Hello from Hanna:
"A couple of wires twisted together is not a wire sensor, but it's a great start. The next steps are soldering, activation and: Tada!
It is truly amazing to see a few pieces of wire and some enzyme-cocktails turn into a device that can measure a real biological marker (Take that, mathematicians! We're making something real and functional: even when 𝝅=3).
It is equally fun to make them, even
repetative work is enjoyable when working with an interesting project along with fun colleagues.
The long-term stability test of our glucose sensors is one of those repetative tasks where you accept the process of dipping wires into different solutions, because you will really enjoy the result of your dedication... In a few months."
A sensor reacts to a physical change. To make a sensor, you first have to know what you want to measure and research possible markers for that parameter. For example: We want to measure stress, and some biological markers for stress are glucose and cortisol.
Once it reacts to the marker of your choise, more work begin: when, what caused it and how will your sensor fail?
Thermal cycling. As a stress-test thermal cycling forces the material to expand and contract repeatedly due to rapid, forced temperature change. The test is over once the test material breaks or it has gone through a given amount of cycles.
However, we only wanted to see the effect temperature has on the signal. We used one thermal cycle (ΔT=14°C, t=34hours) to test the correlation in temperature and signal strength, as well as response time. The increase in current at the end of the test is due to the thermal cycle ending and the chamber starting to return to RT.
Upon designing sensors you already know about several sources of error: like movement, temperature change, surprise-short circuits and the list goes on. We knew from the results of the thermal cycling that we needed to test the stability of the sensors at different temperatures. Our long-term temperature stability test was designed to test how long our sensors would remain functional when temperature was the only factor changed. We set up two seperate tests: One at room temperature (RT) and one with a set of six sensors inside a cabinet at 10°C. Screen printed electrodes (SPE) were included in the tests, as they are considered our gold standard. The tests at RT and at 10°C ran for 120 days and 93 days respectively.
Fun fact: Salmons, as most fish, are poikilothermic, meaning that their body temperature will be relativly close to that of its surroundings. Most of the heat transfer fish/surroundings happens in the blood-filled gills: As the water is filtered through, it cools the blood. The sensors were tested at 10°C because that is considered a reasonable mean body temperature for salmon living in Norway.
The glucose consentrations were changed twice a week: the sensors were put into 0mM glucose solution for 2h, then placed back into a 3mM glucose solution for the remaining time. This was done every Monday and Thursday for the test period.
The difference of output in our sensors (0mM/3mM glucose solution) varies from a maximum of 1.21658 μA to a minimum of 0.0187362 μA.
The finished implantable system has, with optimized power consumption, a predicted lifetime of 3 months and the sensors have proven themselves to work for at least that amount of time. Some drift was noticed but was concluded negligible due to the intended usage of our sensors. The results of the long-term tests were fitted into an exponential function, which we can use to predict values (See? Mathematics does come in handy!).
Until next time,
Hanna T. Bråthen
Engineer at ZP
Comments