The Gatterman Aromatic Formylation
[ Back to the Chemistry Archive ]
Simplification of the Gattermann Synthesis of Hydroxy Aldehydes 
The Gattermann synthesis of hydroxy aldehydes (2) which consists in saturating an anhydrous ether solution of certain phenols and anhydrous hydrogen cyanide with dry hydrogen chloride, sonietimes with the addition of anhydrous zinc chloride, gives excellent yields of products which are readily purified. The method has proved to be invaluable for the preparation of certain intermediates in the synthesis of many natural compounds, and is still the'only avaihble process for preparing many representative hydroxy aldehydes. In spite of the ease with which the reaction generally takes place, its use in the laboratory is not as frequent as might be expected. This is due to the necessity of using anhydrous hydrogen cyanide, a product the preparation and handling of which involve many disagreeable features. More recently (3), this method of preparation of the hydroxy aldehydes has been simplified somewhat by the generation of the hydrogen cyanide and its direct addition to the reaction mixture. Even this procedure does not make the preparation safe for any but the more experienced investigator. Such reactions as the Gattermann synthesis, involving anhydrous hydrogen cyanide, are consequently limited in their use.
Recently Karrer (4) has shown that bromocyanogen may be substituted for hydrogen cyanide in the preparation of certain hydroxy aldehydes. The yields, however, do not appear to be as satisfactory as when hydrogen cyanide is used.(3b) Bromocyanogen is more conveniently made in the laboratory and more easily handled than anhydrous hydrogen cyanide, but even this product is extremely poisonous and much care must be used. Moreover, only a freshly prepared sample of bromocyanogen gives satisfactory results.
This research had as its object the modification of the Gattermann synthesis in such a way as to avoid using anhydrous hydrogen cyanide. Experiments in which hydrogen cyanide was replaced by certain of its salts have been carried out. Of these various salts, zinc cyanide would seem the most promising choice, since it would be changed during the reaction into a mixture of zinc chloride and hydrogen cyanide. Zinc chloride has already been shown by Gattermann to be an effective condensing agent for the reaction of the hydrogen cyanide with certain phenols in the presence of hydrogen chloride. The laboratory results have fully met the expectations. Very satisfactory yields have been obtained with resorcinol, a-naphthol, b-naphthol, orcinol and pyrogallol; in fact, the yields were similar to those obtained with anhydrous hydrogen cyanide. Undoubtedly any phenol that can be converted to an hydroxy aldehyde using the directions of Gattermann can be converted to an hydroxy aldehyde equally well by this slight but most convenient modification. The preparation of zinc cyanide suitable for use in the reaction described above proved to be a problem in itself. Up to the present time zinc cyanide free from zinc oxycyanide has been made by the action of hydrogen cyanide upon zinc oxide or by other methods entirely unfitted for the present problem, the success of which depends upon the ease of preparation of the zinc cyanide. An easy method for preparing a satisfactory zinc cyanide for the Gattermann reaction was finally found. It consisted in treating an aqueous solution of sodium cyanide with magnesium chloride in order to precipitate the impurities of sodium hydroxide and sodium carbonate, Zinc cyanide was then precipitated by adding to the solution a molecularly equivalent amount of zinc chloride in alcohol. The product obtained is about 90% zinc cyanide and contains no impurity that interferes with the Gattermann synthesis.
The literature reports that zinc cyanide gradually decomposes on standing. The product made as described, if kept in a dry atmosphere, apparently decomposes only very slowly. After several weeks it was analyzed and showed a depreciation of about 2%. Nevertheless, it gave just as satisfactory results as when first prepared. No extensive experiments were made to determine just how long the zinc cyanide could be kept, since it can be so easily and quickly prepared.
Gattermann has also described the preparation of aldehydes from phenol ethers by the action of hydrogen chloride, hydrogen cyanide and anhydrous aluminum chloride on phenol ethers. Moreover, he has shown that certain hydroxy aldehydes which could not be made by the direct condensation of the phenol with hydrogen cyanide, hydrogen chloride and zinc chloride could be formed by the action of hydrogen chloride, hydrogen cyanide and anhydrous aluminum chloride upon the phenol. A modification of these last processes in such a way as to avoid the anhydrous hydrogen cyanide has proved successful and is now being studied in detail.
General Method of Preparation of the Hydroxy Aldehydes. A 500ml. wide-mouth bottle was fitted with a stopper holding an efficient mechanical stirrer with a mercury seal, a reflux water condenser and an inlet tube, with wide mouth to prevent clogging from the precipitate, extending nearly to the bottom of the bottle. To this inlet tube was attached a safety bottle and to this a generator producing dry hydrogen chloride. To the top of the condenser was connected a tube leading into a wash bottle containing sulfuric acid; from this a tube was led to a safety bottle and from the latter a tube to the surface of a sodium hydroxide solution.
