*2.1. Spray Methodology*

The Micron Track Sprayer (Micron Sprayers Ltd., Bromyard, UK) consists of a conveyer belt along which a spray head with nozzle can move vertically up and down (Figure 1). The speed of the nozzle is adjustable using an electronic hand-held controller, and power is provided by a rechargeable battery pack. The spray head is connected to a pressurized spray tank of the Micron compression sprayer. Extendable arms at the top and bottom of the track sprayer were set 45 cm from the wall surface, and the equipment was levelled

horizontally using a standard spirit level. The travelling speed was set to 0.45 m/s, which corresponds to 1 m sprayed every 2.2 s as per WHO guidelines [2].

**Figure 1.** A schematic overview of the sprayer adapted from the Micron product manual is shown on the (**left**). A photo showing the set-up of track sprayer in the lab (**centre left**) and in an experimental hut (**centre right**) is shown, and a schematic overview of the filter paper positions on plywood panels, with values in centimetres, is shown on the (**right**).

The Micron track sprayer was compared to manual spraying, performed as detailed in WHO guidelines [2,16]. Before manual spraying, the spray operators were extensively trained using standard operating procedures (SOPs) based on the WHO guidelines which details lance speed and angle, distance from the wall, and speed of movement vertically up and down the walls during application. Spray tanks were calibrated and maintained according to good laboratory practice (GLP) standards. Sprayers were calibrated and deemed acceptable when spraying 550 mL ± 10% per minute. Spraying was carried out at an application rate of 30 mL/m<sup>2</sup> using a 1.5 bar CFV. To ensure comparability between the two spraying application methods, the same spray tank, insecticide solution, 8002E flat fan nozzle, and CFV were used for both manual spraying and automated track spraying.

For both phases of the comparison, a similar protocol was used; minor differences between the round with fluorescein and the round with insecticides are detailed below.

#### *2.2. Spraying of Fluorescein*

The first phase of spraying was performed using 0.1% *w*/*v* fluorescein sodium salt diluted in water. A 3.55 m × 2.00 m tiled wall surface was marked up to accommodate five 75 cm spray swaths (70 cm spray + 5 cm overlap). Each set of five swaths constituted one replicate test. Filter papers (10 cm diameter, Whatman No.5) were held in place in Petri dish lids using a plastic ring; the lids were attached to the wall surface using self-adhesive Velcro strips in a grid pattern with three horizontal positions and five vertical positions per swath (see Figure 1). Vertical positions were located at the following heights: 1.80 m (high), 1.40 m (mid-high), 1.00 m (centre), 0.60 m (mid-low), 0.20 m (low). Horizontal positions were at 0.2 m (left), 0.375 m (centre), and 0.55 m (right) from the left edge of the swath. Six replicates using the track sprayer and five replicates spraying manually were performed using the fluorescein water solution. In total, 1045 filter papers were analysed. Track spraying was performed using a Micron CS10 compression sprayer tank. The target spray rate was 2 metres in five seconds; a metronome app was used to assist the manual spray person to follow an even spray rhythm. The spray time and direction (upwards or downwards) was recorded for each swath.

#### *2.3. Spraying IRS Products*

Three IRS products containing different active ingredients were sprayed in experimental huts: broflanilide (VECTRONTM T500, Mitsui Chemical Agro Inc., Tokyo, Japan, batch no 18I-3671), pirimiphos-methyl (Actellic® 300CS, Syngenta, Basel, Switzerland, batch no BSN9A2383), and a deltamethrin + clothianidin combination product (Fludora Fusion®, Bayer AG, Leverkusen, Germany, batch no EQ13001804). Target application rates were 100 mg/m<sup>2</sup> for broflanilide (BRF), 1000 mg/m<sup>2</sup> for pirimiphos-methyl (PMM), 200 mg/m<sup>2</sup> for clothianidin (CTD), and 25 mg/m<sup>2</sup> for deltamethrin (DLT). A different spray tank (Micron CS14) was used for each insecticide product. For both application methods, the spray tanks were positioned stationary on the floor, which differs from the WHO guidelines for manual spraying where the tank is typically carried over one shoulder. For each insecticide, four panels were sprayed with the track sprayer and four panels were sprayed manually (Figure 1). Filter papers (9 cm diameter Whatman No. 1) were fixed inside Petri dish lids with sticky tack; the lids were pierced in the centre and attached to the panel using shoe tacks. Filter papers were positioned in a grid as shown in Figure 1, with three horizontal and five vertical positions per swath. Each panel with 15 filter papers constituted one replicate test, resulting in four replicates per insecticide product.

#### *2.4. Determining Spray Deposit Using a Fluorescent Tracer*

Filter papers sprayed with a fluorescent tracer were removed from Petri dishes using tweezers and placed, with minimal handling, into individual labelled ziplock bags. 100 mL of 10% NaOH *v*/*v* solution was added to each bag and subsequently stored in the dark for 60 min. Each bag was agitated thoroughly for approximately 1 min to mix the solution and ensure all fluorescein had been extracted from the filter papers. Then, an aliquot of the sample was added to a glass test tube. Fluorescence of each sample was measured using a Sequoia–Turner Model 450 Fluorometer and fluorescein filter set with excitation at 490 nm and emission at 515 nm. The fluorimeter was calibrated before each replicate against known concentrations of fluorescein applied to filter papers. Before analysing samples, a single concentration standard was used to check for any drift in the fluorescence measured over time.

Fluorescence heat maps were generated using Microsoft Excel as a proxy for dosage applied. A three-colour format was used, with the lowest recorded fluorescence value as the minimum (yellow), the second highest recorded fluorescence value as the maximum (red), and a mid-point at 50% of the difference between the high and low points (blue) when recording concentration.

#### *2.5. Insecticide Sprayed Filter Papers*

Sprayed filter papers were left to dry in the experimental huts for a minimum of 24 h, before they were wrapped individually in aluminium foil and stored at 5 ± 3 ◦C. The concentration of active ingredient on the filter papers was determined using highperformance liquid chromatography (HPLC). Samples were extracted from the filter papers at KCMUCo, and dried extracts were shipped to the Liverpool School of Tropical Medicine (LSTM) for HPLC analysis. The HPLC analysis was performed on a Dionex UltiMate 3000 comprising of an autosampler, quaternary pump, and variable wavelength detector. Chromeleon 7.2 SR4 software was used for peak analysis.

Prior to extraction, 12 circles were punched out of the filter paper using a 0.635 cm radius (½ inch diameter) hole punch, to have a consistent exact surface area of 15.201 cm<sup>2</sup> per disc to extract the sample from. A volume of 5 mL of a 100 μg/mL DCP in acetone solution was pipetted into a glass tube containing each filter paper sample and sonicated for 15 min using an Ultrawave U500H Ultrasonic Cleaning Bath (4.5 litre). Then, 1 mL of the sonicated sample was transferred to a new vial and left to evaporate until dry.

Samples were re-suspended using 1 mL of HPLC grade acetonitrile, and vortexed for at least 1 min at 2500–3000 rpm. Subsequently, samples were centrifuged (Eppendorf Centrifuge 5430) at 13,000 rpm for 20 min, and directly afterwards 100 μL of each sample was pipetted into individual HPLC vials. A 250 mm × 4.6 mm HPLC column (Thermo Scientific Hypersil Gold C18) was used for all active ingredients, using an injection volume of 20 μL. HPLC methods were tailored for each active ingredient as detailed in Table 1.


**Table 1.** HPLC methodology per active ingredient in the samples.
