Fig 1.
(i) Circular reaction zones were wax printed onto filter paper. (ii) Ascorbic acid solution was added to each circle in zone 1, Mo/Sb reagent to each circle in zone 2. After drying, the paper was backfolded to align the sample zones. (iii) Devices were sealed by lamination to encase the reagents and prevent contamination. (iv) Slits were cut into the back of the device (zone 1) to allow for sample entry.
Fig 2.
(a) The laboratory analysis workflow involved the paper device being placed into a dish with 20 mL aqueous sample. The sample entered through the slits in the back of the devices. After 3 min incubation, the formation of the blue colour on the upward facing side of the device was captured. (b) The workflow for volunteer sampling involved the collection of a water sample from a river and an aliquot being placed into a container. The paper device was dropped into the sample, the same as in the laboratory, and after 3 min, a photograph of the result was taken and uploaded via the RiverDip app.
Fig 3.
The RiverDip app was custom-developed to record test results in the field, with GPS location and time alongside photographs of the water quality and the surroundings.
(a) The ‘record navigation page’ shows the data that needs to be completed for each sampling. (b) ‘Capturing test PAD result’ will start a 3-min timer, after which an image can be captured and uploaded. Volunteers can compare the result to a colour intensity scale bar and select the colour intensity that matches their results the closest. (c) Screenshot showing a completed result ready for submission with date, time, record code and location and images of the PAD result, location and water quality.
Fig 4.
Formation of the blue PMB complex over time at different phosphate concentrations.
The colour intensity was found to increase for up to 2 min and then plateau (n = 3).
Fig 5.
Comparison of PO43- levels in freshwater samples detected using UV/vis spectroscopy via PMB reaction (n = 3) and modified PMB method on the paper microfluidic device with flatbed scanner image capture (n = 6).
Fig 6.
Typical calibration curve obtained with images captured from a smartphone camera.
The obtained colour development in the sample reaction zones were standardised against both the internal standard (blue square) and the negative controls to account for interferences from varying light conditions and background colour from the water sample (n = 6).
Fig 7.
Data collected via citizen-led sampling sessions across the North Sea region presented for public dissemination via a Google map embedded in the project page website (https://riverdip.com/map).
Since Google’s maps are copyrighted they cannot be displayed here, therefore the figure shown (produced using QGIS Desktop 3.16.10 software republished under a CC BY license with permission from QGIS) is similar but not identical to the map used to update volunteers on the progress of the monitoring campaign. Throughout 2019 and 2021 over 300 samples were collected by volunteer across the region. The location of the marker represents the GPS location of sampling, the colour of the marker represents the obtained phosphate level (none <1 mg L-1, low 1–3 mg L-1, medium 4–6 mg L-1 and high ≥ 7 mg L-1). Each marker links to other details captured on the RiverDip app at the time of sampling; including the user’s own interpretation of the phosphate level, a photo to record water turbidity, and any notes made during the sampling.