Intermediate Ketone: Method II
THC and all the variants can be prepared from alpha-resorcylic acid (3,5-dihydroxy-benzoic acid). As discussed in Appendix One it is also possible to start from gallic acid (3,4,5-trihydroxy-benzoic acid) in the case of THC and THC-II, and perhaps also in the preparation of the other variants. These materials are cheap and not difficult to obtain, however it is also possible to avoid some of the early steps by using commercial materials. At the present time the 3,5-dimethoxy-benzamide required for reaction (d) can be obtained commercially.

Note: Recently a synthesis of 3,4,5-trimethoxy-benzamide which is more convenient than that described here has been described. (See Schiemens; Berichte 92, page 860, 1959, Berichte 92, page 865, 1959).

This synthesis (Method II) has been well tried for THC and all the variants. Possible simplifications are shown on the chart (Reactions (j) and (m)) and described in the appendices. They are briefly as follows.

By use of a cadmium reagent (j) it has been shown that the benzoyl chloride can be converted to the ketone in one step and good yield. An error in procedure caused Suter and Weston (Reference 126) to fail in their attempt to use this reaction (Reference 136), after which they turned to reactions (c) and (d) as the alternative. Their errors have since been corrected and the method described in Appendix Two can be used for THC and all the variants. The main advantage or the procedure lies in the reduction in time required for the reaction (d). (It is very slow due to the insolubility of the magnesium complex in ether; see also Appendix Six). According to a patent (Reference 409) Reaction (j) gives a lower yield than the two-step operation (c)(d), however this may not be so.

If the method of benzylation described in Appendix Three can be used successfully with the 3,5-dihydroxybenzoic acid, then it is possible to eliminate the demethylation Reaction (h) through simultaneous reduction of the olefin and debenzylation in Reaction (m), using palladium catalyst and hydrogenation.

Synthesis Example: THC-V
The following reactions are described for 3,5-dihydroxybenzoic acid ( a-resorcylic acid). When using gallic acid (3,4,5-trihydroxybenzoic acid; for THC or THC-II only) the amount used should be increased (by weight) to maintain the same molar proportions with the other reactants, e.g. in section (b) the amount of benzoyl chloride used should be 66 grams, or about a 10% increase over the disubstituted material. The same increase (10%) should be maintained in subsequent reactions with gallic acid derivatives, up to reactions (k) or (g), after which the 4-methoxy group is removed and the materials are the same as from a-resorcylic acid. Likewise when the hydroxy-groups are protected by benzylation instead of methylation there should be a corresponding increase in the weight of the benzyloxy-compounds used, e.g. a 60%, increase in the weight of 3,5-dibenzyloxy-derivatives over the indicated amount of 3,5-dimethoxy-compounds.

In several of the reactions described below the use of nitrogen is indicated to protect against air oxidation. As described it is also possible to use city gas in some cases. It is suggested that nitrogen be obtained however, as it is cheap and much safer to use than an inflammable and poisonous gas. Reactions (a) and (b) have also been successfully carried out without any protective gas, but the yields are slightly lower.

Reaction (a). To a cold solution of 200 grams (5 moles) sodium hydroxide in 1250 ml. ice-water in a 5 liter flask there is introduced a protective blanket of nitrogen gas. When all air has been displaced the nitrogen stream is continued slowly and 154 grams (1 mole) of 3,5-dihydroxybenzoic acid is added. The flask is immediately closed and the mixture swirled or allowed to stand until all of the acid has dissolved. Then 220 grams (1.75 moles) of dimethyl sulphate (caution; dimethyl sulphate is highly toxic and gives off much poisonous vapor. The specific antidote ammonia solution should be on hand to wipe up any spills, and it is advisable to wash the hands frequently with dilute ammonia) is added at once and the stopper immediately replaced. The flask is cooled with crushed ice held in a towel and swirled gently for 20 minutes. The temperature is not allowed to rise above 30-35 deg. C. The stopper can be carefully loosened occasionally to release any pressure. After 20 minutes a second portion of 220 grams dimethyl sulphate is added and during the next 10 minutes the temperature can safely rise up to 40-45 deg. C., but cooling is still needed. At the end of this time the flask is fitted with a reflux condenser and the mixture boiled under reflux for two hours. To saponify the small amount of ester produced then add a solution of 50 grams sodium hydroxide in 75 ml. water and reflux another 2 hours. The reaction mixture is then cooled and acidified with dilute hydrochloric acid, which causes the 3,5-dimethoxy-benzoic acid to precipitate (dilute sulfuric acid can also be used). This crude product is filtered out and washed well on the filter with cold water. Without drying the wet crude material is added to 3 liters of boiling water and after a few minutes allowed to cool. The pure product is then filtered out and dried. The yield is about 88% of theory as nearly colorless crystals, melting point 180 deg. C. (References 407, 126 and 167)

