I don't normally make a lot of confections but decided recently to try a few new things.

Yesterday I used this recipe for Turkish delight which I originally found on the Hydrocolloid Recipe Collection. It uses agar as the gelling agent and appears relatively easy to prepare (that's why I tried it). Basically you just dissolve some agar in warm water, orange juice, and lemon juice, add sugar, simmer it, then cool and refrigerate to set.

Unfortunately, what I ended up with after refrigerating overnight was closer to the consistency of jam or marmalade than a Turkish delight. It was far, far from delightful.

When I prepared this I didn't actually follow the source URL and noticed a few differences that may account for the problem (but I'm not sure, that's why I'm asking):

  • The original recipe calls for 1/2 teaspoon of agar; the one in the collection specifies exactly 1.2 g.
  • The original recipe specifically says to cover with a cloth; the collection recipe is not so specific, it just says to cover (I used plastic wrap).
  • The original recipe says to refrigerate in an ice-water bath; the collection recipe does not (I just put the dish in the fridge).

Other issues that may have been causes:

  • It turned out that my scale wasn't precise enough to get exactly 1.2 g. The amount I used could have been anywhere between 1 g and 2 g. In retrospect, I wish I had checked the original recipe and simply used the 1/2 teaspoon that it called for.

  • I got the agar from an Asian grocery store, and the packaging clearly said "agar agar", but I did notice that it had two ingredients, the second one (after agar) being vanilla. I'm not sure whether this is normal or whether it might have been diluting the agar mix.

  • I did notice that it did not seem to be gelling at all while it was coming down to room temperature. It was basically a soup until it had been refrigerated.

Can anyone identify the most likely reason why this didn't work out? Did I use too little agar? Too much? Could some of the differences between the two recipe versions be important? Could it be something I haven't thought of yet?

And equally importantly for the purposes of this question, can I fix it or am I going to have to throw it out?

  • When dealing with hydrocolloids, always always always go by weight. The amounts used are very specific, and volume measurements aren't close enough.
    – daniel
    Oct 8, 2010 at 19:13
  • @roux: In general I would agree. In this case (given that I couldn't get the weight completely accurate) I'd probably have been better off using the volume measurement, which the rest of the recipe was evidently built around. In any case, going by mass probably wouldn't have helped me if, as Michael seems to have confirmed, the agar I was using wasn't pure enough to begin with; I would have been wrong no matter what I did.
    – Aaronut
    Oct 8, 2010 at 19:44
  • Yeah, purity was probably the issue. Honestly I'm kind of surprised that T&T didn't have pure... they have everything! (Man, do I miss shopping there when I worked for Thuet... the fish guys would fall all over themselves for us!). I have had very little luck sourcing molecular supplies in Toronto.. have you tried Bakersource in Oakville? They may carry stuff.
    – daniel
    Oct 8, 2010 at 20:40
  • @roux: They did have these gigantic solid... things that they referred to as "agar" also, and there seemed to be different colours (or maybe flavours?). That was the only other thing I saw. At least I found it; try looking for carrageenan or xanthan gum here. Try even looking for a sub-gram scale! (Everyone says Amazon - too bad they don't ship to Canada.)
    – Aaronut
    Oct 8, 2010 at 22:39
  • 1
    Sub gram scales... try, I'm not kidding, some of the head shops on Yonge. I think the big solid things are basically candies; agar is used for candies all over the eastern hemisphere. You can order (expensively) the entire Texturas line from the Adrias; there is a distributor in Toronto but I forget the name. albertyferranadria.com has a list of suppliers.
    – daniel
    Oct 8, 2010 at 23:05

3 Answers 3


I've worked with agar a lot, and made a few things from the HRC. Of all of the things you've identified as possible problems, I think the biggest likely ones are the quantity of agar, and the agar powder itself. Everything else about the cooling issue is normal. Agar powders do vary. You want to find one that is pure, unadulterated agar, not one that is setup to be a pre-prepared dessert as those are weaker.

If just for a laugh you want to try to save this batch, heat it back up until the agar fully melts, shear in a second amount equal to what you did last night, making sure it fully dissolves, then let it set again. I'm pretty sure that will work.

You can also pick up a sub-gram scale on Amazon for around $20. Thanks, drug dealers! They are very helpful when making small batches of "molecular" recipes.

  • I know I definitely need a better scale; I actually thought that mine was sub-gram as it cost about $30. It also seems harder than I expected to find pure agar; even in the gigantic T&T store I only managed to spy this one at the last minute on one of the "South East Asia" shelves. I hoped that it was similar, but now that I think about it, it did have a distinctly sweet aroma and could easily have been nearly half vanilla. The net weight on the package is 7 g; perhaps I'll just bomb the rest of the package in there and see what comes out.
    – Aaronut
    Oct 8, 2010 at 15:14
  • One other question: The preparation steps include boiling and simmering for 20 minutes, is that necessary to do again in order to activate the agar? In the HRC it lists a 90° C hydration temperature but I assume the 20 minute simmer time is simply to reduce the mixture; can I just bring the melted gel to a boil briefly after adding more agar, and then set it?
    – Aaronut
    Oct 8, 2010 at 15:47
  • I ended up doing exactly this - tossing the entire rest of the package into the mix. It actually came out OK, although the nuts were a little soft after having been boiled. I think the recipe itself needs a few tweaks, but at least I've got the general idea now.
    – Aaronut
    Oct 9, 2010 at 2:55
  • @aaronut sorry, I didn't see your followup question until tonight. I've never boiled agar for anything like 20 minutes. As long as it comes up to the hydration temperature and then gets sheared thoroughly, you should be fine. Oct 9, 2010 at 5:21

Well, from what I know, even though your measurements were not accurate or the agar wasn't a hundred percent pure, it should not affect your mixture in such a way . The reason why your turkish delights turned wrong is more likely that you used fruit juice. When using agar agar to make candy or firmer gel bases, it is not recommended to use fruit juice since it contains acids wich makes the agar unable to set properly.

If you want to make turkish delights with agar, I believe you are better off using flavoring and coloring. I would even say you should use these in any recipe wich involves agar agar,unless otherwise specified. You should also check out agar agar candy,wich is in my belief pretty similar to Turkish delight. Since lokums are more jelly like, you should either not sun dry them or maybe just let them dry a little after they set (that doesn't take a lot of time,only let them cool down and they'll be gel like) to get the consistecy you want.

  • You might add that Arrowroot is a possible agar stand-in that stands up to acidic environments, or specify an alternative gelling agent.
    – mfg
    Mar 7, 2012 at 19:43

According to Wikipedia, Turkish delight base is not agar but starch and sugar.

  • 1
    Yes, this recipe is meant to be an analogue, not the traditional recipe. Oct 8, 2010 at 15:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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