In the reaction bottle was placed 20 g. of the hydroxy compound in 150 to 200 cc. of dry ether. Sufficient dry zinc cyanide was then added, equivalent to 1.5 mol. for every mol. of phenol. The mechanical stirrer was started and dry hydrogen chloride was passed in rapidly. The zinc cyanide gradually disappeared with the formation of a milky solution (the milkiness being probably due to the sodium chloride present) and as more and more hydrogen chloride dissolved, the imide hydrochloride condensation product began to separate as a thick oil. In 10 to 30 minutes this oil turned to a solid. At the end of about 1.5 hours the ether was usually saturated with hydrogen chloride. When this point was reached, the stream of gas was passed in more slowly and stirring was continued for 1/2 hour longer to be certain that all the phenol had reacted. The ether was decanted from the solid material and the imide hydrochloride then decomposed with water or dil. alcohol as described below.
The amount of ether used in the above reaction was somewhat greater than that suggested by Gattermann. It was found, however, that the colored by-products which invariably formed with these phenolic compounds were smaller in amount the larger the amount of solvent. It was also possible to use dry chloroform in place of dry ether. Under these conditions the reaction went practically the same, but the crude products were somewhat more colored.
Resorcyl Aldehyde.-With resorcinol the reaction mixture turned pink and the solid material tended to be pink. Previous investigators advised the decomposition of the imide hydrochloride by boiling it with water for about 5 minutes and then allowing the product to cool, It was found that by this procedure the resorcyl aldehyde which separated amounted to about 70% of the calculated amount. It was found to be much more satisfactory to add about 100 cc. of water to the imide hydrochloride, raise the solution t o the boiling point, filter the mixture and then immediately allow the filtrate to cool. In this way about a 50% yield of aldehyde was obtained. This was filtered and the filtrate was allowed to stand. In about 10 to 15 hours, more material separated which, upon filtration, amounted to about 45% of the calculated yield, giving a combined yield of 95%. The resorcyl aldehyde obtained had only the slightest tinge of color and melted very sharply at 135-136C (Gattermann, 136C). By recrystallization from water with the addition of boneblack, a product absolutely free from traces of colored by-products was produced.
Orcinol Aldehyde.-The orcinol used in these experiments must be thoroughly dried to remove the water of crystallization which is ordinarily present. The reaction proceeded smoothly with the development of a pink color. No special precaution was necessary in the decomposition of the imide hydrochloride. It was merely boiled for two to three minutes with 100 cc. of water, filtered and allowed to cool. The yield of product was about 85%; m.p.~ 178-180C (Gattermann, 180C). By a crystallization from water with the addition of boneblack a very pure product was obtained.
b-Naphthol Aldehyde.-The imide hydiochloride was decomposed by boiling it with 100 cc. of water and an 85% yield of product was obtained, melting a few degrees below the correct point. After recrystallization from water with the addition of boneblack, a very pure product melting at 80-81C (Gattermann, 80-81C) was obtained.
a-Naphthol Aldehyde.-In contrast to the hydroxy compounds described above where the reaction mixture was pink, that from a-naphthol became deep yellow. The imide hydrochloride was decomposed by boiling it with 700 cc. of 30% alcohol until complete solution took place; then the solution was filtered and allowed to cool. When water was used in this experiment, the a-naphthol aldehyde, bequse of its insolubility in water, tended to separate immediately and to prevent the smooth and complete decomposition of the imide hydrochloride. The product obtained as described formed in 72% yield and melted slightly low. More aldehyde could be recovered from the mother liquor. By recrystallization from water or 30% alcohol with added boneblack a very pure product resulted; m.p., 178C (Gattermann, 180C)
Pyrogallol Aldehyde.-The ether in the reaction mixture turned more deeply colored than when the other phenols were used, but the imide hydrochloride was light colored. It was decomposed by boiling it for two to three minutes with 400 cc. of water. The yield was about 45%, and the product melted at 158C (Gattermann, 158C). It can be obtained very pure by recrystallization from water containing boneblack.
Simplification of the Gattermann Synthesis of Hydroxy Aldehydes II 
In a recent communication from this Laboratory, a convenient simplification of the Gattermann reaction for the synthesis of certain hydroxyaldehydes was described. Various phenols were treated in dry ether with zinc cyanide, and then dry hydrogen chloride was passed in. By this procedure, anhydrous hydrogen cyanide was formed in the reaction mixture and condensed with the hydrogen chloride and phenol to give a condensation product which was hydrolyzed to an hydroxyaldehyde. The zinc chloride which was produced at the same time acted as an effective condensing agent. Thus the most disagreeable feature of the Gattermann reaction, the handling of anhydrous hydrogen cyanide, was avoided. An easy method for the preparation of zinc cyanide was also described.
