What Is Extraction? – the Theory and How It’s Applied to Food

After writing about making your own extracts (vanilla & orange/lemon) and how to use these it is time we start talking more hard core science. What is extraction from a chemists perspective? And is there something we can calculate here?

Making you interested in food science is one of the my goals of my blog. That’s why I try to write posts in different levels, hence the previous posts on making and using extracts. This more in-depth post is for those who are ready for a next step in food science!

When do we use extraction?

Imagine you have an orange and you’d like to use the flavour components in the zest of this orange. However, you don’t want to use them today only. Instead, you’d like to use some now, and the rest a couple of weeks later. That orange won’t keep so long though. You could buy a new orange every time you need it, but you can also try to take out the flavour components and store these. Whereas an orange might spoil, those ‘extracted’ flavour components won’t.

This is when you would use extraction. Tea is another example of extraction. You extract the flavours and other tea components from the tea leaves when making tea. You don’t drink the actual tea leaves themselves, but still taste the tea!

making lemon and orange zest extract, making the zest
Making lemon & orange zest extract.

What is extraction?

Extraction is an example of a separation technique. It can split two components, separate them apart.

Imagine you have a mixture of components A and B (for example orange zest (A) with flavours inside (B), or a vanilla bean pod (B) with vanilla flavour (A)). We’d like to take B out of A. However, A and B are very well mixed, we can’t wait for B to settle out, or wait for A to evaporate. In extraction B is removed from A by adding a third compent C (the rum in the case of orange and vanilla extracts). This third component will ‘pull’ B out of A.

The reason this works is that B dissolves (better) in C whereas it might not dissolve in A or a lot worse. Let’s apply this to the orange again: the flavour molecules will dissolve well in rum (alcohol) in which the orange zest is layed. As a result, they will sit in the rum and leave the zest itself.

When to use extraction?

There are a lot of different separation techniques. Some techniques (such as distillation) use heat to separate two components. Because of the heat, one of the components will evaporate, whereas the other one will remain behind.

However, if your components are sensitive to heat, distillation is not suitable. In a lot of those cases extraction is a good alternative. Since a lot of food components (especially flavours) are heat sensitive, extraction is used quite often in the food industry. In order for extraction to work, you should have a component available that is good in ‘pulling’ out the molecules you’re looking for.

Types of extraction

There are different types of extraction, the two main ones being liquid-liquid and liquid-solid extraction.

In liquid-liquid extraction the component you want to transfer (called the solute here and called B in the previous explanation) sits in a liquid (A). B has to be extracted into another liquid (C). During extraction, the solute travels from liquid to liquid. What’s very important in liquid-liquid extraction is that the two liquids do not dissolve in one another. If the liquids would dissolve in one another, they won’t be able to split again. In other words, you’ll end up with a mixture of 3 components instead of separating them further.

As you might guess by the name, in liquid-solid extraction the solute will have to travel from a solid into a liquid (or vice versa). Both processes are used in food, but we will focus on liquid-liquid extraction since it can be simplified a little more easily.

Liquid-liquid extraction theory

Liquid-liquid extraction isn’t just used in food. It is a very big topic within analytical chemistry. Analytical chemists often use extraction to isolate or concentration a component so it’s easier to analyze by them. There’s quite a lot of theory available on liquid-liquid extraction so we’ll dive into the topic a little deeper.

It is easiest to explain extraction in the most simple system possible, again using the same coding and the image shown below:

The form you have selected does not exist.

  • Liquid A
  • Liquid B (does not dissolve in A and will not mix)
  • Solute S (stars in the images below)
concept of extraction demonstrated in simplified illustration

Partition coefficient

As you can see in the image above, that extraction process wasn’t very efficient. Only half of the stars actually moved from A to B! We would prefer more than half to move though.

