Yeah, it actually makes more sense to do it directly in the oil or butter right from the start. Different guides suggest different ways like using a rack or putting it in a Pyrex dish with a lid on etc. Loads of variables will be introduced there, say if someone put a cold dish and just the weed into an oven, even if its preheated, it will take quite a long time for the dish to warm up and way, way longer for the weed to heat up because the rate of heat conduction through still air in the dish and between the pieces of weed would be so low, even if the bowl is hot. It'd also depend on if you were planning to take the bowl out to cool or leave it in while the oven coasts down.
There are tons of threads on many sites asking about butter/oil weed ratios etc with a lot of confusing things.
There's this gigantic document entirely about weed from a pharmaceutical/chemistry perspective here...
[/URL]
(WHO expert committee on drug dependence, critical review, delta 9 tetrahydrocannabinol)
Around page 13 you can see the method by which they decarboxylate their sample material which is dry at 105c for 15 minutes and then 120c for 2h to achieve 95% decarboxylation. They then supercritically extract it and purify it via chromatography. They're doing that because they want a sample of just the thc itself to use for other tests.
The simplest approach if you're after butter would be to just melt sticks in a Pyrex dish with a lid, wait for it all to actually come up to temperature, drop the weed in and leave it to simmer around 120c for 2-3h. One advantage to that is it will help keep the stank down a little with more of the volatiles dissolving into the butter/oil rather than just directly evaporating off. You can improve that even more, the stank reduction, if you wrap some ptfe tape round the top first to help air seal the dish better. "Gas fitters tape", the kind that comes in yellow packs, is slightly thicker than regular ptfe. They both cost tens of cents/pence per roll.
An interesting thing is around page 18 they describe the solubility of thc as being about 0.3g/ml of sesame seed oil. So..... saying you had 10oz (280g) of weed at 28%, that'd be (280/100)*28 = 78g of thc in total. At a solubility of 0.3g/ml sesame oil, you could potentially dissolve all of that into 78g thc / 0.3 g/ml = 260ml of sesame oil. That would be the absolute maximum. Butter sticks are usually about 250g and it has a density of around 0.9g/ml (which is why it floats on water), so 1 stick at 250g*0.9g/ml = 225ml liquid butter. The solubility of the thc in butter wouldn't be identical to sesame oil, it'd quite likely be higher (it tricky trying to find actual lab measurements for that). But what that implies is you could potentially dissolve all of the thc present in 10oz of weed into a single stick o' butter. It's important that the mix is fully heated to decarboxylation temperatures for longer enough not just to convert it to the more active form but the decarboxylated version will tend to be more oil soluble. Heat also drives the diffusion process from the bud to the oil/butter; it's part of the formula for diffusion rate calculation.
There are a couple of problems with using tiny amounts of butter or oil to make super concentrate though. Say if your weed is 28%.... the butter won't absorb anymore from the weed once it's own thc content matches that of the weed. 78g of thc into 250g of butter = ~31% thc in the butter. So whether it's butter or oil you'd need to add some more. The weed will also tend to blot all that up quite a lot as well, the butter stick.
One trick used in labs that I've not seen anyone try is in a lab they definitely normally wouldn't try extracting something right at it's maximum solubility and/or all in one go, because the dissolution / diffusion rate starts to slow way down. The standard lab approach would be to use about three times more solvent than necessary, or a bit more, and to then rinse whatever it is three separate times with about a third of it each go.
E.g. using the butter example, to cook it in a couple of sticks of butter, drain / squeeze it off, add another couple of sticks, drain / squeeze, add the last couple, drain / squeeze. That way you maintain a high diffusion ratio between the solvent (the oil or butter) and the solute, the thc in this case. That could be a bit messy and it may be easier just to do it all in one go. You'd also need to be careful not to overly heat it during the second and third stages, just warm it until it's liquid.
Using the standard guide of about half an ounce to a pound of butter, it'd require around 20lbs of butter to do one plant (~10oz dry bud); ~9kg of butter, 8L, ~36 sticks.
I thought I'd go through the numbers on that because some people go with the quarter, half ounce to a pound ratio but I've seen others mentioning they do six or so ounces (168g) to four or so lbs butter (about 7 sticks of butter at 250g a stick), so an ounce per stick. In theory.... using that sesame seed solubility figure above, you could go all the way to 10oz per stick of butter (assuming the solubility is the same) but one stick of butter is unlikely to actually whet out 10z of dry bud on it's own.
I'm definitely not suggesting people should do that because it'll end up insanely strong but it could be useful if you're trying to process bigger batches and concerned about having massive amounts of liquid butter partying around. You can always dilute it down in an actual recipe right, but you can't make it stronger again so easily. And the best place to store it would be air tight in a freezer, so compact would be nice.