Diclazepam (and active metabolites) has a T1/2 of over a week. It undergoes N-demethylation lorazepam, 3 hydroxylation to lormeazepam and BOTH to lorazepam.
We did actually produce a homologue with a 3-trifluoromethyl moiety that was even more potent (in spite of one isomer being inactive) and tried substituting the N-methyl with an N-methylcyclopropyl so it did not undego ANY metabolism but it got stupid. People took 2mg and were still zoned out 2 weeks later.
Of course, pyrazolam was also designed not to undergo any metabolism but it was designed to be redistributed and protein bound so it could be excreted unchanged. If I had to guess at a REALLY euphoric benzo, I wouldn't go for the most potent, I would go for pyrazepam.
Image pyrazepam-sk2 hosted in ImgBB
ibb.co
You may note that it can be made in a single step from bromazepam. The reason it's so interesting is that it's a potent serotonin releasing agent and so it is very euphoric. Generally it's the nitrobenzodiazepines that are serotonin releasers (nitrazepam, nometazepam, flunitrazepam, clonazepam, nitrazolam and so on) BUT their are 2 problems:
1)Nitrobenzodiazepines are toxic. While their are multiple recorded cases of people swallowing multiple grams of diazepam and living, nitrobenzodiazepines kill. In Sweden, 76% of suicides involving poisons were due to nitrobenzodiazepines alone.
2)Nitrobenzodiazepines have a high affinity for the α1 subtype of the GABA receptor. This is one of the sites alcohol has particular affinity for and is responsible for a lot of the negative effects such as aggression, retrograde amnesia, mood lability, vertigo, syncope.
In short, all of the nitrobenzodiazpeines I have come across have a very bad side-effect profile. Replacing the pendant aromatic from a ('2-substituted) benzene to a 2-pyridine overcomes all of these problems. Although we never got around to it, a 3-trifluoromethyl will increase a5 affinity so that it's MORE like alcohol but if you want a REALLY accurate alcohol mimic, benzos will do so. I do not 'similar to', I mean EXACTLY.
It turns out that all of the positive effects of ethanol are mediated by the α5 subtype.... but that means α5β1/ α5β2 α5β3. Finding a ligand that is selective to JUST those three subtypes (from the 28 identified subtypes) is rather complex. I believe Professor James Cook - Milwaukee Institute for Drug Discovery has elucidated all 28 subtypes and even he has not found a single ligand that has the appropriate selectivity. His best bet so far?
Image cook hosted in ImgBB
ibb.co
So good luck making that one!