The amount of solute S which will move to the other phase can be described using the partition coefficent (K). K describes the ratio of the concentration of S in A vs. that in B at the end of extraction:

K ≈ [Concentration S in B] / [Concentration S in A] or [S]B / [S]A

In the example shown above the concentration of S is equal in both A and B at the end of the extraction process. That results in a K value of 1. In the example below you can see different values of K represented.

extraction and the partition coefficient explained

Since we want to extract as much S from A as possible, we’re looking for a B which has a very high K value when extracting from A. Each combination of solvents and solute has a different K and is influenced by all three components. Generally, the more alike A and B are, the closer the value of K to one since the solute won’t see the difference.

Partition coefficient and time

The partition coefficient describes a so-called ‘equilibrium’. In other words, this is the final state the mixture will arrive at over time. However, it might take a while to get there.

Speeding up extraction by shaking for instance doesn’t affect the partition coefficient. In the end the ratio of the concentrations will be the same.

Partition coefficient and pH

That said, the partition coefficient can be influenced by other factors, such as the pH-value (acidity). At a different pH the solute might prefer to sit in another solvent.

Choosing solvents in liquid extraction

In order for liquid-liquid extraction to succeed, it is important the solvents are chosen well. As we discussed before, the solvents A and B shouldn’t mix well, nor dissolve in one another. Instead, they should separate easily.

Also, you need to make sure the K value for your process is suitable. Make sure the solute actually prefers to sit in the solvent you’re using to extract it with.

Solid-liquid extraction

For solid liquid extraction the principles of extraction are the same. The solute you’re trying to extract, will prefer to sit in the other component. However, in this case you cannot shake the two components, they will not mix.

So, instead of shaking you will generally try to cut the solid phase into smaller pieces. The smaller surfaces result in more surface area over which the solute can travel.

Extraction & Food

In analytical chemistry a chemist might be looking for a very specific molecule to extract. By choosing the appropriate solvents, concentrations and time they will be able to extract the molecule. This might take several extractions in a row. If you take out 60% with each extraction, you will keep on extracting molecules, but it will take a while before you’ve extracted 95%.

In food things are even more complicated. There are often a lot of different molecules you’d like to extract (think of tea and vanilla for instance). Each of these will have different solubilities in the liquids and solids you’re using. Whereas one component might be extract very easily, in others, this may take a lot longer or might even never fully extract.

Another consideration is the relatively limited choice of solvents to extract with. Solvents tend to have to be food-grade, that is, fit for human consumption. This limits the wide pantry that analytical chemists have to choose from.

Applying your knowledge

Now that you know what extraction is, you can start extracting yourselves! Give making vanilla extract or orange zest extract a try. Wondering why orange zest and not orange juice?

Good luck! And, if there are any questions, let me know :-)!

The form you have selected does not exist.


  1. I have never espected the extraction in chemistry but because of my reseach i have see today

  2. I have read a lot of your blogs and as a fellow food scientist, I must say they are pretty informative and can be useful even to someone who doesn’t come from a science background! Keep it up 🙂

  3. I’ve seen lots of people claiming that vanilla beans soaked in vodka and similar methods shouldn’t be referred to as “extracts”, as technically they are only infusions. Is there a technical difference?

    • Hi Alex,

      I’m not familiar with the nuances between the definitions for infusions and extracts. As is often the case, they aren’t always used consistently (based on my quick search). If you soak vanilla beans in liquors, you’re making a vanilla extract, though maybe some call it infusion. In both cases the vanilla flavor will transfer over to the liquid, making a liquid with a strong vanilla flavor.

      From a legislatory aspect, soaking vanilla beans in alcohol does make an extract (at least in the US, definitions may differ between countries!).

      Would that make sense?

      • Sounds about right. Most nerdy food discussions around DIY vanilla circle around to Stella Parks’ article on Serious Eats, which vehemently argues that homemade vanilla can’t be called ‘extract’. Seems to be mostly an argument of semantics. Or for those who think homemade vanilla isn’t good, though oddly it seems no one has actually done any side-by-side taste testing to see if that’s actually true.
        I’d say that ‘extract’ is better used as a verb than as a noun. But we can thank the FDA for muddying those waters 😛

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.