successful method for purifying anthranilic acid(quick report)

Hey, this is a quick report on a successful attempt of mine to purify anthranilic acid made from phthalimide via the Hoffman rearrangement.
There's nothing really special about it, but I figured it can't hurt posting it in here, just in case someone would find it useful
---

Stuff you'll be needing:
1. Crude anthranilic acid
2. Soxhelt apparatus
3. n-Heptane(other alkanes might also work, although it seems reasonable to assume that those with higher b.p would work better[or at the very least faster]).

The process:
Fill the Soxhelt apparatus with the crude anthranilic acid, add some n-heptane and start refluxing. That's about it.

If larger crystals are desired, you can add a bit of methanol(after the extraction is over) and slowly boil it off. But keep in mind that some of the acid will react to form methyl anthranilate(evident by a grapy odor). Adding ether might be a better choice.
---

pictures & descriptions:
beginning: a brown powder(somewhat sticky)
a bit after the beginning:

end:

extracted product(after recryst with methanol):

slower evaporation of the methanol:

---

full disclosure:
a relatively pure product can be obtained straight from the formation of the acid via the Hoffman rearrangement, if done meticulously.

Unfortunately, stuff like traces of phthalic anhydride/acid in your phthalimide, or uncertainty in the concentration of your hypohalite solution, might make it a difficult goal to achieve.

[Edited on 7-7-2020 by B.D.E]

Quote: Originally posted by Tsjerk

Why would higher boiling alkanes work faster? The actual extraction is cold right? And how did you evaporate the heptane? I found it quit hard to get rid off in the past.

Quote: Originally posted by mackolol

...

Glad to hear you find it useful

As for your question, I just want to clarify that anthranilic acid is only very slightly soluble in heptane(roughly 1-2 grams per liter at ~80C). (this is also why a soxhlet extractor is an absolute must-have for this workup).

So as long as you don't mind having a powdered end-product, you can just filter the heptane out at rt, right after the extraction is complete, without a significant lost of yield.

[Edited on 8-7-2020 by B.D.E]

Quote: Originally posted by B.D.E

So as long as you don't mind having a powdered end-product, you can just filter the heptane out at rt, right after the extraction is complete, without a significant lost of yield. [Edited on 8-7-2020 by B.D.E]

Quote: Originally posted by mackolol

In case of anthranillic acid, I don't see any other point in having bigger crystals, than just aesthetic matter.

yes, that's about it.
I just wanted a trophy

And this is another reason to buy soxhlet exctractor, which I wanted to do for a long time but I'm always forgetting

Quote: Originally posted by B.D.E

[...] If larger crystals are desired, you can add a bit of methanol(after the extraction is over) and slowly boil it off. But keep in mind that some of the acid will react to form methyl anthranilate(evident by a grapy odor). Adding ether might be a better choice. [...]

Quote: Originally posted by Texium (zts16)

This is fantastic, B.D.E. That is the cleanest looking anthranilic acid I have ever seen, even compared to commercial samples. I will definitely be doing this with mine once I resume home chemistry. Edit: one thing that would be good to include is your percent recovery. If the process is particularly lossy, it may not be worth using in most cases, though if the losses are minimal then this would be really great. [Edited on 2-8-2021 by Texium (zts16)]

I was doing another anthranilic acid synthesis today, getting enough so that I can get enough to put in my soxhlet extractor whenever I get the time to try this out, and after I dropped the pH to 3.5 and the AA precipitated out, I couldn't believe the color, it was the best color I've gotten thus far! Just a very light beige (I think I could get by calling it "off white" even).


I was thinking that maybe I wouldn't have to do the soxhlet extraction using heptane... Instead, I could just give it a quick recryst with some boiling water...


This stuff is so damn touchy How the hell does it get darker in the recryst??

Looks like I'll be trying this out after all. I might even skip the recryst on some of the runs and go right to the soxhlet trick detailed above.


P.S. Here's the other pictures I took, if anyones curious. And some other photos of the previous synthesis runs.

