SamuZai
Darcy S. ONeil
Darcy S. ONeil

patreon


Real Vanilla Extract Cream Soda

Several people have asked about making cream soda using real vanilla, so I've looked at the numbers for everyone and here is how it works.

The first thing you need to know is that vanilla beans typically contain 1% to 2% vanillin (some can be 3%, but that is not typical; closer to 1% is). For experimental purposes, let's assume you have beans at 2% vanillin and get 100% yield (unlikely) from your extract. That means for every 100 grams of vanilla beans, you will yield 2 grams of vanillin, or one-tenth of the Cream Soda Formula, which is fine because not everyone wants 400L of cream soda. This recipe will yield 40L. Though the reason we work with such large volumes is that it allows us to be more accurate with our small measurements, as you will see.

Another important note is that commercial vanilla extract is actually a tincture (see my video on tinctures), which is 10 grams of vanilla per 100 ml which yields 0.2 grams of vanillin, and that is too low to work with when adding the other flavour compounds. We need to make an actual extract (see extract video).

A true vanilla extract is 100 grams of vanilla to 100 ml of 40% alcohol/vodka using the percolator method. Now, I use more alcohol because the vanilla tends to retain some of the alcohol, so I estimate 120 ml. You can buy ground vanilla bean powder but check with the supplier for vanillin content, this powder is often cast-off beans that didn't meet a quality standard so they are ground up and sold at a lower price. Once you have made the extract, you will use the full 100 mL of extract to make the cream soda formula, but the additional ingredients will be used at one-tenth the volume.

Vanilla Bean Cream Soda Formula

100 mL 1:1 Vanilla Extract
6 drops Ethyl Acetate (full concentration)
4 drops Aldehyde C16
2 drops Isoamyl Acetate
3 drops Lemon Oil
1 drop Citral
1 drop 50% Benzaldehyde

For beta-Ionone, we can't measure a tenth of a drop, so you will need to make a one-in-ten dilution (1 drop of Ionone in 9 drops of alcohol and then use 1 drop of the dilution) or skip it. And most of the measures above are slightly more than in the original formula because it is just hard to work with drops as there is some variance between drops. But they are still below the typical maximum usage. And again, you can always add a drop or two more if you want to experiment. There is lots of working room in the formula, though working with real vanilla could get expensive.

The 100 mL of cream soda flavouring will make 5 one-litre bottles of syrup. To make the syrup, you will use 20 mL of this cream soda flavouring in 1 litre of simple syrup. Then you would use 30 mL to 45 mL (1 oz to 1.5 oz) of the syrup per drink.

Important Notes

Depending on the type of vanilla bean you get, your vanilla flavour may not be intense enough. For example, if the beans you get have only 1% vanillin, in that case, it becomes more difficult to extract more vanillin. You could try extracting 200 grams of vanilla bean with 100 mL of alcohol but I suspect your yield will be poor and the costs will get more excessive. There is a limit to how much vanillin you can extract from a bean and keep the vanillin flavour at a workable level in a soda.

If you proceed with this, try to get the best quality bean you can and ask for the vanillin content, you really want a minimum of 2%.

You may find grinding the beans a bit tough as they usually have a moisture content of 25%, so they can form more of a paste when ground, which is hard to extract in a percolator. You can macerate the ground beans and filter them if it gets too thick. You will probably need to press the vanilla mass to get most of the liquid out. 

Just soaking the beans in alcohol will give you a low yield of vanillin, probably 50%. There is a reason commercial vanilla extract is actually a 1-in-10 tincture and not a true extract.

The last issue is your true yield. With all the steps (grinding, extracting, filtering, etc.) your actual yield will probably be around 80%, which means for 2% vanillin content beans you'll yield 1.6 g, which is about 30% lower than the formula, which may make a weaker vanilla flavour in the soda. But maybe the other compounds in the bean will make up the difference. Experiment and find out!

I hope that helps and if you have any questions, please post them below.

Comments

They can be one-dimensional compared to real vanilla, but is some cases like Cream Soda that can be plus as it works well with other flavours. And yes, a simple essence that you listed would work well in root beer, but you don't need 95% EtOH, 40% vodka will work fine as vanilla compounds have good solubility.

Darcy S. O'Neil

Hello Darcy! I saw online that ethyl vanillin and vanillin are too "one dimensional" because even artificial vanilla has other 60ish compounds. Is that why you add other things to the cream soda? Will an essence of vanillin and ethyl vanillin be enough in root beer and cola? If so, 3g of vanillin and 0.5g of ethyl vanillin in 100ml of 95% EtOH work as an essence? Thanks in advance!

Emmanuel Isunza

You can experiment with vanilla beans and such, and if cost isn't an issue, then just use as much vanilla as needed. However, it could get quite expensive. Dehydration won't help, the beans are mostly dry, the small amount of moisture in them would help extraction.

