A novel racemic codeine total synthesis was presented recently by Varin et al. (DOI). Many such efforts for codeine and closely related morphine (free phenol without methoxy group) are known, for example the codeine (-)-stereoisomer DOI or the (+)-stereoisomer DOI DOI
In a nutshell the procedure consists of a esterfication of 1 and 2 to 3, Heck reaction to 4, acetal deprotection to ketone 5 with triphenylcarbenium tetrafluoroborate, organic oxidation to dienone 6 with benzeneseleninic anhydride, cascading amidation and oxa-Michael reaction with N-methylbenzylamine to furan 7 (also involving work-up with potassium sodium tartrate) and organic reduction with lithium aluminumhydride to alcohol 8.
This compound reacts in a Eschenmoser-Claisen rearrangement (decalin , 215 °C) to amide 10, reduction with phenylsilane / titanium isopropoxide leads to aldehyde 11, a Friedel-Crafts hydroxyalkylation / dehydration reaction with p-toluenesulfonic acid gives 12, the benzyl group is replaced by a tosyl group in 13 (chloroethyl chloroformate to the free amine then tosyl chloride), allylic oxidation adds a hydroxyl group in 14 but with wrong stereochemistry necessitating an additional Dess-Martin oxidation / sodium borohydride ox/red sequence to inverted 16 and hydroamination with lithium, ammonia, t-butanol yields codeine 17.
This procedure is not scoring points with respect to enantioselectivity (it is just a racemate), with respect to having to fix incorrect stereochemistry along the way and with respect to overall chemical yield: 0.72%.