This file is a part of the Rhodium site archive. This Aug 2004 static snapshot is hosted by Erowid
as of May 2005 and is not being updated. > > Back to Rhodium Archive Index > >

Synthesis of Mandelic Acid and Derivatives

Mandelic Acid

This a-hydroxy acid is prepared from benzaldehyde through the cyanohydrin, which is produced by interaction of the benzaldehyde bisulfite addition compound with potassium cyanide. This interchange reaction eliminates the hazard of working with volatile, toxic hydrogen cyanide, but even potassium cyanide is dangerous if safety measures are not observed. The interchange reaction is reversible and excess cyanide is used to shift the equilibrium in favor of the cyanohydrin. Mandelonitrile, a liquid, undergoes changes on standing and hence should be processed further without delay; it is extracted with ether, the solution is carefully washed free of cyanide ion, and the nitrile is then hydrolyzed with hydrochloric acid. An intermediate ketimine hydrochloride, C6H5CH(OH)C=NH*HCl soon gives way to mandelic acid and ammonium chloride, both of which are soluble in the aqueous acid. The solution is cooled and extracted with ether, and the ether is then displaced by benzene for crystallization, since mandelic acid is much less soluble in this solvent than in ether.


Potassium cyanide is extremely poisonous if it gets into the blood through a cut, if transferred from the hand to the mouth, or if it comes into contact with acid and liberates hydrogen cyanide. Never touch the solid with the fingers; be very careful not to spill it and if you do clean it up at once. Take care not to spill a solution or reaction mixture containing the substance, and discard the waste solution directly into the drain and wash down the sink with excess water.


In a 125 ml Erlenmeyer flask dissolve 11 g of sodium bisulfite in 50 ml of water by brief swirling, add 10 ml. of benzaldehyde and swirl vigorously and stir until the oily aldehyde is all converted into the crystalline bisulfite addition compound. Cool to room temperature but not below, add 14 g. of potassium cyanide (caution) and 25 ml. of water (rinse down the walls), and take the mixture to a hood. Swirl and stir for about 10 min. until all but a trace of solid has dissolved (break up lumps with the rod). Mandelonitrile separates as a thick oil. Pour the mixture into a separatory funnel, rinse the flask with small amounts of ether and water (and at once wash the flask free of cyanide), and then shake the mixture vigorously for a full minute to insure complete reaction. Add 50 ml. of ether, shake, and run the aqueous solution down the drain. Wash the ether extract with 25 ml. of water and then with 25 ml. of saturated salt solution (note and account for the difference in appearance of the ether layer). Then run the solution into a 125 ml distilling flask containing 15 ml. each of coned. hydrochloric acid and water. Add a boiling stone, stopper the flask, and mount it on a steam bath for distillation through a condenser into an ice-cooled receiver. Distill off the ether and then disconnect the condenser, note the time, and continue heating on the steam bath with frequent swirling to mix the layers and promote hydrolysis. Note that the initially lighter-than-water layer of mandelonitrile gradually changes to an oil of density greater than that of the aqueous acid. In a little less than 1 hour the oil dissolves to a clear solution; however if the solution is cooled at this point it will become cloudy from separation of unhydrolyzed oil still present. Hence continue heating for one half hour more to complete the reaction and then cool to room temperature. Measure 100 ml. of benzene into a 250 ml. distilling flask and make a mark at the level of the liquid. Transfer the acid solution to a separatory funnel and rinse the flask with a little ether (5 ml). Then add to ml. of ether, shake well, let the layers separate, and draw off the aqueous layer into a second separatory funnel or a dry flask, and wait a few minutes for separation of a little more aqueous solution. Run the ethereal extract into the flask containing benzene, and reextract the aqueous solution with two further 20 ml portions of ether, and add the extracts to the flask containing benzene. Add a boiling stone, connect the flask to a condenser and an ice-cooled receiver, put a thermometer in place, and distil the solvent, cautiously at first since the flask is very full. Water is eliminated gradually by azeotropic distillation and the boiling point rises as ether and water are eliminated. Continue the distillation until the solution has cleared, water is no longer evident in the neck of the distilling flask, and the volume of the solution has been reduced to the 100 ml. mark. Disconnect the flask and inspect the bottom to see if a trace of ammonium chloride has separated, as either a solid or a gum, make sure that no flames are near, and decant the hot solution into a 250 ml. Erlenmeyer flask. Crystallization usually starts soon; yield 10.0 gram, mp 118-119 deg C. In case no crystals appear, pour the solution into a 25 ml. distilling flask, evaporate the solvent, scratch to induce crystallization, and digest the solid with benzene.

Note: When mandelic acid that has been crystallized from benzene is allowed to stand in contact with the mother liquor for several days, the needles gradually change into granular, sugarlike crystals of a molecular compound containing one molecule each of mandelic acid and benzene. These crystals are stable only in contact with benzene and at temperatures below 32.6 deg C on exposure to the air the benzene of crystallization evaporates and the mandelic acid is left as a white powder.


To a 40C solution of 40 g NaHSO3 and 250 mL water 34 g benzaldehyde (0.324 mol) is added with vigorous stirring. After 20 minutes cool the reaction mixture down to 0C, add 200 mL Ether and drip in a solution of 17 g NaCN in 70 mL water. Agitate for 30 minutes at 0C, then seperate the organic phase and extract the aqueous phase twice with 50 mL portions of ether (or toluene). The combined ether extracts are washed with 2x100ml conc. NaHSO3 solution to remove traces of unreacted benzaldehyde (dont skip this step), washed 1 x 50ml brine then dried over anhydrous calcium chloride. After rotovaping the solvent under a vacuum, the remaining oil is allowed to solidify upon cooling then placed in the vacuum dessicator to remove last traces of water to give mandelonitrile in high yield.

Acetylmandelic Acid

In a 500-cc. Claisen distilling flask with a low side tube connected to a condenser, are placed 105 g. (0.69 mole) of mandelic acid (m.p. 118) and 151 g. (137 cc., 1.92 moles) of acetyl chloride. A reaction sets in without the application of heat (Note 1). As soon as a clear solution results, the flask is warmed on a water bath and the mess acetyl chloride is distilled. The last trace of acetyl chloride may be removed by prolonged drying in a vacuum. The acetylmandelic acid then crystallizes in large, round, white clusters after one or two days standing. The yield is 130-133 g. (97-99% of the theoretical amount) (Note 2).

Acetylmandelyl Chloride

To the crude acetylmandelic acid still containing some acetyl chloride obtained as described above, is added 250 g. (149 cc., 2.1 moles) of thionyl chloride. The reaction starts at once without warming but it is necessary to reflux for four hours to complete the reaction (Note 3). The excess thionyl chloride is then distilled and the residue distilled under reduced pressure (Note 4). The yield is 115-120 g. (79-82 per cent of the theoretical amount) of almost colorless liquid boiling at 125-130/10 mm. (150-155/33 mm.).


  1. Occasionally the application of a little heat is necessary to bring about a more rapid acetylation.
  2. The melting points given in the literature range from 39 to 80. The acetylmandelic acid is difficult to crystallize but may be purified from benzene or chloroform, preferably the former. The product thus obtained melts at about 79-80.
  3. Prolonged refluxing of the acetylmandelic acid with the thionyl chloride tends to lower the yield.
  4. The boiling point of acetylmandelyl chloride has been reported(1) as 129/10 mm. The pressure should be reduced as low as possible, to avoid the formation of tar during the distillation.