Gattermann(6) found that, in order to prepare aldehydes from certain phenols or phenol ethers, it was necessary to use anhydrous aluminum chloride as a condensing agent with the phenol or phenol ether, hydrogen cyanide and hydrogen chloride, zinc chloride being unsatisfactory. A study has been made of the synthesis of the aldehydes from various phenols and phenol ethers, using aluminum chloride, hydrogen chloride and zinc cyanide in place of hydrogen cyanide. The reactions run very smoothly, and the expected products were obtained thus making the synthesis of these aldehydes possible without the use of anhydrous hydrogen cyanide. Phenols or phenol ethers and zinc cyanide were mixed with benzene as a diluent so that stirring could be accomplished more readily, and dry hydrogen chloride passed in to the saturation point. Anhydrous aluminum chloride was then added and hydrogen chloride again passed in. The imide hydrochlorides of the aldehydes separated. These were decomposed with hydrochloric acid and the aldehydes obtained by one or another of the usual procedures.
The zinc chloride which is formed in the reaction mixtures does not interfere in any way with the condensations. The only noticeable difference between this procedure and the one using anhydrous hydrogen cyanide is that the imide hydrochlorides almost always separate from the reaction mixtures as a sticky mass, probably a mixture or a compound with zinc chloride. In contrast to the imide hydrochlorides obtained from those phenols (1) which formed by condensation in dry ether with zinc cyanide and hydrogen chloride, these imide hydrochlorides were not so readily decomposed by water or very dilute hydrochloric acid, but required refluxing with 10-20% hydrochloric acid. They were then rapidly decomposed with the production of aldehydes in yields equal to those usually obtained, provided anhydrous hydrogen cyanide was used in place of zinc cyanide.
The method has been applied to the preparation of aldehydes from anisole, p-cresol methyl ether, resorcinol dimethyl ether, diphenyl ether, b-naphthol methyl ether, o-cresol and thymol with excellent results. These represent a sufficient number of condensations to warrant the conclusion that any aldehydes that can be synthesized by the older Gattermann procedure may be prepared equally well by this modification.
Zinc Cyanide.-Zinc cyanide for this investigation was made according to the procedure described in the previous communication on the modification of the Gattermann reaction. It has been found that conversion of 200 g. of zinc chloride to zinc cyanide can easily be carried out in about 30 minutes. When the 96-98% potassium cyanide was used for the preparation, no magnesium chloride was necessary and the zinc cyanide was 95-98% pure.
Phenols and Phenol Ethers.-The phenols and phenol ethers were made by the usual procedures and boiled over a range of 2C.
General Method of Preparation of Various Phenols and Phenol Ethers. A 600cc. wide-mouth bottle with a flat bottom was fitted with a stopper holding a mechanical stirrer with a mercury seal, an adapter holding a reflux water condenser, and an inlet tube, preferably with wide mouth to prevent any chance of clogging from the precipitate, extending nearly to the bottom of the flask or bottle. To the inlet tube was attached a dry hydrogen chloride generator. To the top of the condenser was connected a tube leading to a wash bottle containing sulfuric acid; from this a tube was led to a safety bottle and from the latter a tube to the surface of a sodium hydroxide solution.
In the bottle was placed about 30 g. of the phenol or phenol ether in approximately three times its volume of dry benzene. Sufficient powdered dry zinc cyanide was then added, equivalent to about two molecules for every molecule of phenol or phenol ether. The mixture was cooled, the mechanical stirrer was started and dry hydrogen chloride was passed in rapidly for half an hour to an hour. After this treatment, the mixture was kept cool, and then, without stopping the stirring, the condenser was removed and 1.5 molecular equivalents of powdered anhydrous aluminum chloride was slowly added at one time. Stirring was continued and hydrogen chloride again passed in very slowly while the mixture was heated at 40-45C for three to four hours.
The material was added to an excess of 10% hydrochloric acid, which generally caused a heavy precipitate of imide hydrochloride to separate. This mixture was refluxed for half an hour, thus causing a rapid decomposition with the formation generally of an oily product, the aldehyde. The aldehyde could then be extracted directly or filtered from this hydrogen chloride-benzene mixture or it could be steam-distilled and then extracted from the distillate and finally purified by distillation or crystallization. The benzene was used so that mechanical stirring could be carried on more satisfactorily. Although Gattermann did not use this solvent in all cases, it is necessary to do so when zinc cyanide is employed.
An attempt was made merely to use powdered sodium or potassium cyanide in the above reaction in place of zinc cyanide. The results were entirely unsatisfactory, the yields of products being very low. An attempt was also made without success to replace the benzene by other solvents.
