The manufacture of codeine from morphine

Abstract

The methylation of morphine is one of the key operations of the opium alkaloid industry, as up to 90% of the manufactured morphine is converted into codeine, the methylether of the phenolic hydroxyl of the morphine. The latter is a rather precious raw material for a commercial-scale manufacture, and many attempts have been made in the Course of almost one century first to realize and then to bring to perfection this phenolic methylation of the morphine. After discussing chronologically this development, the present paper describes a methylating method which can be seen to be its logical consequence.

Details

Author: Walter R. Heumann
Pages: 15 to 17
Creation Date: 1958/01/01

The manufacture of codeine from morphine

D.Sc. Walter R. Heumann
Associate Professor of Chemistry, University of Montreal, Montreal, Canada

The methylation of morphine is one of the key operations of the opium alkaloid industry, as up to 90% of the manufactured morphine is converted into codeine, the methylether of the phenolic hydroxyl of the morphine. The latter is a rather precious raw material for a commercial-scale manufacture, and many attempts have been made in the Course of almost one century first to realize and then to bring to perfection this phenolic methylation of the morphine. After discussing chronologically this development, the present paper describes a methylating method which can be seen to be its logical consequence.

In the chemical equations throughout this article, the abbreviated formulae M.OH and M.OCH 3 will be used for morphine and codeine respectively.

The first to achieve the methylation of morphine was Grimaux (5) in 1881, using methyl iodide as the methylating agent in alkaline solution:

(1) M.OH + CH 3I + KOH → M.OCH 3 + KI + H 2O

During the years following Grimanx's synthesis other classical methylating agents such as dimethyl sulfate or diazomethane were used to prepare codeine (7, 8). All these reactions had one important disadvantage - namely, that the nitrogen of the morphine reacted easily with one molecule of the methylating agent, forming quaternary ammonium compounds. These secondary reaction products are unstable and decompose easily, thus causing considerable losses of morphine. Numerous German patents (4) are evidence of the efforts made to overcome this drawback. A relatively satisfactory solution was found in 1909, when the German manufacturer C. H. Boehringer (1) introduced the use of methyl groups containing quaternary ammonium compounds as methylating agents. Such quaternaries cannot, of course, give the above-mentioned side reactions with the nitrogen of the morphine. It was thus that trimethylphenylammonium chloride was to become the outstanding methylating agent for morphine for many years. The reaction involved was described by the inventor as follows:

(2) M.OH + NaOH + (CH 3) 3C 6H 5N.Cl → M.OCH 3 + (CH 3) 2C 6H 5N + NaCl + H 2O

The reaction was carried out in ethanol solution, and it worked satisfactorily only at temperatures between 120°C and 130°C. The solution was therefore heated in an autoclave; whereby the pressure reached 4 to 5 atm. A detailed descrip-tion of a commercial-scale application of this method was given by Schwyzer in 1927 (10). This author dissolved the morphine in an alcoholic solution of sodium ethoxide, added then the equivalent amount of trimethylphenylammonium chloride, and heated the autoclave until the pressure reached 4 atm. which was maintained for one hour. The methylation is not complete, and between 5% and 10% of the morphine can be recovered as such. The yield of codeine is approximately 90% of the theoretical equivalent of the morphine used up by the reaction. The loss of about 10% of morphine may be due to the rather crude conditions of temperature and pressure combined with both the strongly alkaline reaction of the mixture and the presence of water. It is known that morphine is easily destroyed by oxidation in alkaline aqueous solution.

A certain simplification of the process was a method first published by the British forces' intelligence reports on German technical processes after World War II (2). This method, which according to this report had been used by a German manufacturer since 1910, avoided the use of the autoclave as follows: The morphine was dissolved in a solution of potassium hydroxide in absolute ethanol. To this solution of potassium morphinate an equivalent amount of trimethyl-phenylammonium chloride is added. This mixture is then added slowly to xylene, which is kept at 125°C, whereby the alcohol distills off together with the water present. Only 2% of the morphine remains unchanged, according to the report, which, however, fails to mention the yield of codeine. That the methylation apparently goes further than in the autoclave seems to be due to the fact that the solution is immediately freed of water as soon as the temperature rises. As will be shown later, there is strong reason to believe that water has an inhibiting effect on the reaction.

