One will first require to defat the tobacco. Acidification to render the nicotine in salt form, followed by several extractions with dichlor or chloroform, TCE or whatever similar chlorinated hydrocarbon floats your boat. DCM is nice for the low BP.
Plant parts often have a ton of waxes, fats, etc. in leaves, stems, seeds, rootbarks are sometimes extractable, or dried barks, with straight-to-base techniques, but aerial portions of most plants you need to defat first.
If you don't, you can and likely will end up with an unholy emulsion and a load of saponified shite to deal with. Which trust me, you really would rather never have to deal with either.
Don't forget, if you make a total alkaloidal isolate from tobacco, there are things like beta-carbolines (harmala alkaloids in fact, the MAOIs we know and love, harmine for example) in there, as well as, aside from the main pyridine alkaloid, nicotine itself, smaller amounts of anatabine, cotinine, desmethylnicotine and anabasine. Also depending on the tobacco species, with tree tobacco, Nicotiana glauca, IIRC containing significantly more anabasine.
Furthermore, I have a conjecture to offer, based on my experience with both smoking tobacco, and using a nicotine-only fag juice.
The fag juice is based on a propylene glycol base (I do not allow glycerin to be present in mine, nor will I use off the shelf ones containing it, because of the dehydration of glycerin to acrolein, a most unpleasant substance, which has the nature of a tear-gas, only far more toxic than the military or filth-issue tear agents, but a noxious-smelling, noxious tasting, highly toxic, violently irritant unsaturated aldehyde, that can be formed by dehydration of glycerin. Indeed that, in my first, and up until the present day, only deliberate act aimed at it's synthetic preparation.
In that case, I employed the sodium bisulfate from my first ever kiddie's chemistry set plus some glycerin from a local pharmacy; can't remember exact age, but under 10, running it on a test tube scale, purification via distillation from tube to tube, after mixing bisulfate and glycerine, heating, and collecting the distillate, redistillation under welding argon to prevent oxidation during the distillation, and more or less, doing it because it was a novel experiment, and those in the booklet that came with the chem set, and it's highly limited resources, were for the most part boring, like watching cobalt chloride and CuSO4 change color between hydrate and anhydrous forms. Not really fodder for my young and enquiring mind, and besides, nothing I hadn't already known and in the copper sulfate case already done with a blowtorch (the set was actually my first supply of cobalt resources, but once you've seen one coloured crystal hydrate turn to another colour with a blowtorch, and you know whats going to happen, what's the point? to me, not very much, so I found something else to do, that was more intriguing than the proposed experiments the set had guides to.
Make, distill and bioassay dilute concentrations of atmospherically dispersed lachrymatory agents, extremely dilute in the case of acrolein knowing the toxicity. Not on anyone else, or animals, mind you. I'm no animal abuser, never have been, never will be, and neither do I tolerate it in others if I can get within range to act upon such an individual.
All went swimmingly, and choking, spluttering, coughing, eye-streamingly once a sample was tested in an enclosed space and allowed to vaporise.
However not knowing any better, and considering the small quantity prepared, once done with it, and I hang my head in shame at this, but it went down the kitchen sink. And flushed with a lot of water. Unfortunately, and to teach me a lesson for dumping it in such an irresponsible manner, I'd turned the tap the wrong way, and flooded the U-bend with steaming hot water. Whilst standing at the sink.
You can guess the rest, after the entire remainder of the batch was exposed to hot water causing the sink plug hole to vomit forth the most noxious miasma back up into the kitchen, and right back at me. Pretty funny looking back at some of the crap I got up to as a wee'un, but I wasn't nearly as amused at the time.
Anyhow, amusement for your consumption aside, I use propylene glycol, commercial flavourings and nicotine dissolved in PG to make a calibrated volumetric dilution.
The nicotine, is not a tobacco extract of whole composition, and it has quite different properties when administered in an e-fag or transdermally, than does smoking tobacco.
The e-fag sourced nicotine alone, is to me, pretty devoid of compulsivity and ultra-notorious-grade addictive potential. I've been a smoker since age 9, of tobacco fags, long before e-fags were all the rage, although the nicotine alone has let me cut down to the point where I only enjoy occasional cigars, particularly as the icing on an opioid injectable-shaped 'cake', so to speak. It just fits, smoking tobacco. the nicotine from an E-fag lacks this particular reinforcing, satisfactory pleasure enhancement when delivered as an inhaled dose via the E-fag (obviously transdermal is a bit slower, so was aiming for rapid delivery to compare.
Tobacco does, it really makes things a lot more satisfying, addictive, but nicotine, without tobacco constituents, does not, in me, act thusly in the same fashion.
I have hunches, but the primary one, is, knowing the presence of harmala type beta-carboline alkaloids with MAOI properties, that these, in tobacco smoke, dramatically increase both addictive tendencies of the product used without opioids in context or at least, when using only at the doses I must maintain to remain free of withdrawal and to have the chronic pain in my knee joint, patellar tendon, and the trochanteric bursitis I have bilaterally not scream in red hot howling agony, such as nasal lower dose oxy, or plugged morphine rather than IM or IV or subcut morphine.
In short, I believe the harmala alkaloids present, are modulating the effect of the delivered nicotine when tobacco smoke is intaken, and doing so in such a manner as to render the end result far more compulsive and addictive.
So, I postulate, and it's only, I should make clear, a theory of mine as of the present, as I have yet to prepare a beta-carboline MAOI-spiked sample of fag juice nicotine and compare it for addictive, reinforcing, compulsive tendencies, both whilst at baseline, and after the administration of intravenous doses, or IM doses if the vein-gods won't smile on the endeavour, to then test PG/flavouring/nicotine spiked with known concentrations of harmine, harmaline and tetrahydroharmine alone and in combinations with one another; that the result of using tobacco as a source for extracted nicotine, would likely pull along with it, quantities of these MAOI beta-carboline harmala alkaloids. And make the fag juice more addictive than it otherwise would be due to their presence.
I feel it would be irresponsible if I were not to share this hypothesis, despite not yet having performed the bioassays on myself, since I'll have to either perform a Pictet-Spengler, on appropriate tryptophan derivatives to prepare the harmala alkaloids in single pure forms, or, to extract Banisteriopsis caapi liana, or syrian rue seed and remove the likes of vasicine, vasicinone, peganine and what have you, for the reason that if this were done, using tobacco to make fag liquid, it could end up being just as addictive, as cigarette smoking is and in a sly, underhanded, unexpected guise.
At this, I will leave you to draw your own conclusions from the reported results of the bioassay of tobacco smoking vs delivery of inhaled nicotine, both alone, and with the highly divergent effects upon opioid bolus dosing via injection type routes. VERY, very different animals.