[Edited on 12-2-2021 by SuperOxide]

@B.D.E - Can you tell me how much crude anthranilic acid you processed and for how long you refluxed? I'm doing this right now, and I planned on sharing my results. On the run I'm doing right now, I'm processing 10g of anthranilic acid with about 120mL of heptane (probably not n-heptane, since it mostly distilled over at 91 °C, but it seems to work regardless).
I ran the soxhlet for probably about 7 hours straight, and while it's obvious that it worked by the amount of solids that were precipitating out in the bottom flask (once the heptane got saturated), it wasn't as much as I had hoped. Did you just let it run for a full 24 hours or something?

Edit: And I think that if the goal is purity, then a recryst after the soxhlet/heptane treatment is definitely important. The crystals were a light brown when they were in the soxhlet, and when they were precipitating out in the flask they were a weird looking yellow color:

Now they're pretty close to white (I think real chemists would call it "off white"):


Definitely an improvement :-)

[Edited on 8-3-2021 by SuperOxide]

Quote: Originally posted by Texium (zts16)

What was your percent recovery?

Just an update - I gave this a go again yesterday, and while it technically works, it takes a loooooooong... long time.. Honestly it's not worth it unless you can set it and forget it for like 24 or 48 hours, or you just need a small sample of a nice white crystalline anthranilic acid sample to test something out on.

Procedure

I added 20g of anthranilic acid (which was made via the Hofmann rearrangement of phthalimide, then recrystallized in boiling H₂O) to my 150mL soxhlet extractor, added the heptanes (extracted from Super Tech Starting Fluid, which distilled over at 91 °C - 93 °C, suggesting it's not n-heptane) and cranked up the temperature. I ran it for a little over 12 hours. I timed how long it took to cycle in the soxhlet extractor, and it was aprox 7 min and 20 sec, so that means it cycled approximately 100 times (give or take a couple considering as the anthranilic acid got extracted, there was more room for solvent in the soxhlet apparatus and less room for it in the flask)

20g of recrystallized anthranilic acid


Acid placed in the soxhlet extractor (in a coffee filter to prevent clogging, which was a problem before)



At the end of the 12 hours, there was a yellow solid that was forming in the boiling flask, which was sticking to the stir bar pretty strongly.

The light yellow solid that precipitated out and clung to the stir bar



Flask with yellow solid from a different angle (notice there seems to be quite a bit on the sides)



I took the remaining anthranilic acid out of the soxhlet apparatus, dried it and weighed it at 15.69g - meaning only 4.31g was extracted over 12 hours. Thats about 0.359g extracted every hour. Even if one were to assume that the extracted solid was pure enough to use (which it seems to need a recryst, so not likely), this is still a pitiful yield. Obviously this means the soxhlet is absolutely required if you want to use this route.

From here I kinda messed up a bit. I decided to just filter off the yellow solid and do a recryst with water, which was a bad idea. The crystals were also yellow, as opposed to white (or slightly off white) that I expected to see from my previous smaller scale shorter run (also mentioned in a previous post above).

Recrystallizing the yellow anthranilic acid in boiling H₂O seems to yield long needle like crystals, still light yellow in color



Filtered off (still slightly wet, but they looked the same when dried)



I did another recryst with MeOH which seems to have cleaned it up a bit more (though it's still in solution), but not as colorless as I was hoping.

Crystals of anthranilic acid starting to form in the MeOH as it cools


Same flask, but you can see some of the color seems to come off when the flask is tilted on its side and the solvent is allowed to settle to the side (gives me hope it may clean up once filtered and dtried)



If anyones interested, I may post some photos of the filtered/dried crystals if they're any lighter in color.


This doesn't mean that a white solid can't be attained, since on a smaller scale, I added MeOH directly to the heptane/anthranilic acid mixture, then heated it up with strong stirring and this yielded perfectly white crystals:
Crystals NOT from this run, just used as a comparison.


The fact that I was just messing up the recryst seems to be evident in that whenever I dissolve the anthranilic acid in MeOH, then pour it into a different container after heating it up, the dribble that runs down the outside always turns out pure white!



