One would have to be careful experimenting with something unknown thus far, in case one accidentally ended up with a GABA antagonist, or something even funkier. For instance, IIRC there are kappa-agonist benzos, as well as AMPAr antagonist 2,3-benzodiazepines, and I think, possibly, selective antagonists for AMPARs in the barbiturate class also, most barbs act as AMPA antagonists of a sort, stabilising AMPA receptors in a glutamate-bound, desensitized state, worryingly, with a very, very VERY long washout time (long enough to make me think it may require endocytosis and resynthesis of the receptor, 300 second wash time failing to remove more than a few percent, less than 20% by far in a patch-clamp evaluation of pentabarb) and there was the very concerning incident a while back with the release of methylmethaqualone, which caused muscle tremors, twitching and would likely cause siezures in many people, or at sufficient dose, AMPA antagonism for instance could potentially cause PERMANENT damage, ala domoic acid, which selectively fries the hippocampus, and buggers up muscarinic ACh receptor-bearing neurons coexpressing AMPARs located in the prefrontal cortex/basal ganglia, and causes potentially/likely permanent damage to both, resulting in severe, permanent inability to learn, form new memories, also personality changes and can wipe out executive function.
Also flurothyl, an oddball amongst the fluoroethers, in that it isn't an anaesthetic, but a GABAa antagonist (or inverse agonist, not sure which) and thus panicogen and convulsant.
So treat carefully, working up from a very, very small, submilligram quantity of any new Z-drug derivative and have somebody else on hand with the ability to quickly administer an IV or IM shot of a GABAa agonist. Would be scary indeed to accidentally end up with a competitive antagonist or inverse agonist, or worse, glutamatergic neurotoxin, the former could easily prove lethal in sufficient doses, and at the very best would prove to be a hideously unpleasant experience, I.E delirium tremens-inducing, psychosis producing, epileptogenic panicogen, the latter on the other hand, could do worse than kill you, but instead leave the user a literally permafried mindless zombie, but still alive-ish nevertheless.
Although there have been benzo derivatives discovered that are interesting GABAa antagonists, in that they are subtype-selective, and there are some that are weak partial inverse agonists at some of the more amnesic subunit-containing heteropentamers, such as alpha1betaX and alpha5betaX binding sites, whilst not being panicogens or convulsants, but instead, at some doses, possibly useable as nootropics, although one would have to take care with how much was used, and what if anything it was taken with.
zolpidem.pdf