Is it possible to figure out Coca Cola's recipe using analytical chemistry? by Christopher VanLang
Answer by Christopher VanLang:
This is a pretty common water-cooler discussion amongst analytical chemists and was probably the one of the first questions I asked when I started analyzing mixtures. "Why couldn't you just put Coca-Cola in an HPLC-MS/MS and figure out what's inside?"
The general consensus is that it's definitely possible but would require a lot of effort and money. Then again if Coca-Cola sells $50 billion in non-alcoholic beverages and you're a competing company that spends billions in R&D and start with the letter P…
Pfizer could probably figure it out. But if they had the formula, they would be keeping it private.
Back to the science. The general feeling is that while you should be able to use an HPLC to separate to get approximate ratios of the components and the MS/MS to determine the actual ingredients, that in of itself only provides a small portion of the puzzle. The truth is, the ingredients that go into Coke aren't exactly a mystery, it's basically sugar water + hydrolysates. However, because of the caramelization process that is presumed to be a part of that sugar water, you now have a very complex heterogeneous mixture full of polydisperse polymers of god knows what. Thus, the HPLC peaks aren't going to be clean and your MS/MS data is going to be very noisy. A very basic analysis would just tell you that you have a lot of sugar but very little insight into the actual properties of your complex sugars.
So now you're in the domain of some really challenging carbohydrate analytical chemistry. Assuming that you do have a good understanding of the composition of your sugar water, that in of itself doesn't tell you much. Now you're in the regime of very challenging process chemistry. Coke is known for doing very strange things to ensure the protection of their trade secrets like breaking up their process steps to very physically distant factories and obscuring their supply chain so that no one can figure out what exactly is going on. But there are two things that are generally known about the Coke process:
- There are a lot of components that get produced in several locations and those components get passed around so that the end to end process can never occur in a single place.
- There is a critical component of Coke that is between the pH of 2.0 and 2.5
Why does everyone know #2? According to the EPA, corrosive waste is defined as a chemical with a pH of less than 2.0 or greater than 12.5. Why 2.0 and not 2.5? Because Coke told the EPA that it didn't want to drive trucks around with giant signs that says "corrosive". Also Coke is 2.52. Coke classic is closer to 2.50.
As I was saying. Caramelization is an extremely unpredictable process and very difficult to scale. The order of addition of the ingredients is very particular and there are a huge number of permutations that would make sense. If you're going to put in that much effort, you might as well call it something else.
Maybe something blue. How about Pfizer-Cola?
Is it possible to figure out Coca Cola's recipe using analytical chemistry?