A further attempt towards improvement was the use of the free trimethylphenylammonium base instead of its chloride, first described by Rodionov in 1926 (9). This author added the salt of the ammonium base to an alcoholic solution of sodium ethoxide and thus obtained a solution of the free base. The reaction is described by the author as follows:

(3) (CH 3) 3C 6H 5N.Cl + NaOH → (CH 3) 3C 6H 5NOH + NaCl

The sodium chloride precipitates and is filtered off. To this solution morphine is added, and the temperature is raised to 110°C, whereby the alcohol distils off. The author gives the following equation for the reaction:

(4) M.OH + HO 2N(CH 3) 3C 6H 5 → M.OCH 3 + C 6H 5N(CH 3) 2 + H 2O

According to Rodionov - who described only laboratory experiments - about 20% of the morphine was recovered, the remaining 80% yielding approximately 85% of theoretical equivalent of codeine.

It may be noticed that Rodionov used a solution of alkali ethoxide to prepare the free quaternary base, but gives an equation for the reaction with alkali hydroxide. Furthermore, he obtained morphine by heating to 110°C only, instead of the usual 120°C to 130°C. The solution of his methylating agent was perhaps a mixture of the free quaternary base with its ethoxide, which could form as follows:

(5) (CH 3) 3C 6H 5N.Cl + Na.OC 2H 5 → (CH 3) 3C 6H 5NOC 2H 5 + NaCl

Such alkoxides of quaternary ammonium bases were first described by Ingold & Patel in 1933 (6). They had to be prepared under strictly anhydrous conditions, as they were found to be very easily decomposed by water. Rodionov's solution was prepared and handled without these precautions, and was therefore either a solution of the free quaternary base or the above mixture of free base and its ethoxide.

One German manufacturer, as reported by the British forces' intelligence service (3), obtained better yields of codeine with a certain modification of Rodionov's method. The solution of the free quaternary base is prepared by adding its hydrochloride to an alcoholic solution of potassium hydroxide, whereas Kodionov used sodium ethoxide. The morphine is then added slowly to xylene, which is kept at 125°C. The yield of codeine is reportedly 95%.

As shown by equations (2) and (4), the methylation has been described hitherto by one summary equation only. However, the reaction proceeds in fact in two distinct steps, as will be demonstrated by the new method described below. The morphine reacts first with the quaternary base to form the morphinate of the latter. This first step is an acid-base reaction involving an extremely weak acid, the morphine. Such reactions are difficult or even impossible to achieve in water, but they are favoured by solvents of lower dielectric constants, and could possibly become quantitative in solvents such as anhydrous alcohols, provided the formation of water by the reaction itself is avoided. This may explain why, in the methods cited, a certain amount of morphine remains unreacted. As will be shown later, this first step takes place at ordinary temperature. In the second step, the morphinate of the quaternary base splits into codeine and dimethylaniline by the action of heat.

It thus seems that water affects the methylation in two ways. As already pointed out, it may contribute to the oxidative destruction of morphine and thus lower the yield of codeine. On the other hand, it may also prevent the first step of the methylation from proceeding to completion. Water is introduced by using commercial morphine base, which always contains one molecule of water of crystallization, and also by the solvent ethanol, which is hygroscopic and absorbs up to 5% of water, when the reaction is carried out in contact with the atmosphere.

When the reaction is carried out according to equations (2) or (4), water is also formed in the first step as the product of neutralization. If, however, the free quaternary base or its salt were replaced by its ethoxide, ethanol would be formed instead of water, and the methylation would thus become easier. Following the line of these considerations, it appears logical to use the pure ethoxide of the quaternary base and to work under anhydrous conditions. Such a methylation method is now to be described here.