Basically the same procedure is used to methylate gallic acid. 50 grams are added to 80 grams sodium hydroxide in 500 ml. cold water immediately stoppering the flask. Without opening the flask is shaken occasionally until all the acid has dissolved. Half (89 grams) the total quantity of dimethyl sulphate is then added all at once, cooling by running cold water on the flask to keep the temperature below 30-35 deg. C. Once methylation is under way the danger of oxidation is reduced so that the stopper can be occasionally lifted to release pressure. The second and last portion of dimethyl sulphate (89 grams) is added after 20 minutes and shaking continued another ten minutes, during which the temperature can rise to 40-45 deg. C. The flask is then fitted with a reflux condenser and boiled for 2 hours, another 20 grams sodium hydroxide in 30 ml. water is added, and reflux continued another two hours. After cooling and acidifying with dilute hydrochloric acid the precipitated trimethyl gallic acid (3,4,5-trimethoxy-benzoic acid) is filtered (best with suction) and washed on the filter with cold water. The yield is 90% of theory, 50-52 grams. (Reference 364)

Methylations on a larger scale can be carried out conveniently with yields above 95% by adding the dimethyl sulphate dropwise from a dropping funnel, mechanically stirring, and protecting against oxidation with a nitrogen stream passed slowly into the flask. Highest yield is obtained when the temperature is kept below 5 deg. C. by slow addition and by cooling the reaction flask in an ice or ice-salt mixture bath. It is also possible to use city gas instead of nitrogen, but the stirrer must be sealed to prevent ignition and explosion from the motor sparks and the overflow gas must be safely vented through a rubber tube. When the addition is completed the protective gas flow is stopped and refluxing continued as described above. Possibly use of a protective gas can be avoided by attaching a Bunsen valve to the flask. The valve serves to prevent entry of air but allows gas to escape. It can be simply made by cutting a vertical slit in a piece of rubber tube closed at one end, and attached by the other to a glass tube in the flask stopper. When addition is over the Bunsen valve can be replaced by the reflux condenser. (Reference 363 and 364)

Reaction (b). Conversion of 3,5-dimethoxybenzoic acid to the benzoyl chloride has been carried out with thionyl chloride (with or without solvent benzene), or with phosphorus pentachloride. Use of thionyl chloride is easier and preferable to phosphorus pentachloride. Contrary to the results obtained by most workers, Hurd (Reference 365) obtained much better yields of 3,4,5-trimethoxy-benzoyl chloride using phosphorus pentachloride than with thionyl chloride.

Example: Into a 250 ml. flask is introduced a stream of nitrogen, then 60 grams of 3,5-dimethoxybenzoic acid and 136 grams thionyl chloride. The flask is fitted with reflux condenser and the reaction mixture stirred and heated under reflux for 3 hours. The reflux condenser is then removed and replaced as a take-off condenser and the thionyl chloride is distilled off under water-pump reduced pressure. The brown residue is then distilled under a Claisen head (see a text book on distillation) using a protective nitrogen atmosphere. The yield of pure 3 5-dimethoxy-benzoyl chloride {boiling point 158 deg. C./20 mm.) is about 87% of theory (52 grams). (Reference 410 and 407)

Example: 60 grams 3,5-dimethoxy-benzoic acid, 200 ml. thionyl chloride, and 300 ml. thiophene-free benzene are warmed in a flask on the boiling water bath. When reaction is completed the benzene and excess thionyl chloride are rapidly distilled off, first at ordinary pressure, then the last portions are removed under water-pump reduced pressure. (The reference does not indicate time required for the reaction; it may be necessary to perform the reaction under a reflux condenser). (Reference 167 and 365)

Example 60 grams 3,5-dimethoxy-benzoic acid and 69 grams phosphorus pentachloride are mixed together and heated at 100 deg. C. on the boiling water bath for 1/2 hour (15 minutes was found sufficient for the trimethoxy-benzoic acid, Reference 365). After cooling the product can be isolated by extracting it with ether (leaving behind insoluble inorganic material). In the original work the ether solution was treated by passing in ammonia gas (Reference 126) but probably the ether could be distilled off (water-bath, electric hot plate) and the residue dropped into concentrated ammonium hydroxide. Alternatively the crude reaction mixture has been distilled under water-pump reduced pressure to remove the volatile phosphorus oxychloride and some unreacted pentachloride, the residue being then treated with concentrated ammonium hydroxide. (Reference 126, 168 and 365)

(Note It is likely that the reduced pressure distillation purifications described above can be omitted for the present purpose. Most of the thionyl chloride can be distilled off at ordinary pressure, then the crude residue cooled and cautiously poured into ice-cold concentrated ammonium hydroxide. Perhaps a larger excess of ammonia should be used than is given in (c). Also, if purification is to be omitted, it is particularly desirable to minimize formation of impurities by using a protective gas during the reaction reflux period.

