Team:UFRGS Brazil/Measurement

Glyphosate is a really difficult analyte to measure because its molecular structure do not have charged atoms or functional groups that are easily distinguished (Teofilo et. al, 2004; Gil et. al, 2013; Oliveira et. al, 2018 ) from other molecules. Also, it is a very small compound and it is found in very small concentrations (ppb), which makes it harder to couple with identification methods.

The existing golden standard method to quantify glyphosate is Liquid Chromatography-Mass Spectrometry (LC-MS), which consists in the application of the sample dissolved in a specific solvent in a column that operates through liquid chromatography. After this initial step, the sample reaches the mass spectrometer and is ionized by electrospray, then passes through the quadrupole mass analyzer and is measured by the photomultiplier detector.

Why not to use the traditional method?

However, this method depends on high technical skills, extremely expensive equipment and a well trained staff. Moreover, the determination of analysis parameters depends on extensive previous studies and high comprehension of the analyte and its vehicle.

In light of this, we decided that a new approach should be used to detect glyphosate. We needed a method that is easily operated, time saving and cost-effective. This could be used in our project as a tool for quickly and faithfully determine glyphosate concentration in rivers and lakes, being linked to our filtering system. In Brazil’s context, this could be used for routine tests in freshwater, improving fiscalization and providing a cheap technique for science application, given that Brazil is facing its biggest scientific funding crisis. Lastly, it could also provide a device which could be applied for other iGEM teams’ projects and adapted to a broad range of analytes.

What it did

In order to find a new manner to measure glyphosate, our research reviewed a set of techniques based on voltammetry, which comprises a collection of analytical methods involving the application of different electrical potentials in a sample in order to determine its response at electric current levels.

Overall, based on Oliveira et. al, 2018, we built an electrolytic cell with three electrodes - reference, working and counter - which should be able to detect glyphosate indirectly as its not electroactive natively.

We based our quantification method on the principle that copper ions naturally reduce themselves in an electrolytic environment. Literature describes that glyphosate interacts with copper ions in an unknown manner that amplifies the electrical current generated by copper reduction.

Our tests were designed to search for a working electrode that would be able to detect analytically this current change created by glyphosate interaction with copper. The main system remained unchanged and consists of an Ag/AgNO3 reference electrode and a Pt counter electrode where oxidation occurs. The electrical potentials were produced by a potentiostat and changes in the current were measured by it.

Eletrolytic cell example. In this image, carbon-ink 3D-printed working electrode (In 3D-printed electrodes, reference and counter-electrodes are printed together with the structure over a chip)

Our Measures

As we wanted to develop a cheap method, we did not perform chemical modifications at our electrodes or treatments at our analytical cell, in contrast with literature reports.

First, we tested the glassy-carbon as the working electrode. This carbon arrangement provides a surface that acts as a conductor to charge changes. We performed measures using water as the cell solvent and tested three different buffers: sodium acetate, ammonium nitrate and phosphate-buffered saline (PBS). These three buffers maintained pH at 4, 9.5 and 6.8 respectively. pH variations alters the interactions between copper ions and glyphosate. Also, we tested the addition of KCl 3M as the supporting electrolyte. The tests were conducted with three different methods: Squared-wave, Cyclical and Pulsed voltammetry. Glyphosate was added at 20 µL drops from a solution of 5000 ppb.

For all tested conditions and possibilities, the results stayed the same: we did observe changes in electrical current in response to voltage, however, these ups and downs were not analytical. For example, the same sample did not produced the same variation, or the variation was not bigger than copper addition at the beginning of our tests. As a result of this experiment, we conclude that this system cannot be used in this way.

We also tried to repeat the measures using different working electrodes as it follows: carbon-ink 3D printed, carbon nanotubes 3D printed, copper wires, and golden particles.

For all the possible combinations, our results remained the same: no analytical data could be observed.

