Public Lab Research note

Detection of Distilled Spirit Adulteration (with water) Using Visual Light Spectroscopy

by ygzstc | | 3,367 views | 4 comments |

Read more:

ygzstc was awarded the Basic Barnstar by aldehyde for their work in this research note.

Disclaimer: All the information (including hardware, software, experimental setup, procedure, and results) in this research note is provided "as is" without warranty of any kind. Author makes no warranties, express or implied, that they are free of error, or are consistent with any particular standard of merchantability, or that they will meet your requirements for any particular application and/or problem. They should not be relied on for solving a problem whose incorrect solution could result in incorrect claims which may or may not lead to any kind of monetary loss related to trade and/or legal liability. If you do use them in such a manner, it is at your own risk. The author disclaims all liability for direct, indirect, or consequential damages resulting from your experiments and claims based on their results.


Distilled spirit adulteration may also be considered as an example of “Economically Motivated Adulteration (EMA)” [1, 2]. Similar to other types of adulteration, most common types of adulteration techniques for distilled spirits are also “dilution” and “substitution” with one or more of the possible adulterants such as, water, ethanol, methanol and/or low-quality versions of the same or similar type of spirits.

So the question is: Can we come up with a cheap but effective method (based on UV-Vis spectroscopy) for household use to detect distilled spirit adulteration?

Setup, Sample Preparation and Data Collection

In this study, Jack Daniel’s whiskey and water (for adulteration) are considered. In addition, Public Lab’s spectrometer (, and its data collection software “Spectral Workbench” ( and few extra tools are used (See Figure 1).


Figure 1 - Setup (left) and Jack Daniel’s (right)

Total 7 samples are created. Starting from pure Jack Daniel’s whiskey, 90%, 80%, 70%, 60%, 50% and 40% diluted (with water) samples are prepared.

Once the spectrometer is calibrated with CFL, physical setup (shown in Figure 1) is set. First, a spectral data is recorded with the empty petri dish (90 mm diameter, glass) and using a 6000K LED ( We call this spectrum as “baseline”.

Later, from each sample, 20 ml is taken and poured in the petri dish and the spectral data are collected. (Those spectral data are on the Public Lab’s website and nomenclature of the data is provided in Appendix.)

Collected spectral data is then smoothed with 5th order Savitzky–Golay filter and the difference between each sample spectra and the baseline spectrum is calculated. Resulting spectral data (i.e. absorbance) is shown in Figure 2.


Figure 2 – Absorbance spectrum of varying concentrations and various (zoomed) regions of interest

It is worth noting the fact that, it is possible to detect/model the level of adulteration using the spectra of Red, Green or Blue Channel (or combination of them) data as well. However, optimal region selection may/will differ for each color channel.

As a measure for evaluating the level of adulteration (relative Jack Daniel’s concentration in the mixture), area under the curve (AUC) is selected [3]. This is more robust measure compared to “peak height” which fluctuate more causing noisy measurements. Also note that, in order to have stable measurements (for omitting negative value from numerical integration), all spectra are lifted up by 5 (see Figure 2).

To calculate the AUC, three different wavelength bands ([430-500 nm], [560-640 nm] and [720-740 nm]) are selected. Note that the [560-640 nm] band is different than other two bands in the sense that, the absorbance values of different concentrations are in reverse order. In other words, absorbance values of higher concentrations are lower in this band. Therefore, AUC values calculated at [560-640 nm] band are subtracted from the summation of the AUC values calculated at [430-500 nm] and [720-740 nm] bands. More specifically:

AUCfinal = AUC[430-500] - AUC[560-640] + AUC_[720-740]

Values of the area under the curve (AUC_final) with respect to different adulteration levels are shown in Figure 3. It is clear that the level of adulteration and AUC_final exhibit almost perfect linear relation.


