$1 Graphene Sensor Identifies Safe Water

If you live in a place where you can buy Arduinos and Raspberry Pis locally, you probably don’t spend much time worrying about your water supply. But in some parts of the world, it is nothing to take for granted, bad water accounts for as many as 500,000 deaths worldwide every year. Scientists have reported a graphene sensor they say costs a buck and can detect dangerous bacteria and heavy metals in drinking water.

The sensor uses a GFET — a graphene-based field effect transistor to detect lead, mercury, and E. coli bacteria. Interestingly, the FETs transfer characteristic changes based on what is is exposed to. We were, frankly, a bit surprised that this is repeatable enough to give you useful data. But apparently, it is especially when you use a neural network to interpret the results.

What’s more, there is the possibility the device could find other contaminants like pesticides. While the materials in the sensor might have cost a dollar, it sounds like you’d need a big equipment budget to reproduce these. There are silicon wafers, spin coating, oxygen plasma, and lithography. Not something you’ll whip up in the garage this weekend.

Still, it is interesting to see a FET used this way and a cheap way to monitor water quality would be welcome. Using machine learning with water sensors isn’t a new idea. Of course, the sensor is one part of the equation. Monitoring is the other.

15 thoughts on “$1 Graphene Sensor Identifies Safe Water

  1. So currently it is one use only:
    “His team is also studying ways to remove the contaminants from the graphene to make the sensors reusable.”

    I wonder how the contaminates congregate in a large volume of water. Are E-coli equally dispersed through out the water or mostly attracted to the internal surfaces of the container. And is there a slightly higher concentration of heavy metals closer to the bottom of a container than the top due to gravity. Anyone know ?

    1. E. coli can be motile and may tumble randomly, or may be attracted towards or repelled from certain directions depending on what else is in the solution. They can also colonise surfaces depending on the conditions and which mutations it has. A quick stir should redistribute most of them, and even if you’ve got biofilm there should be *some* free-living cells to detect.

      At a quick skim of the paper the authors seem to think they start getting a signal from ~5 cfu/mL (roughly culturable cells per mL), while at another quick glance the infective dose of some E. coli strains is as low as 10-100 cells, or high as tens of thousands of cells, so it looks like they’re in the right ballpark for some strains. It also looks to be antibody-driven, so it should be possibly to tune it to specific organisms?

    2. *Particles* of heavy metals, e.g., tiny metal flakes, would indeed fall to the bottom. But *dissolved* heavy metals aren’t metallic, they are dissolved ions with water molecules surrounding them. In a solution the concentrations of such ions are the same in any part of the solution. Molecular motion of other solvated ions and of water molecules keeps the ions suspended regardless of where in the solution they are.

      1. Analytic Electrochemistry is a well developed field. I built one with an Apple 2 for my dad’s lab 40+ years ago. He had originally done that science when they were cutting the peaks out of paper and weighing them to estimate concentrations. Today the units are hand held commodity products.

        IMHO Detect ions that way and focus efforts on bacteria.

  2. The fabled microchips! /s

    Lets be real for a moment. This is a self-published report. It has not been reviewed and accepted by peers within the scientific community. The contributor, Micro-Raman Spectroscopy, isn’t even a real entity and instead is a spectroscopic technique used to determine vibrational modes of molecules found within a substance.

    Get your shit together, pack it up, and take it to the shit-store and sell it because nobody here wants to buy your shit.

  3. Have we forgotten “extraordinary claims require extraordinary evidence”.

    Demonstrate both the ability to distinguish between compounds and the minimum concentration required for any of these. Then start it on mixed cocktails of contaminants. Oh yeah, peer review and repeatability…all that.

    Back in the 1980s people discovered that semiconductors would sense organic vapors – I was involved in one of these trying to repeably detect levels of a specific compound. Of course it didn’t work because it detected *all* the compounds (and a detector for that compound still hasn’t surfaced though it would make someone a tidy sum). DIdn’t matter because the results were cherry-picked, the publications flowed, and the funding was continued. Quite disillusioning for a wet behind the ears grad student.

    1. The organic vapor is interesting. Bio sensors can be made of protein which activates in the presence of a given type of vapor. To take an exerp from organic chemistry; form equals function. I have not seen this approach yet but you or anyone in the field are more than welcome to use the idea.

    2. They’re (grant sources) actually looking specifically for a general purpose detector. They don’t care about exact quantitation they want a qualitative sensor. When it comes to water sources and occupational safety knowing that *some* contaminant(s) is/are present is as useful as knowing what contaminant it is.
      So while you see detecting all of the compounds at once as “not working” and a bad thing, for a lot of cases it’s not. Toxic compounds and pathogens are harmful at threshold values and they’re frequently additive with other contaminants so having a cheap Go / No-Go sensor still has a place.

        1. Yes. But that’s not the use-case of this sensor.

          This is for water intakes, discharges, or some other scenario where you don’t (immediately) need to know what the contamination is, just that it is contaminated. It’s not meant to break things down to their constituent parts because for a lot of scenarios knowing the sum is just as useful as the value of the constituents.

  4. This as addition to a finger print sensor. PC: User your skin bacteria have changed please eat more vegetables. In Hopspitals: Geeez Dr. please go back and wash your Hands again. Or a instand tester for sexual infections, for one night stands. Or test hygiene on puplic toilets, Tester: if you use these toilets you get the worst arfrcan diahear ever :-D Or icecream tester for samonella.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.