Fig. 1- Overall measures. Measures performed on glassy-carbon electrode, with addition of KCl 3M on Squared-Wave mode. Glyphosate was added as volume from 5000 ppm solution as indicated

Fig. 2- Detail. Detail from measure of 60 microliters of glyphosate from previoues rsultes. As denoted, duplicate samples did not performed the same measure, which was observed in all tests

Fig. 3- Overall measures. Measures performed on glassy-carbon electrode, with addition of KCl 3M on Pulsed mode. Glyphosate was added as volume from 5000 ppm solution as indicated

Fig. 4- Detail. Detail from measure of 20 microliters of glyphosate from previoues rsultes. As denoted, duplicate samples did not performed the same measure, which was observed in all tests

Our conclusions

To measure glyphosate through a very simple system, we could not observe analytical variation that could correlate with glyphosate concentration. As denoted before, glyphosate is an analyte very difficult to work with, as it rapidly decomposes to AMPA and also, has no charged or electroactive groupments in its structure. However we did observe changes in electrical current in our designed experiments. This could imply that more experiments and variables introduced in the analytical cell could help to stabilize interactions and produce regular changes in the signal. At the end, more experiments need to be replicated in this system to make it useful.

How to calibrate a pHmeter arduino shield

We used to do our monitoring system a analog sensor pH-4502c. It has 5 pinouts: To, Do, Po, GND, GND and VCC. The first 2 pinouts weren’t used in this project, although we have to connect the other 4: Po is the data pinout read by A0 port, GND and GND, both linked to Ground at Arduino and VCC to 5 volts.

Usually, a pH sensor correlates 5 volts to pH 14 and 0 volts to pH 0, but pH-4502c correlates 0 volts to pH 7, and it means that acidic solutions will go into negative values and it won’t be read by the analog port on Arduino, and that’s when we have to do the first part of calibration of this pH sensor: we will have to create a offset of 2.5 volts and correlates it to pH 7. It can be done adjusting the offset pot in the shield, that is the blue pot close to the NBC conector (the part that connects the shield to the probe).

The first step it to assembly the circuit as if we would measure something but disconnect the probe and connect a wire inside the NBC and write a simple code just to read the analog value. The other tip of the wire will be touching outside the NBC, creating a short-circuit. This short-circuit corresponds to the pH 7. As we want the pH 7 to correspond to 2.5 volts, we have to adjust the offset pot to it by spinning its screw -that works like a potentiometer and reading the values of the short-circuit until it hits 2.5 volts. The images below shows how the short-circuit is and the screw we have to turn to calibrate the offset.

Fig. 5 - Potentiometer

6- Short circuit.

The second part of calibration is algebraic: we will need at least two different pH solutions to calculate a equation for a straight line. Now with the probe connected to the NBC, write a code to read the analog value. We tested with a pH 10 and 4 solution. The Y axis corresponds to the pH and the X axis to the volts. Now draw a line connecting the 2 dots and calculate the equation y=mx+n, where m is the difference between pH over difference between volts and n is when x=0. When we have this values, we can implement that in code, like pHvalue=mVolts+n and it will give the right number of pH.

References

Martinez Gil P., Laguarda-Miró N., Soto Camino J., Masot Peris R. Glyphosate detection with ammonium nitrate and humic acids as potential interfering substances by pulsed voltammetry technique. Talanta. 2013;115:702–705. doi: 10.1016/j.talanta.2013.06.030.

Oliveira, P. C., E. M. Maximiano, P. A. Oliveira, J. S. Camargo, A. R. Fiorucci, and G. J. Arruda. 2018. "Direct electrochemical detection of glyphosate at carbon paste electrode and its determination in samples of milk, orange juice, and agricultural formulation." J Environ Sci Health B no. 53 (12):817-823. doi: 10.1080/03601234.2018.1505081.

Pajares G, Peruzzi A, Gonzalez-de-Santos P. Sensors in agriculture and forestry. Sensors (Basel). 2013;13(9):12132–12139. Published 2013 Sep 10. doi:10.3390/s130912132

TEOFILO, Reinaldo F. et al. Experimental design employed to square wave voltammetry response optimization for the glyphosate determination. J. Braz. Chem. Soc. [online]. 2004, vol.15, n.6 [cited 2019-10-20], pp.865-871. Available from: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0103‑50532004000600013&lng=en&nrm=iso ISSN 0103-5053. http://dx.doi.org/10.1590/S0103-50532004000600013.

Zhang, C., She, Y., Li, T. et al. Anal Bioanal Chem (2017) 409: 7133. https://doi.org/10.1007/s00216-017-0671-5