Recent examples of the conversion of the acid to chloride have been described (Lambooy; J.A.C.S. 76, page 137, 1954: Davies; J.C.S. page 2784, 1955: and Bycroft; J.C.S. page 2063, 1962).

Reaction (c). For conversion to the benzamide, the crude benzoyl chlorides obtained as above from 60 grams 3,5-dimethoxy-benzoic acid are gradually stirred into 600 ml. of ice-cold concentrated ammonium hydroxide, cooled in a flask by an ice-salt mixture. After completing the addition stir the white flocculent suspension for another 1/2 hour before filtering and wash the product on the filter with water. The overall yield from the acid through the benzoyl chloride is about 90% of theory for the 3,5-dimethoxy-material and at least 80% for the 3,4,5-trimethoxy-benzamide. (References 167 and 365).

When the crude 3,5-dimethoxy-benzyol chloride is used without purification the product will be impure and should be recrystallized to obtain a nearly colorless product for the following reaction. Several recrystallizations may be needed. For this purpose the crude colored material should be dissolved in boiling water containing decolorizing charcoal (5 grams), boiled a few minutes, then cooled in a refrigerator after filtering out the charcoal. The minimal amount of water should be used. Other solvents which can be used are methanol, benzene, and ligroin. Also it has been recommended that purification be carried out by dissolving the crude material in chloroform, then by adding petroleum ether, the pure 3,5-dimethoxy-benzamide is precipitated. (References 365, 167 and 411)

Reaction (d). This is a Grignard reaction; references (References 185 and 186) should be consulted for the general method and precautions. The previously prepared 3,5-dimethoxy-benzamide is reacted with the Grignard reagent (alkyl magnesium halide) selected to give the appropriate ketone intermediate for the desired THC or THC-variant. The best yields are obtained when alkyl bromides are used and these are preferable. Alkyl chlorides can be also used, and are more often available commercially. The following chart shows (1) the alkyl halide used for the particular THC or variant (as bromides), and (2) the product ketone with it's boiling point as found by earlier workers at reduced pressure, then also calculated for water-pump reduced pressure. The calculated boiling point is only approximate, but will serve as a guide for purification. The symbol R- stands for 3,5-dimethoxy-phenyl-, so that for example under THC, R-(n-butyl)-ketone should be read; 3,5-dimethoxy-phenyl-(n-butyl)-ketone.

THC
1. n-butyl bromide (1-bromo-butane) (1.25 mole = 170 grams)
2. R-(n-butyl)-ketone, b.p. 176 deg. C./ 11 mm. or 190-200 deg. C./25 mm. It is a solid at ordinary temp., m.p. 42.5 deg. C. Yield 80% of theory.

THC-II
1. n-amyl bromide (1-bromo-pentane)(1.25 mole = 186 grams)
2. R-(n-amyl)-ketone; b.p. 175 deg. C./ 7 mm. or 200-210 deg. C./25 mm. A solid, m.p. 53 deg. C., yield 88%.

THC-III
1. n-heptyl bromide (1-bromo-heptane) (223 grams)
2. R-(n-heptyl)-ketone; b.p. 155 deg. C./ 0.5 mm. or 235-45 deg. C./ 25 mm., yield 60%.

THC-IV
1. n-octyl bromide (1-bromo-octane)(1.25 mole = 240 grams)
2. R-(n-octyl)-ketone; b.p. 180 deg. C./ 1 mm. or 250-260 deg. C./ 25 mm., 77% yield.

THC-V
1. 2-bromo-heptane (methyl-n-hexyl bromide)(223 grams)
2. R-(1'-methyl-hexyl)-ketone, b.p. 147 deg. C./ 1 mm. or 210-220 deg. C./ 25 mm., yield 82%.

THC-VI
1. 2-bromo-octane (methyl-n-heptyl bromide)(240 grams)
2. R-(1'-methyl-heptyl)-ketone, b.p. 157 deg. C./ 0.1 mm. or 270-280 deg. C./ 25 mm., yield 63%.

Sources and preparations for the required alkyl halides are given in Appendix Seven.

