ shows a single acetate signal near 2.0 ppm. Check any new bottle the same way if you have the instrument.

Use 224.50 g/mol (the formula weight per palladium) for all calculations, so that 1 mol% means one palladium atom per hundred molecules of limiting substrate. 22.4 mg is 0.100 mmol of palladium.

For an aryl bromide and an arylboronic acid on a 1 mmol scale:

1. Charge a dry flask or vial with the aryl bromide (1.00 mmol), the boronic acid (1.3 mmol), potassium carbonate (2.0 mmol, finely ground) or potassium phosphate (2.0 mmol), palladium(II) acetate (2.2 mg, 1 mol%) and the ligand: SPhos (8.2 mg, 2 mol%) for general use, or triphenylphosphine (10.5 mg, 4 mol%) for reactive substrates.
2. Add 1,4-dioxane (4 mL) and water (1 mL), both degassed by sparging with nitrogen for 15 minutes or by three evacuate–backfill cycles. Oxygen is the main cause of boronic acid homocoupling and of palladium black.
3. Seal under nitrogen and heat at 80–100 °C with good stirring for 2–16 hours, monitoring by TLC or LC-MS.
4. Cool, dilute with ethyl acetate, wash with water and brine, dry and concentrate. Filter through a short pad of silica or Celite to remove the bulk of the palladium before chromatography.

Expected: full conversion within a few hours for electron-poor and electron-neutral bromides; electron-rich or ortho-substituted bromides may need 100 °C and the full 16 hours. Aryl iodides couple faster and tolerate lower loading (0.1–0.5 mol% Pd with PPh₃). Aryl chlorides need the bulky, electron-rich biaryl phosphines (SPhos, XPhos, RuPhos) or an N-heterocyclic carbene; with triphenylphosphine they do not react.

Choosing the ligand is the main decision. Triphenylphosphine (2–4 equivalents per Pd) is adequate for iodides and activated bromides and is cheap. SPhos is the best general ligand for bromides and chlorides, including hindered biaryls; XPhos for heteroaryl chlorides and amination; and tri-tert-butylphosphine (as its air-stable HBF₄ salt) for very hindered couplings. For aqueous or ligand-free conditions with aryl bromides, palladium acetate alone in ethanol–water with potassium carbonate at 80 °C works for many simple substrates and is easy to scale.

Boronic acid problems are the second most common failure. Two-pyridyl, 2-furyl and polyfluoroaryl boronic acids protodeboronate quickly under basic aqueous conditions; use their pinacol or MIDA esters, or the trifluoroborate salts, and keep the water content low. A 1.2–1.5-fold excess of boronic acid compensates for modest decomposition.

For an aryl iodide or bromide and an acrylate or styrene on a 1 mmol scale:

1. Charge the aryl halide (1.00 mmol), the alkene (1.5 mmol), palladium(II) acetate (2.2–4.5 mg, 1–2 mol%) and tri(o-tolyl)phosphine (6–12 mg, 2–4 mol%) for bromides; iodides often need no ligand at all.
2. Add triethylamine (2.0 mmol) or potassium carbonate (2.0 mmol) and 2 mL of degassed DMF, DMAc or NMP.
3. Heat under nitrogen at 100 °C (iodides) to 140 °C (bromides) for 4–24 hours. The mixture typically darkens and may deposit some palladium black towards the end.
4. Cool, dilute with water, extract with ethyl acetate, wash several times with water to remove the amide solvent, dry and concentrate.

Electron-poor alkenes (acrylates, acrylonitrile, enones) react at the terminal carbon and give the (E)-product selectively; styrenes give mainly the (E)-stilbene with some 1,1-isomer. For aryl iodides at low temperature, Jeffery's conditions (tetrabutylammonium bromide 1 equivalent, potassium carbonate, DMF, 60–80 °C, no phosphine) are mild and fast. For high turnover numbers on bromides, the Herrmann–Beller palladacycle, itself made from palladium acetate and tri(o-tolyl)phosphine, runs at 0.01–0.1 mol%.

The Heck reaction tolerates less: aryl chlorides, alkenes bearing β-hydrogens on both sides and substrates with free thiols or terminal alkynes are all problematic. Alkene isomerisation after the coupling is suppressed by added silver or thallium salts in classical work, or simply by stopping the reaction as soon as the halide is consumed.

• Degas properly. Three freeze–pump–thaw cycles are ideal; 15 minutes of nitrogen sparging is usually enough for Suzuki couplings at 1 mol% Pd. Low-loading reactions are the most oxygen-sensitive.
• Weigh palladium acetate last and do not let it sit in solution with the ligand before heating. Mix, degas, heat.
• Grind inorganic bases and stir fast: Suzuki couplings with potassium phosphate are often mass-transfer limited.
• Concentration of 0.2–0.5 M in limiting halide is a good default; very dilute reactions favour catalyst decomposition.
• Palladium black appearing early (within minutes) means the ligand-to-palladium ratio is too low or the solvent is not degassed; appearing late is normal.
• Scale-up changes heat transfer and gas exchange rather than chemistry; keep the same concentration and stir rate, and allow a longer induction period.

Pharmaceutical intermediates are limited to low parts-per-million palladium (ICH Q3D gives an oral permitted daily exposure of 100 µg per day, often translated to 10 µg/g), and even for research compounds residual palladium can interfere with biology and subsequent catalysis. Options in increasing order of effort: filtration through Celite; a wash with aqueous N-acetylcysteine or thiourea; stirring the crude solution with a thiol- or thiourea-functionalised silica scavenger (1–5 equivalents relative to palladium, 1–16 hours, then filter); activated carbon; recrystallisation. Collect all palladium-containing solids and solutions for recovery rather than disposal.

• No conversion, starting material intact: oxygen in the solvent, wet or old base, wrong ligand for the halide (triphenylphosphine with a chloride), or the catalyst never formed. Repeat with fresh degassed solvent, fresh palladium acetate and SPhos.
• Homocoupled biaryl from the boronic acid: oxygen. Degas harder; add the boronic acid last.
• Dehalogenated arene: hydride sources such as alcohols or formate, or excessive heating. Switch solvent to dioxane–water or toluene.
• Conversion stops at 50–70%: catalyst death. Add a second portion of palladium acetate and ligand, or move to a precatalyst (Pd G3/G4) that forms the active species cleanly.
• Inconsistent results between bottles: check the palladium acetate by ¹H NMR; a second acetate peak or insoluble residue means an impure batch.