Doe 1

Using Density to Estimate the Alcohol Content of Common Liquors

John Doe

Chemistry 101 January 18, 2013 Dr. Maria Smith

Doe 2

Abstract

Commercial producers of alcoholic beverages use several methods to measure the amount of

alcohol by volume (ABV) that can be found in their products. Because the density of ethanol and water are both known, this study attempted to determine whether density could be used to approximate ABV. A series of ethanol/water solutions were prepared and weighed to create a density curve, and several commercial available liquors in varying proofs were sampled for density. Overall, the density curve created by the series was a good predictor of ABV, with the curve over predicting the whiskey, rum, and grain alcohol samples by 1.2%, 1.7%, and 1.8%, respectively. The likely cause for this small difference was likely ingredients that were added to the alcohol for flavor and which changed the density of the solution. Temperature and human error also likely played a role in the difference.

Doe 3

Introduction

Alcoholic beverages are a solution of ethanol (the psychoactive ingredient) water, and other substances, usually added for flavor. In the United States, the ethanol content of commercially available liquor is noted on the bottle as the proof, defined as being twice the percentage of alcohol by volume (ABV) in the solution. So a bottle of liquor that was 40% ABV would be labeled as 80 proof. This study will attempt show that density can be used as a means of measuring ABV.

Historically, the idea of alcoholic proof developed as a literal test of whether a liquid contained the claimed amount of alcohol. The test relied on the flammability of ethanol, namely that gunpowder will ignite in a mixture of ethanol and water that is greater than 57.15% ABV and will not in a solution that is less than 57.15% ABV. Eighteenth century sailors, who were paid in rum rations, would test whether their rations had been watered down by dousing gunpowder with the rum. It ignited the run was considered to have been “proved.” If the gunpowder did not ignited, the rum was considered “under proof.”1

Today, distillers use complex instruments that measure alcohol content by comparing the

density of the solution to the density of water. Density is defined as the amount of mass per unit volume. For liquids, density is measured in grams per cubic milliliter (g/mL3). Because of the expansion and contraction of liquids, density varies with temperature. At 20 degrees C, water has a density of 1.00 g/mL3, and ethanol has a density of 0.789 g/mL3. To test the labeled proof on commercially available liquors, a ABV versus density curve was created using a series of ethanol/water mixtures of known volume. The density of the purchased liquors was calculated and compared to the curve.

Materials and methods

ABV versus density curve

A volumetric pipette was to create a series of 10 mL ethanol/water solutions ranging from 35-95% ethanol in 5% increments. A 5 mL aliquot was then weighed and the density of the solution calculated. All solutions were kept at 20 degrees C. The density was then plotted against ABV and a best-fit curve calculated.

Density of commercial liquors

Representative brands were chosen for several types of liquors to represent a range of ABV: Jim Beam Original, white label, 80 proof (whiskey), Bacardi 151, 151 proof (rum), and Everclear, 190 proof (grain alcohol). 5 mL aliquots were taken from each bottle and weighed to calculate density.

Results

Doe 4

ABV versus density curve

The measurements and densities for the curve samples are shown in Table 1. Table 1. Weight and density for ethanol/water solution series.

35%40%45%50%55%60%65%

Water (mL)6.56.05.55.04.54.03.5

Ethanol (mL)3.54.04.55.05.56.06.5

Weight (g)47.4045.7544.7244.6544.3544.0143.75

Density (g/mL3)0.9480.9150.8940.8930.8870.8800.875

Table 1 (con.).

70%75%80%85%90%95%

Water (mL)3.02.52.01.51.00.05

Ethanol (mL)7.07.58.08.59.09.5

Weight (g)43.4542.7542.541.4041.0539.95

Density (g/mL3)0.8690.8550.8500.8280.8210.799

The scatter plot showed a linear relationship between ABV and density with the equation: y = -0.002x + 1.0012

The R-squared value was .9477, indicating a high correlation.

Fig. 1 ABV versus density scatter plot with best fit line.

Density of commercial liquors

The weight and density of the 5 mL aliquots of commercial liquor are in Table 2.

Table 2. Weight and density of commercial liquors.

Weight (g)Density (g/mL3)

Jim Beam (whiskey)46.320.926

Bacardi (rum)43.500.870

Everclear (grain alcohol)39.250.785

Doe 5

The difference between the sampled and predicted densities was as follows for whisky, rum, and grain alcohol, respectively: 0.011 g/mL3, 0.015 g/mL3, and 0.014 g/mL3. The percent error for each prediction was 1.2%, 1.7%, and 1.8%.

Discussion

The alcohol industry relies on indirect methods to calculate the amount of alcohol in their finished product. This experiment tested the accuracy of using density as a proxy to approximate ABV by developing an ABV versus density curve, then comparing predicted densities on commercial liquor labels.

The ethanol/water solution series resulted in a density curve with a high R-squared value,

indicating a strong correlation between ABV and density. This curve consistently over predicted the calculated density of the commercial liquors, meaning that the bottle contained a higher volume of alcohol than was predicted by the curve. However, the percent error in the predictions was consistent, suggesting that the curve was precise if not accurate and that the difference was due to factors that stretched across all three brands.

There are several possible reasons for this difference. The first is the presence of outliers that

can be seen in Figure 1. The outside data points, at 35% and 95%, do not fit the curve, and most likely skewed the curve. These outliers may be the result of human error in measurement or may indicate that the curve is less reliable at the highest and lowest ABV values.

Another possible reason for the consistent difference between the predicted and sampled ABV valued is that commercial alcohols often contain ingredients other than ethanol and water. Most important among these are likely to be ingredients added for flavor. For example, both the sampled whiskey and rum liquors were aged in barrels so that they would pick up flavor from the wood.2,3 However, the grain alcohol sample indicates that added ingredients may not be the main cause of the discrepancy, because Everclear is unflavored.

The last factor that may account for the discrepancy is the effect of temperature on density. This experiment was done at 20 degrees C, but it’s possible that proof alcohol for commercial labels is calculated at a different temperature. It’s also possible that the temperatures, and hence the densities, of the solutions changed after they were prepared and allowed to rest at room temperature.

Overall, however, it should be noted the percent error between the predicted and sampled ABV was small. The curve proved to be a very reliable predictor of relative ABV, meaning that the curve could be used to accurately predict the difference between ABV in an ethanol/water solution. The curve was also a somewhat reliable predictor of ABV.

Doe 6

Works Cited

1. Gately, I. Drink: A Cultural History of Alcohol; Gotham: New York, 2009.

2. Pacult, F.P. American Still Life: The Jim Beam Story and the Making of the World’s #1 Bourbon; Wiley: New York, 2003.

3. Gjelten, T. Bacardi and the Long Fight for Cuba: The Biography of a Cause; Viking: New York, 2008.