I'm posting recent results following a different procedure in order to convert salts -- here, fumarate salts -- to their freebase form.

The experiment was driven by an increasing distaste for the standard procedure using Na2CO3 -- the freebase is hard to gather together in one piece, there is usually an excess of sodium carbonate which has to be dealt with (if one wants to), and most importantly for me, as I wrote out the symbolic conversion I simply couldn't account for the whereabouts of some components of the reaction, and the whole thing seemed inefficient.
The following theory was postulated: one could bypass the unwanted byproducts of the standard reaction [using Na2CO3] by simply using NaOH, thereby producing sodium fumarate and water as the sole products (in an ideal situation). It was speculated that the absence of anecdotal reports of similar experiments using NaOH instead of Na2CO3 was maybe most likely due to a notion that using NaOH is overkill, and since the Na2CO3 solution will generally sit at a nice pH of ~11.7(+/-), there isn't any need to go stronger. Nothing seemed inherently wrong with using NaOH, though, so the new procedure was tested :


Basically,

1. Dissolve salts in H2O, [use minimal but I don't think it really matters]. I used heated water
2. Prepare NaOH solution.
-- I began with 1N NaOH solution, (pH ~12.2), but after initial testing it seemed like more would be better. An additional ~0.9g was added to the ~20mL 1N NaOH solution, so approximately this is equivalent to 1.75g NaOH in 20 mL solution (let's say 18.25g H2O, but doesn't really matter)
3. Add dropwise the NaOH solution to the dissolved salts.
-- This process has its benefits: (Pictures below)
The solution will start yellow. Adding the NaOH produces opaque white clouds initially, and upon adding more the solution becomes entirely white and somewhat opaque. The freebase will begin to surface and coagulate. I screwed the cap onto the vial and shook for several seconds, and immediately all of the freebase coalesced into a single mass, floating on top of the surface. Excess NaOH was added to help this. So the benefits are, visually, you know how much is necessary to add but as a matter of fact excess can be used with hardly any impurities binding alongside (such as the sodium carbonate / fumarate - freebase paste). The only product should be the sodium fumarate, which will stay in solution. Shake or swirl until all the freebase has combined.
4. Remove the freebase chunk from solution, [I did not do this at the time, but looking back..] rinse with H2O (when the water evaporates, NaOH will get left behind. The hexane/heptane pulls shouldn't collect this, but it's better to remove anyway, and it's quick to do), and dry (dries extremely quickly)
5. Make into finer pieces, and pull with heated hexane / heptane. You can see that very minimal residue is left from the freebase chunk -- many of what is left behind is freebase that wasn't gathered all the way. Some is undoubtedly impurities.
NOTE: some freebase will stay behind in the NaOH solution. Currently seeing just how much that is, but on the whole, the process seems entirely efficient.

Attached at the end is an example of a heptane pull from a previous batch. Seems like the process changes the crystal habit, because the crystals are unusual for heptane.
For another post perhaps.

(Fun fact: Slow-cooling this freebase crystallization [attributed to the attached pictures] by 5 deg C / 20min beginning from 120C, and 65C is a changing point, where the majority of the excess freebase from the supersaturation comes out of solution and begins to take crystal from. It's interesting. If anyone wants to try, do a repeated cool-heat from (70C --> 55C --)n-->, it should be a continuous crystallization-dissolution process with slow crystals formed during the process...)

Thoughts?