How much Ethanol is in my beer?

Whether you drink a beer from a major manufacturer, the local brewer down the street, or even brew the beer yourself, knowing the ethanol content of that beer is an important piece of information.  Up until 1995 the alcohol content wasn’t allowed to be on beer labels at all until the Coors Brewing Company successfully challenged the law in the Supreme Court.  Now, brewers are free to either disclose that information on the labels or not.  

How is Ethanol content, or Alcohol By Volume (ABV), measured?  The old fashioned way is by what is called a hygrometer.  A hygrometer is a sealed glass weighted tube that will float in liquids to give the specific gravity of the liquid.  The brewer would take the measurement of the pre-brewed beer, called a wort, and then a second measurement after brewing and use a simple calculation to determine the ABV.  

Another traditional technique would be a distillation, where the beer would be placed in special glassware and heated.  The vapor would be cooled and collected into another vessel and the masses of each would be used to determine the ABV.

Both of these techniques use a lot of the beer to run properly and are time consuming.

A more modern technique is to test by GC-FID.  This technique uses less than a milliliter of the beer sample and once the GC and methods are set up, the method is very simple and quick.

In this post, we take a go at measuring alcohol content in several canned and bottled beers off the shelf.

Method Selection, Standards, and Calibration Curve:

The first step is to find a GC method.  We found a good starting point online for Ethanol (the specific version of alcohol in beer) that we started with:

Next, we need to run some standards on this method to confirm that the method does in fact work.  We ran standards of 0.5% Ethanol in water, 1.0% Ethanol in water, and 5.0% Ethanol in water to cover the basic range of Ethanol found in beer.  We started with a 20% Ethanol in water stock solution, created a 10% Ethanol in water working solution, then diluted the working solution to create the three standards (0.5%, 1.0%, and 5.0%).

You can see a screenshot of the 1.0% Ethanol standard below, which shows a large Ethanol peak about halfway through the run.  The same is true of the 0.5% and 5.0% standards.

Chromatogram of a 1% ethanol standard

Next we need to create calibration curve using the 3 standards we ran, so that we can properly quantitate our unknown samples.  Creating a calibration curve using these standards shows a reasonable R2 of 0.9990, indicating that we should be able to get reasonable quantitation of Ethanol in water based samples in this concentration range.

Calibration Curve of Ethanol

Next, we grabbed some beers from various vendors including a popular macrobrew beer, the light version of that macrobrew, an 8% beer, and two different nonalcohol beers, which are labeled 0.0%, but the fine print explains that this means less than 0.5% ABV.  We also included a 6% ABV selzer for fun.  Note: Best practice would have been to add a 10% point to our calibration curve so that all samples fell within the concentration range of the calibration curve, but we proceeded anyway.

Sample A: selzer, 6% ABV

Sample B: popular macrobrew beer, 5% ABV

Sample C: light version of the popular macrobrew beer, 4.2% ABV

Sample D: nonalcoholic beer, < 0.5% ABV

Sample E: nonalcoholic beer, < 0.5% ABV

Sample F: strong beer, 8% ABV

Sample Preparation:

For sample preparation we simply poured about 100mL of each beer into a 150mL beaker and allowed the beers to set overnight to decarbonate and reach room temperature.  We then sonicated the beer samples for 30 minutes to further assist the decarbonation.  Next we took a 1mL aliquot of each sample and diluted it to around 1% alcohol concentration based on the labeled amount of alcohol (which corresponds to around 80ng on column using a 100:1 split).  This was to ensure that we were loading an appropriate amount of sample onto the column for each run.

GC-FID Procedure:

We then loaded the samples in the GC-FID and let them run on the method outlined above with the calibration curve show above with 1uL liquid injections of each.  We only did one injection for each sample, and the following screen shots from the software show the results of each run.

Results:

The following table summarizes the amount of ethanol in each sample based on the calibration curve we ran the samples against.  Sample A, the selzer, is very close to the labeled amount, Samples C and F, the light beer and the high alcohol content beer are within 0.5% ABV of the labeled amount, and the two nonalcoholic beers are both below 0.5% as the label indicates they should be.  Sample B, the macrobrew beer reports almost 1% higher than the labeled amount.

One factor that could be affecting the accuracy of our measurement is the amount of sample we’re putting on the column.  The peaks do appear to be fronting, which would indicate we’re overloading the column, which is supported by the fact we’re putting around 80ng of Ethanol on a 0.25mm ID column, which depending upon which recommendations you look at is either at the high end of recommended analyte on column or above the recommended amount of compound on column.  A next step from here would be to run this experiment on a larger diameter column (like a 0.53mm ID column) to see if the peak shapes and results improve.

But for a quick and simple first run at measuring Ethanol in beer, not too bad.