The same conditions as were found most suitable by Gattermann for the preparation of each of a large number of aldehydes can undoubtedly be used when zinc cyanide is employed in place of hydrogen cyanide. The only difference is the more vigorous hydrolysis of the imide hydrochlorides after they have formed and the necessity of always using a solvent in the initial condensation.
Anisole. From 30 g. of anisole, 52 g. of zinc cyanide, 45 g. of aluminum chloride and 65 g. of benzene was obtained 35.5 g. (quantitative) of anisaldehyde boiling at 246-248C. After decomposition with hydrochloric acid, the mixture was steamdistilled, The benzene was removed and the anisaldehyde, with traces of anisole, collected. This was separated and shaken with sodium bisulfite. The traces of anisole were thus removed, and the anisaldehyde was obtained by the action of alkali and then distilled.
p-Cresol Methyl Ether. From 30 g. of p-cresol methyl ether, 52 g. of zinc cyanide, 45 g. of aluminum chloride and 75 g. of benzene, was obtained 29.5 g. (80% of the calculated amount) of aldehyde boiling at 250-252C. It was obtained from the reaction mixture in exactly the same way as the anisaldehyde.
Resorcinol Dimethyl Ether. From 25 g. of resorcinol dimethyl ether, 40 g. of zinc cyanide, 38 g. of aluminum chloride and 100 g. of benzene, was obtained 25 g. of aldehyde (an 80% yield). The product was obtained as described under the two previous aldehydes but was purified by crystallization from ligroin, from which white needles were formed melting at 71C.
Diphenyl Ether. From 30 g. of diphenyl ether, 52 g. of zinc cyanide, 45 g. of aluminum chloride, and 70 g. of benzene, was obtained 18 g. of aldehyde (a 50% yield); b.p., 188-190C (20 mm.). The aldehyde was obtained as described above.
o-Cresol. From 30 g. of o-cresol, 52 g. of zinc cyanide, 45 g. of aluminum chloride and 90 g. of benzene, was obtained 14.5 g. of aldehyde (a 38% yield).
Since the aldehyde could not be steam-distilled, the unchanged o-cresol and benzene were removed by steam distillation and the aldehyde salted out from the residual solution, then extracted with ether. It was readily purified by crystallization from hot water with boneblack. It yielded white crystals melting at 118C.
Thymol. From 20 g. of thymol, 37 g. of zinc cyanide, 30 g. of aluminum chloride and 60 g. of benzene, was obtained 23.5 g. of aldehyde (a quantitative yield). The aldehyde was isolated in exactly the same way as that from o-cresol and formed white needles melting at 133C.
8-Naphthol Methyl Ether. From 30 g. of @-naphthol methyl ether, 52 g. of zinc cyanide, 45 g. of aluminum chloride and 150 g. of benzene, was obtained 36 g. of aldehyde (a quantitative yield). The aldehyde was isolated as were those from thymol and o-cresol. It was purified by crystallization from ligroin, giving white needles; m.p. 83C.
Preparation of Zinc Cyanide 
To a 5% excess over 1 molecular equivalent of technical sodium cyanide calculated as 100% pure, regardless of its purity, dissolved in about 25% more than an equal weight of water, was added magnesium chloride solution until no more precipitate of magnesium hydroxide and carbonate formed (this requires an amount of magnesium chloride sufficient to precipitate a quantity of sodium carbonate equal in weight to about 74 % of the sodium cyanide used). The precipitate was filtered off immediately and the filtrate added at once to 1 molecular equivalent of zinc chloride dissolved in as small an amount as possible of 50% alcohol. The zinc cyanide precipitated and was filtered off. When the magnesium hydroxide and carbonate were not removed immediately and the zinc cyanide precipitated at once, the reaction mixture turned dark until finally it became almost black. When the zinc chloride was added to such a colored solution the zinc cyanide formed was always colored. The zinc cyanide was washed on the filter with alcohol, then with ether and dried in a desiccator or in an airbath at 50C.
The only important precaution in this preparation is to insure an excess of zinc chloride over sodium cyanide. If the sodium cyanide is in excess, the zinc cyanide invariably precipitates as a sticky mass which is difficult to filter and unsatisfactory for the preparation of the hydroxy aldehydes.
The product was analyzed for the amount of cyanide present by titrating it with standardized silver nitrate solution, and was shown to be about 90%, pure zinc cyanide. The remaining 10% was presumably for the most part sodium chloride, with small amounts of zinc chloride, magnesium chloride and perhaps traces of basic zinc cyanide. These do not interfere with the Gattermann reaction.
 Adams, J Am Chem Soc, 45, 2373 (1923).