of course you can make it work but it’s more expensive and harder than normal commercial synthesis. with such limited analytical methods you might not even be able to tell that something went wrong (melting points can be completely fucked up by traces of solvent, same with IR) naloxone is actually a nice example of how you can design synthesis in order to make your life easier, in terms of purification at least. one of commercial syntheses looks like this

anyway that’s what i found in one chinese patent (CN104230945A). they start from oxycodone, which they make from thebaine, which they don’t have to make because it’s a natural product. there’s a very useful purification technique, column chromatography, that works like this: you take a tube, fill it with silicagel, then pour hexane through it, then put your compound on top of that silicagel column. silicagel is polar, and the more polar compound is, the stronger it is bound to it. so now in order to get your pure compound you wash this column with increasingly polar solvents, and this way you can recover various compounds in order from least polar to most. this allows you to get pure compounds most of the time

now, small changes in or near polar groups mean that polarity changes little. in particular, we can ignore for five minutes all these hydroxyls and just focus on amine because it’s the most polar part and any changes at other groups make polarity change in the same way anyway. step 2 is von braun degradation, and it removes methyl and turns amine into amide. this amide is very likely to have much lower polarity than starting product and so it can be easily separated on column. here specifically, it’s no longer basic so unreacted acetyloxycodone can be just washed away with acid extraction, which is good for both low tech and large scale. step 3 is hydrolysis, which gives us amine back, which again gives more polar product which means again that separation is easy. if some acetyloxycodone was left there, it would give very similar compound, just with that N-methyl left and separating it would be pretty hard. this here is especially important as N-methyl and N-allyl compounds have opposite effects

that weird synthesis with burgess reagent and everything else also allows for separation of these unreacted starting materials, but this requires column chromatography in the middle. if you skip it, maybe it’ll turn out alright or maybe you just fucked up for the last time. maybe mCPBA was old, or maybe burgess reagent or THF was wet, or any variety of random bullshit that someone inexperienced would not even register. and this is only one point that brings us to point 2: that jarduino is really juicero of backyard chemistry. to begin, you need dry solvents, and this is something you should do yourself if you want it to be done right. this means you need to reflux THF with drying agent of your choice, then distill it away to clean dry bottle, fill it with argon, ideally add molecular sieves and tape it shut. then you have to reflux something, this requires all glass flask because that 3d-printed head will melt under these conditions. at some point you also need sep funnel and rotovap and column, and so on and so on and it turns out that you already have all glassware to do it all the normal way and jarduino isn’t even needed for anything. (also i don’t understand their choice of peristaltic pumps, as they already put everything in syringes, this means that you can just use syringe pumps and use normal medical long needles for piping of sorts. it’s very chemically resistant as long as plungers aren’t rubber, which their demos had for some weird reason)

this is closer to the synthesis i thought they were using

if you swap oripavine for oxymorphone and cyclopropylmethyl bromide for allyl bromide, you should get naloxone, probably, in 2 steps. this doesn’t work keeping cyclopropylmethyl bromide and using oxymorphone gets you naltrexone which also works as antagonist and even slightly more potent one, it’s from this paper https://cdnsciencepub.com/doi/10.1139/cjc-2014-0552

then there are all the sterility requirements, as they want to make injectables in garage. i’m not pharmacist or biologist, but i expect it would be a bit harder to make it reliably and repeatably than what they make it look like. i take that your wife is a medical professional so if she weighed in on that, that would be nice

also, part of their advice is to just eyeball powder as a part of identification. this is a bad idea because how stuff looks like and behaves can change depending on humidity, traces of solvents if any, and ambient temperature, especially for low melting solids, and crystal habit can vary from batch to batch anyway and it’s not a reliable indicator of anything