Appendix One

When it is desired to prepare either THC or THC-II it is possible to modify the synthetic method described in the main section in several ways. Instead of the three steps shown on the chart as (e,f,g), a single reduction step is used to convert the ketone to the alkyl benzene (k). From 1,3,5-trichlorobenzene this gives either THC or THC-II in only five steps, or from a-resorcylic acid in 6 or 7 steps.

An additional advantage when THC or THC-II are to be made is that it is possible to replace the a-resorcylic acid (3,5-dihydroxy-benzoic acid) by cheaper and more readily available garlic acid (3,4,5-trihydroxy-benzoic acid). When using garlic acid (see main section) the ketones obtained have three methoxy- groups on the benzene ring; when these ketones are reduced (k) with sodium metal and an alcohol the methoxygroup in position 4- is simultaneously split off, giving the desired 3,5-dimethoxy-(n-alkyl)-benzene.

It was shown by Suter and Weston (Reference 126) that this reduction method is not satisfactory for ketones which are in the first place unsubstituted in the 4- position (such as prepared from a-resorcylic acid, or by method I, from 1,3,5-trichlorobenzene). Also from similar reactions it seems unlikely that benzyloxy- substituted ketones would split off the group at position 4-. When 3,5-substituted ketones are to be reduced directly in the preparation of THC or THC-II the operation is carried out by the Wolff-Kishner method, described in Appendix 4.

The modification starting from gallic acid does not apply to preparations of the variants III through VIII, because here the ketone is not directly reduced but must be converted via (e) and (f) to an olefin, which is then reduced as described in main section (g) by methods which will not split off the methoxy group at 4-. Quite possibly the 3,4,5-trimethoxy- compounds could be used through reaction (f), and then the reduction of the olefin in (g) could be carried out as described here with sodium and alcohol rather than as described in the main text, with the result that simultaneous reduction of the olefin and splitting off of the 4- methoxy- group would occur just as in (k). However this has not been tried and experiment is needed. Olefins can be reduced very well with sodium and alcohol and there seems no reason why the splitting off of the 4- group would not occur just as when reducing the ketone.

The reduction is best carried out in anhydrous amyl alcohol since in this case no water cooling is needed for the reflux condenser. Such large amounts of vapor are liberated that a glass reflux condenser is inadequate and should be replaced by a simple aluminum or copper tube. The reaction flask should be of sufficient size to contain rather violent boiling and it should be set in a sand bath, not on a steam bath or water bath, or wire support (the thickness of the sand between the bottom of the flask and the directly-heated metal pan should be about 1/8 inch to 1/4 inch).

In place of isoamyl alcohol it is usually possible to use n-butyl alcohol with as good results, but in all cases the alcohol should be dry. In recent work sodium hydride has been substituted for sodium metal in similar reductions and it may be possible to use sodium amalgam instead of sodium. The latter reagent has no advantage for the present procedure, however a method has been described in which the amalgam is generated by electrolysis of alkali with a mercury electrode. The apparatus is described by Ingersoll in (Reference 374), and may be advantageous when larger amounts are involved. It is not recommended that these variations be experimented with unless resources are adequate to identify the products as those desired.

Reduction by Sodium
In a one-liter Pyrex flask place about 11 grams of 3,4,5-trimethoxy-phenvl-(n-butyl)-ketone (for THC; the same weight of the n-amyl ketone is taken for THC-II) and 1/2 inch dry isoamyl alcohol. The flask is fitted with a rubber stopper penetrated by a length of aluminum pipe, about 1/2 inch in diameter and three feet long. The flask is tied in place on a sand bath and gently heated by a flame (the temperature of the bath should not exceed 100-110 deg. C.). A piece of sodium metal is added through the aluminum tube before the solution has become very warm so that the violence of the reaction can be judged. A total of about 60 grams sodium (in not-too-small pieces) and 600 ml. dry isoamyl alcohol are then added in small portions alternately. By occasional use of the external heat on the sand bath and control of the rate of sodium-alcohol addition, the reaction mixture is kept at as gentle a reflux as possible. After the sodium has all dissolved the reaction solution is poured carefully into two liters of water and acidified with hydrochloric acid. The alcohol is then distilled off arid the residual aqueous solution extracted with ether in several portions. The extract is dried with an anhydrous salt and after filtration obtained as an oil by distilling off the ether. Purification (if necessary) can be carried out as described in the main text; Asahina purified the product by distillation under reduced pressure (b.p. 145 deg. C./ 4 mm.) and obtained a colorless oil in a yield of about 8 grams. (Reference 160)

Unless the alcohols are completely anhydrous the yield will suffer. Such alcohols are commercially available, or drying techniques are given in Organic Syntheses CV I, page 249: Organic Syntheses CV II, pages 469 & 475.


Preparation and Use of (+)-p-2,8-menthadien-1-ol
Appendix Two