Darcy S. O'Neil

My intended use for a vanilla extract is to have it on hand to make a syrup with. I would think it would follow the same rules as using a ginger extract to make a ginger syrup, 1 oz. to one liter of simple. Are there any suggestion you'd make for that purpose? Would dehydrating the beans help?

Guy Jackson

You can, but you will need a lot of vanilla extract and you need to calculate how much alcohol is going in, you may push your soda above the 0.5% level turning it into an alcoholic beverage.

Darcy S. O'Neil

I'm looking to be making and selling my own root beer and cream soda and I want real vanilla in the products. So if it's not practical to buy vanilla extract because it's too weak to use easily in the flavoring, I'd use something more potent. As long as the math worked out I'd do it. Could one use vanilla extract and add it in when adding the flavor to the simple syrup?

Isaac Holzwarth

You can use it, but it will get quite expensive for Cream Soda.

Darcy S. O'Neil

I see that some of your suppliers have CO2 extracted vanilla "essential oil" (which isn't essential oil really). Would that be able to be used effectively for cream soda and root beer? I read that it has about a 6% vanillin content.

Isaac Holzwarth

Most C16 is safe to use but Perfumer's Apprentice is a good source (search for C-16), check out this video so you what can be used in food and why most suppliers don't say food use: https://youtu.be/FSL6pVrV8Pw

Darcy S. O'Neil

Do you have a recommendation for Aldehyde C16? Every source I find says not for internal/food use.

Brandon

For the Sigma Aldrich stuff, you'll unlikely be able to order from them, they don't sell to individuals, only companies. Important for people wanting to commercialize soda, but not for hobbiest. Try this link for ethyl acetate: https://creatingperfumes.com/products/ethyl-acetate

Darcy S. O'Neil

Looking at the creating perfumes shop and the shipping was 11 for a 22 $ order and giving this the only thing I needed so checking elsewhere. On the Sigma Aldrich site I see a few options and am unsure which is appropriate https://www.sigmaaldrich.com/DE/en/search/ethyl-acetate?focus=products&page=1&perpage=30&sort=relevance&term=ethyl%20acetate&type=product

Rob

Yes, that is exactly how you would make your own artificial vanilla. In the GRAS documents, there is a mention of syrups going up to 20,000 ppm (2%) but 0.3% vanillin seems typical. And having fun doing something more important than just buying stuff. Hence why I do this.

Darcy S. O'Neil

I use a 100% ethyl acetate, I'm not sure why they do a 10% solution now. If you contact them they can provide pure ethyl acetate, or there are other suppliers available: https://creatingperfumes.com/products/ethyl-acetate

Darcy S. O'Neil

Perfumer's Apprentice's Ethyl Acetate is 10% in Benzyl Benzoate. Is this the version you used, or do I need to find 100% Ethyl Acetate? Thanks!

Kent

So just out of curiosity, if I were to make my own Imitation Vanilla Extract like you buy at the store (it has been shown that in baked goods even the best tasters had a hard time differentiating between real and imitation vanilla as most of the complex aromatic compounds get cooked off). Could I just mix 0.2% vanillin and 0.04%ish Ethyl Vanillin with 35% ABV alcohol? In theory how high could I go if I wanted it to be stronger? Now that I've got all this vanillin from the cream soda recipe, I though it might be fun to make an extra potent imitation vanilla for anything that it is cooked in, and save my good homemade stuff for everything else (yes I know I can buy good cheap imitation extract, but what's the fun in that).

Chris Robinson

It should be fine, as long as it is pure. The site seems to indicate their products are pure.

Darcy S. O'Neil

Would the banana oil from Perfumers work?

Perfumers Apprentice (https://shop.perfumersapprentice.com) is my go-to for most of the individual flavour compounds.

Darcy S. O'Neil

I found most of the ingredients on amazon, do you have a preferred vendor for citral, isoamyl acetate or items not on amazon? Thanks for everything.

Sounds good, not planning to commercialize. I work in a pharmaceutical lab and I like how you go into the chemistry aspects like solubility, and pH. My pipette isn,t calibrated but I take into consideration if ithas a bias. I like soda making as a hobby and I prefer not to make a big batch since I end up adjusting for flavor.

Nancy Ellis

A metric drop is considered to be 0.05 mL, so 20 drops per mL. For most of the formulations I factor in the density of the oils at roughly 0.9, so a drop is 45 mg. And yes, drops are not an exact measure, but for most people starting out, this is the easiest way and the results work out fine. Obviously, anyone who wants to commercialize would need more accurate measures, but they'd be dealing with larger volumes as well, which is easier. Calibrated pipettes are great, but can be expensive.

Darcy S. O'Neil

So, what do you estimate a drop to be by volume? I remember in school it was ~50uL but in commercial labs drops is not exact enough for use. I want to scale down and have a pipettor.

Nancy Ellis


More Creators