Example: In a flask fitted with reflux condenser containing a solution of n-butyl magnesium halide in 250 ml. dry ethyl ether, prepared from 1.25 mole of 2-bromo-heptane (223 grams; weights of other alkyl bromides given above) with 31 grams of magnesium in the usual way, there is added as rapidly as refluxing of the ether will permit 46 grams (0.25 mole) of 3,5-dimethoxy-benzamide. A further 300 ml. anhydrous ether is added and the reaction mixture refluxed and stirred under a protective atmosphere for 50 hours. Ordinary city gas can be used to protect against entry or air, only a slow stream of gas is passed into the top of the air-tight reaction flask. The overflow gas is conducted by rubber tube from the top of the reflux condenser to a safe vent.

At the end of 50 hours, the product is hydrolyzed by adding a mixture of ice and a little water to which has been carefully added 80 ml. concentrated sulfuric acid. After stirring for a time the ether layer is allowed to separate from the aqueous portion by settling and then removed with a separatory funnel or by decantation. The ether solution of product (3,5-dimethoxy-(n-butyl)-ketone) is dried by shaking with anhydrous magnesium sulphate, filtered, and the ether distilled off. About 48 grams of the crude ketone are obtained.

In early work these ketones were purified by Adams using reduced pressure distillation alone (a mechanical pump capable of giving a vacuum of around 0.1-1 mm. mercury was used, in all probability the 20-30 mm. vacuum obtainable with a water-aspirator pump will give satisfactory results). It may be possible to use the crude ketone without purification in the subsequent reaction, and if not, then partial purification by distillation at water-pump reduced pressure may be sufficient to give a satisfactory material. In later papers (Reference 187) Adams states that the best method of purification is by use of the Girard reagent T (commercially available). In this method a water soluble salt of the ketone is prepared with the Girard reagent T and the impurities are then extracted from the aqueous solution by ether. The salt is then decomposed with acid and the free ketone can then be extracted into ether, from which it is recovered. The following procedure is based on similar operations and should give satisfactory results. Sufficient Girard reagent T is used to form a salt with all of the ketone and a small excess does no harm. The first ether extract can be treated with a small amount of additional reagent T and re-extracted with water to be sure that sufficient is being used for a given weight of ketone (if any ketone can be isolated from the water extract then more than 50 grams is needed).

To the crude ketone obtained as above there is added excess Girard reagent T (50 grams should be sufficient for the first trials), 1 liter methyl alcohol, 100 ml. glacial acetic acid, and the whole is refluxed for 10 minutes. Sufficient sodium carbonate solution is then added to neutralize 90%-95% of the acetic acid (the solution can be brought to neutrality, then 5 ml. acetic acid added, or a Bromo-thymol Blue testing paper can be used directly to determine the end-point for alkali solution addition)

Afterwards the solution is made ice-cold in a salt-ice mixture or freezer and while still cold the impurities are extracted by severa1 portions cold ether. The ether extracts are themselves extracted with a small amount of water to recover a small amount of ketone which may be removed with the impurities. The main water layer and the aqueous extract are combined and the ketone liberated from complex with Girard reagent by adding dropwise a measured amount of concentrated hydrochloric acid or 10% sulfuric acid until the solution is just acid as tested by Congo Red indicator paper; at this point an additional amount of the same acid equal to one-tenth the volume already used is added. After standing at room temperature for 35 to 60 minutes, the pure ketone is extracted with ether in several portions. The ether extracts are washed with dilute sodium carbonate solution, dried with an anhydrous salt such as Drierite or magnesium sulphate, filtered, and the purified ketone isolated by distilling off the ether. (Reference 197).

THC and the other variants are prepared in the same way, except that the appropriate alkyl halide is substituted for the 2-halogen-heptane used here for THC-V. The same molar ratio of halide to benzamide should be maintained, e.g. the amount (weight) of the alkyl halide used should be 1.25 mole, or 125% or the molecular weight of the halide, in grams.

This is a five-fold molar excess of alkyl halide (0.25 mole benzamide vs. 1.25 mole alkyl halide), the excess being used to ensure a good yield based on the prepared or expensive substituted benzamide.

The Grignard reaction itself has been successfully used for the intermediates for THC and all the variants in satisfactory yields. The yields obtained by early workers are shown in the chart above (References 126, 129, 130, and 187) and can probably be improved. The long reaction time is the major disadvantage of this synthesis (it is due to the immediate formation and precipitation of an ether-insoluble magnesium salt of the benzamide). It is possible that the modification of the Grignard reaction described in Appendix Six will improve the solubility of the reactants, in which case the required time will be shortened, even to a few hours. Alternatively the method described in Appendix Two can be used. This latter method has much to recommend it and it is probably the preferable technique, even better than the two-step reaction (c)(d).


Intermediate Ketone: Method I
Conversion of the Ketones