So basically I messed up the recryst on this longer run by not just adding the MeOH directly to the heptane/anthranilic acid solution. Somehow that yielded a much cleaner product - Perhaps the recryst from boiling H₂O caused some oxidation due to the higher temperature? Not sure about that.

Two things I may try next before running away with my tail between my legs:

1. Running the soxhlet extraction using hexane instead of heptane - Will this be better? I have no idea. But the boiling point is much lower than heptane (69 °C instead of ~93 °C). Perhaps lower BP = more cycles = more extraction? no idea.
   
2. Try a multi-solvent recryst (of the brown crystals right from the Hofmann rearrangement of phthalimide) with heptane and diethyl ether - Anthranilic acid is basically insoluble in heptanes, but very soluble in ether, so I'm just curious what will happen if I dissolve it in some ether and crash it out with heptane.
   

Question for B.D.E: How long did you let the soxhlet extractor run? What was your yield? How long was your vacation you took in the interim? (I'm assuming a month).

[Edited on 19-4-2021 by SuperOxide]

Quote: Originally posted by draculic acid69

Water seems to be a no go for this substance. Going off everyone's before and after pictures water seems to make the stuff worse not better.

Yep... I'm now convinced of that, 100%. I wish I could go back in time and slap myself before I made the stupid decision to use water (I tried to wash the product with ice cold water, since it's much less soluble in ice cold water - still didn't work well). That's 12 hours, wasted, lol.

I may actually just try the whole synthesis again, and instead of using water when I recryst the crude product, I may use MeOH. The crude product in my experience is almost white:


But when I use boiling water for the recryst, this is the same product:


So yeah, lets see if it keeps its color using MeOH as the solvent, keeping water out of it entirely.

Quote: Originally posted by draculic acid69

Just a quick question; was tap or distilled water used? Maybe that's a factor maybe not.anyone have experience doing this with distilled vs tap water my tap water has lots (LOTS) of crap in it. Maybe it's the water causing decomposition or oxidation maybe it's gunk in the water? I mean cl2, fluoride substances, flocculant, carbonates of mg+ca, other minerals,copper steel or plastic traces from piping etc are all possibilities.

Quote: Originally posted by draculic acid69

Last question:. Is this recrystallisation problem route specific or common to all anthranillic acid synthesis?

So apparently there are several threads dedicated to purifying anthranilic acid, and everyone reports crap forming with the product. I surmise, but don’t take my word for it, that the crap is caused by oxidation (of impurities in phthalamide?) caused by either the hypochlorite/hypobromite itself or, for those who use bleach, by the sodium chlorate produced when bleach decays.

From what I observed, the crap-like colouring develops as temperature rises. Temperature control of the initial reaction doesn’t amount to much, since it needs to be pushed to 80 °C anyway later to complete the process. I've tried to add various reducing agents after heating the solution, but it doesn’t really do much either. Adding reducing agents before heating is counterproductive since it will destroy bleach/hypobromite.

I (almost) successfully purified very dirty crystals this way: recrystallise from boiling water. You should obtain a very dark solution, and probably no crystal will form when cooled. This is due to a wrong pH value. Add hydrochloric acid until the pH is way more acid than anthranilic's pKa (say, 2): the crap will polymerise and fall out, while the anthranilic acid will remain dissolved (anilinium form). Leave overnight in the fridge. In the morning, you should come up with a light yellow solution over a thick layer of sticky tar deposited at the bottom of your beaker (clean with acetone). Decant the clear light orange liquid above, and cautiously add sodium hydroxide: nearly pure anthranilic acid will precipitate. Collect and dry.

[EDIT] I've tried this rearrangement using in-situ hypobromite, commercial bleach and home-made bleach from calcium hypochlorite.

