Even without the rise of online pharmacies, there have been multiple cases of counterfeiting of food and medicines, some due to carelessness, others due to greed. An unfortunate part of the story is that most counterfeiting cases are quite inconvenient, and lives could have been saved had we had simple and widely available tests for contaminants.
That’s exactly what a team of engineers recently tried to achieve. They’ve taken some pretty old ideas and flipped them over to create a rather innovative testing system that can detect adulteration in liquid medicines and maybe even food.
The challenge in creating a generic test for contamination is that anything can get into food and medicine. The key to this new idea is that you don’t necessarily need to know what’s been added, just that it’s different from the standard wording. In almost all cases, changing the formulation changes the density of a liquid. A sensitive mass sensor must therefore be able to detect drugs that have not been produced properly.
The key to this sensitivity lies in the acoustics: the tone of an instrument depends on how the sound wave is resonated by the instrument. For example, the pitch of a guitar string depends on its tension, mass, and length. This property has been used with great success outside of music. For example, a quartz crystal will change its resonant frequency as its mass changes. This provides a very sensitive balance to measure small changes in mass. But quartz mass balances, as they are called, are quite expensive.
What these researchers were looking for is an inexpensive alternative. Especially in the developing world, hospitals and individuals cannot spend thousands of dollars on highly sensitive mass balances. They do have access to musical instruments.
However, the principle remains the same: change the mass of the vibrating material and the tone of the note it produces will change. To that end, the researchers decided to use a modified version of an ancient African instrument called the mbira. The mbira consists of a set of steel teeth that hang over a sound box. The length, width and density of the teeth determine their nuts.
The researchers removed the teeth and replaced them with a single hollow metal tube tuned to a G sharp two octaves above middle C. When the tube is filled with water, the tone drops to F sharp in the same octave. The change is clearly discernible by ear.
Tuning in to smaller changes
Smaller changes in density, such as adding small amounts of salt to the water, are indistinguishable by ear. To help in these cases, the researchers created a script to analyze the sound online. A user records the sound on his smartphone and uploads the sound file to the researchers’ website. The script finds the instrument sounds and tries to find a point where only the root – the root is the lowest note the tooth can produce – is present. It uses that sound window to calculate the frequency of the tooth for the user.
The researchers showed that a single-note mbira-like instrument can detect salt concentrations with a resolution of about 12 mg/ml, which is enough to distinguish seawater from freshwater, but not much in between. However, that test is almost a worst case, because the mass difference between seawater and freshwater is only 2.5 percent.
The researchers also showed that a common counterfeit — swapping glycerol for diethylene glycol during cold medicine production — was easy to detect.
However, this only works if drugs are reliably consistent in their production, as they are the baseline. To test this, the researchers obtained cold medicines produced in several batches and measured their frequency. They found that the drugs were all the same down to 0.025 percent, while the change between glycerol and diethyleneglycerol is more than one percent.
Now most people don’t want to destroy musical instrument to make drug detector. However, you don’t need to do this to do this. The same result can be obtained with an instrument consisting of a solid plank as a sounding board with something holding the tube firmly in place above the plank.
I like cheap sensors
This could even be a record in terms of cheap sensors: scrap wood, a few screws and some metal tubing, plus a cell phone. And as the author notes, smartphones are becoming more and more common worldwide.
I also think that the sensitivity can probably be improved and is probably a necessity. At the moment, the mbira detector can detect really clumsy counterfeiting, but it would struggle with smaller density differences. Perhaps that can be improved, either by using a more refined analysis or by changes on the instrumental side.
ACS Omega2018, DOI: 10.1021/acsomega.8b01673 (About DOIs)