A Digital Microfluidics Platform for Loop-Mediated Isothermal Amplification Detection

Digital microfluidics (DMF) arises as the next step in the fast-evolving field of operation platforms for molecular diagnostics. Moreover, isothermal schemes, such as loop-mediated isothermal amplification (LAMP), allow for further simplification of amplification protocols. Integrating DMF with LAMP will be at the core of a new generation of detection devices for effective molecular diagnostics at point-of-care (POC), providing simple, fast, and automated nucleic acid amplification with exceptional integration capabilities. Here, we demonstrate for the first time the role of coupling DMF and LAMP, in a dedicated device that allows straightforward mixing of LAMP reagents and target DNA, as well as optimum temperature control (reaction droplets undergo a temperature variation of just 0.3 °C, for 65 °C at the bottom plate). This device is produced using low-temperature and low-cost production processes, adaptable to disposable and flexible substrates. DMF-LAMP is performed with enhanced sensitivity without compromising reaction efficacy or losing reliability and efficiency, by LAMP-amplifying 0.5 ng/µL of target DNA in just 45 min. Moreover, on-chip LAMP was performed in 1.5 µL, a considerably lower volume than standard bench-top reactions.

. DMF platform integration. a) Overview of the DMF system. An AC signal is amplified 200 times by a high voltage amplifier, and is further processed by a high voltage switching unit, which will enable either an ON or OFF state on each electrode/reservoir of the DMF platform via commands transmitted by an Arduino control board. Finally, a temperature control system assures that the onchip temperature corresponds to the set point temperature. b) Zoom on the region where the DMF chip is placed.

Improved sample input/output method
The improved DMF droplet input/output method was firstly tested, as well as all the fluidic operations. A top plate inlet system was developed for direct input and output of reaction reagents and products, through ports drilled on the top plate, partially overlapping the reservoirs. Figure S3 shows the droplet input process. This test was performed using 1× Bst enzyme buffer with blue dye for easier readout analysis. After sample insertion, the respective inlet reservoir is activated, allowing the droplet to spread. Figure S3. Sequential video frames, evidencing the sample input process.

Device working conditions and droplet speed
Regarding operating conditions, namely working voltage and frequency, this configuration allows the movement of 1× enzyme buffer droplets with 5 kHz and 40 VRMS (standard operating parameters) at reasonable speed, with the possibility of lowering voltages to 8 VRMS (see Figure 3A in the manuscripts). The movement of a solution containing DNA (0.5 ng/µL in LAMP reaction buffer) was also tested and the solution droplets were easily moved, as well as droplets containing all LAMP reagents. Droplet velocity was determined by averaging both head and tail velocities. Velocity measurements were performed by determining the amount of time the droplet took to move from one non-activated electrode to an adjacent activated electrode, considering a total motion distance of 0.83 mm.
Dynamic characterization of the temperature control system From Figure 3B in the manuscripts, used to characterize the temperature control system, as well as the results obtained for temperature measurements at bottom (T1) and top (T5) plates, T3 was determined simply as the average between inner bottom and top plate temperatures (T2 and T4), considering that the system is approximately symmetric, which is deemed a good approximation. Furthermore, the hydrophobic, dielectric and electrode layers were omitted from all calculations and representations, since they are too thin in comparison to the glass substrates, or even the oil, and may be neglected.
The temperature rise in the thin film resistor is quite fast (200 seconds for initial temperature increase to max temperature) and heat is quickly transferred to the top plate. As temperature in the bottom plate approaches the set point temperature, there is a small overshoot, consequence of high speed controller settings, which is not visible in the top plate due to the thermal resistance of the materials between the points where T1 and T5 were measured. The PID controller is also functional, since there are no deviations from the set point temperature after reaching the stationary state.

Heat transfer for a single reaction droplet
In addition to the chip heat transfer analysis and temperature controller optimization, the temperature gradient across a LAMP reaction droplet was theoretically studied. Since on-chip LAMP reactions should occur within the optimal temperature range (60 °C to 65 °C), the droplet thermal gradient was studied for bottom plate temperatures corresponding to both lower and higher temperature limits. Considering first a bottom plate (T1) temperature of 60 °C, to which corresponds a top plate (T5) temperature of 59.1 °C (see Figure S5), it is possible to determine the heat between both locations from an adaptation of Fourier's law, equation eqS1: Where q is the heat, Δx represents the thickness of each layer, A is the overlapping area of all layers involved and k represents the thermal conductivity for each material. Table S3 shows the various parameters of the material layers involved.    Figure S5. Electrophoretic analysis of LAMP products obtained for the study of reaction volume reduction, with 60 min. (lanes 1 to 9) or 90 min. (lanes 10 to 17) reactions. The reaction volumes used in this study are as follows: 20 µL (2, 10), 15 µL (3, 11), 10 µL (4, 12), 5 µL (5, 13), 2.5 µL (6, 14),