Clarification of stock and other liquids

In a comment to the last post, Chad asked how the clarification with laboratory glass ware works. Here’s how. Basically it’s a filtration. But if you would use a normal filter paper (such as a coffee filter) and let gravity pull the liquid through the filter, it would take ages. By applying a vacuum to the back side of the filter, the stock is sucked through (or pushed if you like by the atmospheric pressure). The are several possible sources of vacuum. The simplest and cheapest is a water aspirator or a handpump. More expensive solutions include a membrane pump or an oil pump. The particles you want to remove are from 0.0001 mm and upwards to > 1 mm. The best thing would be to first pass the stock through a cheese cloth or a muslin, followed by one or more filtrations using filter paper. This would gradually yield a perfectly clear solution. Pictures of a Büchner funnel, Erlenmeyer flask and a water aspirator can be found on the tools page of Khymos. Pictures of a complete setup can be found by googling. If doing this in a kitchen, you would want to have an Erlenmeyer flask of at least 2-3 L as this is where the clearified stock is collected. The Büchner funnel should preferably have a diameter of 12 cm or more.


The fascinating thing about a filtration like this is that you can also remove color. At the EuroFoodChem XIV conference I was told by Jorge Ruiz of Lamaragaritaseagita that you can make perfectly clear tomato juice by succesive filtrations, starting with a coarse filter and moving to finer filters. All in all, 3-5 filtrations should be sufficient.


  1. My very first taste of molecular gastronomy was in 2003 at Clio in Boston: a crystal clear tomato juice served on a Martini glass and decorated with an olive.

  2. Thanks for that quick diagram and tutorial. The idea of clarification through physical means has intrigued me for awhile from a distance. Although we have not tried it, I would love to have a similar set-up in the kitchen. The process of forcing the liquid through means of pressure reminds me of an aquarium filter that relies on pressure and reverse osmosis. What kind of filters should be used towards the end, and from what kind of source would they be available (medical supply? lab supply?)
    Jorge Ruiz has emailed me in the past about transglutanimase and meat curing processes. His knowledge on the subject is incredible. It’s really great that you guys were able to meet and share information.
    I will be attending the StarChefs congress in less than 2 weeks in New York and I hope to make some great contacts and meet others with the same interest once we are there. The anxiety and anticipation are building everyday.

  3. the filter membranes to use must be hydrophilic, like cellulose acetate. These are available from lab suppliers, but at a price. Note that filtering with a vacuum line is common practice in many labs, but the amount of suspended solids or particulates present in typical sample is usually much less than that be found in tomato juice, for example. I would not recommend this kind of filtration for such a task, much better would be a depth filter, with say cotton wool, to remove gross solids.

    It is possible to remove even colloidal sized particles, such as the caesine micelles in milk with membrane filters. I managed to produce a few drops of crystal clear liquid which tasted exactly like milk! with a hand syringe and 0.1µm membrane.

  4. ceramic membranes (tubular membranes or ceramic discs) as used for the clarification of apple juice, sterilization of milk, etc.. are easier to handle and easier to clean. They are available with porosities varying from a few micons down to a few nanometers.

  5. Thanks guys for your comments about myself (I’m turning red). I am pretty sure that my knowledge is very, very limited.

    It is true that the lab filter membranes are not cheap… and they are blocked with only a small amount of juice. It could result very interesting to perform first a clarification following the gelatine-freezing procedure (today in Harold McGee’s column at the NY times) and thereafter the cellulose acetate filtration.

    I haven’t tried with ceramic membranes. Pierre, could you tell us something about the price and the use. I am really very ineterested.

    Martin, was really great to meet you and all the other people (Juan, Rachel, Erik… all these names that now have a face) in Paris. We had also time for having interesting meals and to taste Spanish food and French wines at Juan’s home (are you gonna post something about the molecular meal?)


  6. I tried the gelatine freeze filtering technique a while ago, inspired by a post by the blogger M who used it for TGRWT #3 to make a clear strawberry consommé (M in turn was inspired by Ideas in food and eGullet). In my experience it was extremely slow, and after a couple of days I just gave it up. The coffee filter gave a clear liquid to start with, but probably got clogged pretty fast. Any suggestions to what went wrong? Should I have used more gelatin (=> bigger particles, less clogging?) or less gelatin (=> less visous solution, easier filtration?). I can’t decide which is more important.

  7. I find Erlenmeyer flasks a pain to clean and too expensive (especially if they got an extra side arm and a precision ground neck). I usually use the “˜normal’ coffee filter method, after first separating the big particles by letting the stock settle (i.e. using gravity as a centrifuge) and cool down. After that I reheat the top and the carefully cleaned gelatin-laver (if any) for filtration. There are coffee filters with different mesh sizes (usually from about 4um (marked “˜strong [coffee]’) to 13um (“˜mild [coffee]’)). Furthermore using filters designed for professional coffee machines helps the clogging, because the filter(-paper) area is much greater (e.g. Bonamat).
    Have you tried gelatin assisted filtration, yet? It seems like a very clever idea, if the aroma molecules really do not bind or attach themselves to tightly to the gelatin mesh.

  8. Instead of the laboratory glass, you could use a water cooler bottle and apple cider jug. The necks of 2-liter bottles make good valves, and you can seal everything airtight with a glue gun.

  9. I have been playing a lot with gelatin filtration lately. I have made clear stocks of green olive, olive oil, gorgonzola dulce, lobster coral-eau de fleur oranger (flour and orange broth from Provence), among others. I use about 1% gelatin by weight when I do it. It does take time but I am really happy with the results that I have gotten. Oh yea watch out for putting too much starch if you flavor something that is starchy because it will hold onto all of the water molecules even when the ice crystals melt. good luck

  10. Is laboratory filter paper food-safe? It works great in a Buchner funnel setup to filter my cold-brewed coffee (I’ve experimented with it on samples that aren’t going to be consumed) but I doubt it’s safe to use with foodstuffs. If this is the case, does anyone know where food-grade filter paper can be obtained? I’ve tried coffee filters and I can’t pull much of a vacuum on them, which makes the process much more time-consuming. So I’d really like to find a filter paper that works and is safe.

  11. I can’t imagine why it shouldn’t be food safe apart from possible cross contamination if you take an opened box from a chemistry lab. To be sure you could perhaps wash them with water before use?

    But I would certainly make sure that I use a brand new Buchner funnel – not one from a lab. Such funnels can be notoriously difficult to clean since you don’t really see whats inside.

  12. Thanks for the reply! I had originally been afraid that maybe filter papers would leach metals or solvent residues or some such thing left over from the manufacturing process, but lately I was starting to come to the conclusion that they were probably pretty residue-free since people use them to filter out bacterial cultures and such. You wouldn’t want them to include anything that would affect the bacteria so it seems like there would be nothing toxic.

    Thanks again for your input. I feel better about using them now. I had a new unopened package that I was starting from, and a new unused funnel, so I think I am good to go. When funds allow I might get a nicer Coors funnel since the one I’m using has some unglazed parts.

Leave a Reply

Your email address will not be published. Required fields are marked *

The reCAPTCHA verification period has expired. Please reload the page.