Figure 3 – Adulteration level and AUC_final exhibit almost perfect linear relation

Results and Discussions

Results of this preliminary study show that, using UV-Vis light spectroscopy (using the spectrometer developed by Public Lab), it’s possible to detect/model/measure distilled spirit adulteration (with water) in an efficient and simple way. Furthermore, these results indicate that it might be possible to detect and measure other type of alcoholic beverage adulterations as well.


[1] “Application of Mid-infrared Spectroscopy for the Measurement of Several Quality Parameters of Alcoholic Beverages, Wine and Raki”, Burcu Ozturk, Dila Yucesoy and Banu Ozen, Food Anal. Methods (2012) 5:1435–1442.

[2] “A Flow-Batch Analyzer for UV-Vis Spectrophotometric Detection of Adulteration in Distilled Spirits”, Elaine C. L. Nascimento, Mário C. U. Araújo and Roberto K. H. Galvão, J. Braz. Chem. Soc., Vol. 22, No. 6, 1061-1067, 2011.

[3] “A New Method of Area under the Absorbance-Wavelength Curve for Rats Total Metabolomic Pharmacokinetics from Yangxue Injection with Multicomponents,” Lihong Zhang, Xiaojin Xiao, Zhenzhen Yang, Mengli Jiang, and Xiaodong Li, Journal of Spectroscopy, vol. 2013, Article ID 919023, 8 pages, 2013.


test2: CFL spectra used for calibration

d4-l: Spectra of the LED lamp

d4-b: Spectra of empty petri dish

d4-JDx: Spectra of sample – x% Jack Daniel’s (x=40, 50, 60, 70, 80, 90 and 100)


I proclaim this study to be highly replicable, and encourage others to reproduce Yagiz's results. Not that we should have any doubts about Yagiz's work, although he never mentioned what happened to all that bourbon and water.

The area under the curve approach at selected wavelength ranges seems to be a very robust way to make this type of comparison. I have been wondering about doing absorbance spectroscopy with a camera that always uses automatic exposure control. The more diluted samples of bourbon absorb less light, so when compared to pure bourbon, their spectra should be brighter. But the camera adjusts for each image so the average brightness of all the spectra will be about the same. Therefore standard absorbance spectroscopy is not really possible. In that approach, the pure bourbon spectra would be subtracted from the diluted sample spectra, and the difference would be the difference in absorbance. This won't work when the camera adjusts the spectra for all the samples. Yagiz's approach seems to solve that problem.

Yagiz, is this why you used this approach? Or is this approach a standard way to compare absorbance spectra?

Great note, Chris

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

Thanks Chris :)

As far as I know, this seems to be the standard way to compare absorbance spectra...But modelling part may vary depending on the tools and their limitations. For example, automatic exposure control (which is disabled in the webcam comes with the desktop spectrometry kit - as far as I know - am I right? Jeff, Mathew ?) problem you have may be solved by looking into RGB channels and some ratios between those bands (similar to NVDI may be?) and defining that as a metric (which may be not affected by automatic exposure control) can be used to model the concentration variations. Mathematically speaking, as long as you decide the right modelling scheme (depending on the problem you have in your hand), you can come up with a way to measure whatever you want to measure...

And, I never discard my samples after I am done with data collection :) -yagiz

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

Automatic exposure control is not disabled -- but some tests have shown webcams to be pretty linear: But you can also take two spectra at the same time, as we believe webcams have equal exposure across a single image or frame.

This is great.

Reply to this comment...

Himmm...I was aware of straylight's great research note but the linearity you mentioned there is not about webcam but about peak heights vs. concentration (i.e. Beer's Law) I believe...Beer's Law says the log of the ratio of the intensities (I_out/I_in) gives you the absorbance A at a given wavelength... And limitations of this law are more significant for very high and very low concentrations (

AUC is little different and shown to be more robust...

But I have to dig little more to figure out if I am correctly evaluating my findings...Thanks a lot for the comments...


Reply to this comment...

Login to comment.