[NOTE: don’t use calcium hypochlorite directly as a substitute for bleach: it will react with sodium hydroxide to form calcium hydroxide that will precipitate out, reduce the alkalinity of the mix and all sorts of problems. So if you want to use calcium hypochlorite as a source of ClO⁻ ions, prepare sodium hypochlorite from calcium hypochlorite in advance by pouring it into a COLD sodium hydroxide solution. Don’t dissolve sodium hydroxide directly into it, as this will cause the solution to heat up and bleach to degrade. Then filter.]

Nothing really works, they all produce the same crap. I also recrystallised the phthalimide I made in ethanol rather than using it straight away from the urea + phthalic anhydride reaction. It didn't change much. Maybe I’ll have to use recrystallised phthalimide plus in-situ made hypobromite, which I didn't do (I ran out of bromine). Another way to explore is to use N-bromosuccinimide instead of hypobromite. I read somewhere that it could be used for Hofman rearrangements, and probably wouldn't be as oxidising as bleach/hypobromite (but way more expensive).

Lastly, a picture from the recrystallisation.


[Edited on 3-8-2025 by Keras]

So, here is what I finally found on preparing pure (white) anthranilic acid:

• Use recrystallised phthalimide. 25 g of phthalimide made from phthalic anhydride and urea can be recrystallised in ~ 600 mL of standard 90°+ denatured alcohol. It crystallises in fluffy, white needles.

• Use bromine and make hydrobromite on the fly following the method in Vogel p. 898. Don’t let the temperature of the initial sodium hydroxide solution rise above 0°C. Don’t remove the ice bath until you have added the second sodium hydroxide solution. The temperature will rise to ~ 30 °C and then maxx out. When the temperature starts to fall down again, remove the ice bath and heat up to 80 °C. The solution should slightly change colour, from a straw like yellow to a golden one, but never turn darker than that.

• Precipitate the anthranilic acid as described in Vogel. The yield is somewhat impressive, but the powder is grey/yellow, somewhat dirty due to oxidation. Recrystallise in ~ 200 mL of boiling water, add a gram or so activated carbon, and after 10 min let it cool and add a 100 mg circa of sodium hydrosulfite (dithionite). The solution should be pale yellow at this point, and the anthranilic acid will separate in white crystals.

• Don’t – I repeat don’t – put the recrystallisation beaker in the fridge. Put it in an ice bath, and filter as soon as it is cold enough (5 °C). Letting the crystals bathe in the yellow solution will recolour them after a while.

• Dry if possible in a vacuum dessicator and store.

Picture taken during the recrystallisation.



Edit: Yield of slightly dry product > 20 g whereas Vogel states 14.

[Edited on 17-8-2025 by Keras]

Having had plenty of experience making anthranilic acid from methyl anthranilate, as the first part of IBX synthesis, I'd like to make an additional point regarding crystallisation from water.

The solubility of anthranilic acid in water increases sharply between 80 and 100 °C, as does the rate at which it oxidises. The least discoloured crystals I've ever got from the hydrolysis resulted from not letting the temperature exceed 80 °C while adding acetic acid - crystallisation was very rapid, no water had to be boiled off, and said crystals were pale beige. The photo shows what was recovered from the mixture on the pump after cooling overnight.

By contrast, heating the same solution to boiling required nearly two-thirds of it to be evaporated before any precipitate formed, and the solution itself was nearly black.

Quote: Originally posted by Lionel Spanner

The solubility of anthranilic acid in water increases sharply between 80 and 100 °C, as does the rate at which it oxidises.

I completely agree with that. The picture below shows a sample (660 mg) of my last batch. It is light beige. I generally use a mix of two methods to purify it, the first being standard recrystallisation in boiling water followed by addition of sodium hydrosulfite to reduce any oxydation product, the second being re-dissolution in a basic solution (basically adding sodium hydroxide), adding sodium hydrosulfite, filtering and re-precipitation using acid until the pH is around 4.

I've already observed what you describe, i.e. that following a 100 °C recrystallisation the solution gets dark and would not crystallise back. This is, in my experience, caused by a wrong pH. Adding a few drops of HCl immediately precipitates the acid, though not in the purest form.



[Edited on 23-8-2025 by Keras]