I'm trying to get as pure citric acid as possible from a mandarin. It's not because I desperately need citric acid, but because I want to achieve this.

citric acid filtration

I started by grating the mandarin and filtering the juice. After filtering overnight, I was left with a slightly yellow liquid. I heated it while avoiding boiling and I reduced the volume to 50%. When the volume was reduced, the solution started to become saturated and crystals appeared. At this point, I stopped.

Now, I'm looking for a way to remove sugar from the solution, as this is the most annoying part. It's sticky.

When crystals appeared while I was boiling the solution, I came to think, that perhaps that could allow to separate sugar. Citric acid has solubility in water 1477.6 g/L (20 °C), while sugar has 2000 g/L (25 °C). Wikipedia unfortunately doesn't provide values for the same temperature.

From the above, it seems that I can have more sugar dissolved than citric acid. I think that means that citric acid will crystallise first. Does it work like this? Are the crystals I've produced the citric acid crystals?

Is there another way to separate the compounds? Does something react with sugar to produce a precipitate that can be filtered?

filtrated and concentrated citric acid solution

  • 2
    $\begingroup$ This would be very complicated as mandarin juice would contain thousands of compounds. Do you mind if the citric acid is potentially contaminated with a lot of those compounds- mainly organic compounds? $\endgroup$ Sep 2, 2014 at 10:06
  • $\begingroup$ No I don't. But I want it to be acid, rather than sweet. As I said, this is an experiment for the sake of experiment. I will do as good as I can, but there's no cost for failure. $\endgroup$ Sep 2, 2014 at 10:11
  • $\begingroup$ I don't think citric acid not sugar (which?) would be a major component in this juice that would single out from other components. Anyway, crystallisation has very limited value in separating such a diverse mixture. You can purify sugar with crystallisation, but cannot remove sugar from something else this way. $\endgroup$
    – Greg
    Sep 2, 2014 at 10:13
  • 2
    $\begingroup$ I would guess you need a lot of mandarins to isolate a tiny bit of citric acid. You could try fermenting the sugar to alcohol, that will take some time, and I am not sure how well it will work. Of course you won't get rid of any other contaminating components. $\endgroup$ Sep 2, 2014 at 10:56
  • 2
    $\begingroup$ First for the amount of citric acid, let's assume 0.01mol/L (answers.com/Q/…), ~ 2g /L. Sugars wildly guessing 100g/L. Without mentioning other compounds, citric acid is better considered impurity, rather than constituent. If I would have to do it, I would probably turn the citric acid into calcium citrate, separate (again very impure) and turn back to free acid. $\endgroup$
    – ssavec
    Sep 3, 2014 at 9:21

1 Answer 1


Citric acid is relatively soluble in ethanol (62g/100g) whereas sucrose is relatively insoluble in ethanol (1g/170ml). There are many different kinds of sugars and sucrose is probably not the primary one in this extract; branched sugars... oligosaccharides (which have generally higher solubilities than sucrose) are likely, and I'm sure there are a few other organic acids involved with a crude, fruit extract, but the different solubility characteristics should be relatively helpful for extracting citric acid from sugars and some other organic acids.

So, I would try extracting the citric acid from the dried fruit or dried water extraction with 99%+ ethanol. I would then allow the alcohol to evaporate to help concentrate the acid.

Dilute citric acid (from a concentrated extract) can be precipitated with calcium hydroxide as practically insoluble calcium citrate, which is then reacted with sulfuric acid to form (food grade) citric acid and calcium sulfate (gypsum) as a recoverable by-product.

To really purify the stuff, an ion exchange resin or some other chromatographic method would be the last step.


Not the answer you're looking for? Browse other questions tagged or ask your own question.