The ethoxide of trimethylphenylammonium can be prepared according to equation (5) by dissolving under strictly anhydrous conditions the hydrochloride of the quaternary base in a solution of sodium ethoxide in absolute ethanol at a temperature not exceeding 14°C. The sodium chloride which crystallizes is separated by decantation or vacuum filtration. The resulting clear light-brown solution of the quaternary ethoxide is stable for at least several days when protected from humidity and stored at or below 14°C. Its concentration is determined immediately before use by dissolving an exactly weighed sample in at least ten times its weight of water and titrating with 0.1 N standard acid.

Anhydrous morphine base dissolves readily in this cold solution to form the morphinate of the quaternary base. This is the first step of the methylation, which can be presented as follows:

(6) M.OH + C 2H 5O.N(CH 3) 3C 6H 5 → M.ON(CH 3) 3C 6H 5 + C 2H 5OH

When this solution is heated up to 110°C, the second step takes place and the morphinate splits into codeine and dimethylaniline as follows:

(7) M.ON(CH 3) 3C 6H 5 → M.OCH 3 + C 6H 5N(CH 3) 2

When anhydrous conditions were to be maintained throughout these reactions, commercial morphine base could not be used as such, because of its water of crystallization. It releases this water only slowly when heated to at least 110°C, and prolonged heating can cause partial decomposition of the alkaloid. The water is, however, released rather easily and without any damage to the alkaloid, when the latter is heated with toluene, which boils at 111°C and causes the water to distil off in an azeotropic mixture at a temperature somewhat below the boiling point of the toluene.

The methylation can be carried out as follows. The commercial morphine base is suspended in pure toluene by stirring. The mixture is heated to bring about distillation, which starts between 90°C and 100°C and is continued until the boiling point reaches 111°C. The remaining suspension mixture, which is now perfectly free of Water, is kept protected from humidity and cooled down to 14°C under stirring. Then, an appropriate quantity of the titrated solution of the methylating agent is added, this quantity containing an excess of 5%, The morphine dissolves immediately, and the solution is then brought to boiling; which starts at about 80°C. The alcohol distils off with a part of the toluene and distillation is continued until the boiling point again reaches 111°C. The remaining solution contains now codeine and dimethylaniline, but does not contain any morphine. The codeine can be extracted in the usual manner with dilute mineral acid, whereby the aqueous phase is kept only slightly acid to litmus (pH 4 - 4.5) to prevent dimethylaniline from being extracted together with the codeine.

The reactions are quantitative, and no morphine remain unchanged, nor is morphine lost by decomposition or secondary reactions, when temperature and moisture are carefully controlled. This method has been used successfully in commercial-scale manufacture for several years, and under favourable conditions the yield of codeine can reach up to 98%. It has been applied so far only to the methylation of morphine. Taking into account the ease with which the reactions proceed to completion under rather mild conditions, it can be expected that it will prove useful also for the methylation of phenolic hydroxyl groups in other nitrogen containing compounds.

References

  1. BOEHRINGER, C. H. & SON, German Patent 247,180 (1909)

  2. British Intelligence Objectives Sub-Committee, B.I.O.S. Final Report No. 116, p. 90, H.M.S.O., London, 1946.

  3. British Intelligence Objectives Sub-Committee, B.I.O.S. Final Report No. 766, pp. 220 and 228, H.M.S.O., London 1947.

  4. German patents 39,887, 92,789, 95,644, 96,145, 102,364, 107,225, 108,075, 131,980, 189,843, 214,783, 224,388, 247,380.

  5. GRIMAUX, M., Comptes rendus, 92, 1140, 1228 (1881).

  6. INGOLD, C. K. & PATEL, C. S.,J. Chem. Soc. (London), 1933/I, 69.

  7. PECHMANN, H. VON, Berichte, 27, 1888 (1894).

  8. PECHMANN, H. VON, Berichte, 28, 1624 (1895).

  9. RODIONOV, V. M., Bull. Soc. chim. France (4), 39, 305 (1926).

  10. SCHWYZER, J., Die Fabrikation der Alkaloide, Julius Springer, Verlag, Berlin 1927.