Droplet-based chemistry on a programmable micro-chip

Abstract

We describe the manipulation of aqueous droplets in an immiscible, low-permittivity suspending medium. Such droplets may serve as carriers for not only air- and water-borne samples, contaminants, chemical reagents, viral and gene products, and cells, but also the reagents to process and characterise these samples. We present proofs-of-concept for droplet manipulation through dielectrophoresis by: (1). moving droplets on a two-dimensional array of electrodes, (2). achieving dielectrically-activated droplet injection, (3). fusing and reacting droplets, and (4). conducting a basic biological assay through a combination of these steps. A long-term goal of this research is to provide a platform fluidic processor technology that can form the core of versatile, automated, micro-scale devices to perform chemical and biological assays at or near the point of care, which will increase the availability of modern medicine to people who do not have ready access to modern medical institutions, and decrease the cost and delays associated with that lack of access.

Fig. 1

Overview of an 8 × 8 substrate. Electrodes of the programmable fluidic processor (PFP) are controlled via gold leads individually connected to pads along the upper and lower edges of the substrate as shown. In order to access and view the reaction surface, experiments were carried out using an open chamber formed with an o-ring epoxied to the glass substrate.

Fig. 2

Experimental apparatus. The PFP was exercised on a motorised microscope stage. A programmable function generator and AC amplifier generated electrical signals controlled from the laptop computer and microcontroller board. A manual syringe pump delivered fluid to the PFP and controlled the hydrostatic pressure within the micropipette injector. Experiments were monitored by the CCD camera system and recorded on videotape for analysis.

Fig. 3

Multiple droplet movement. Droplet motion and fusion on an 8 × 8 PFP with 30 μm electrodes. Frame (a) shows four droplets of 3 × H₂O, three of which contain 0.065-0.18 nl each, have been manoeuvred into a line on the 2nd column of electrodes, and a 0.38 nl droplet that has been manoeuvred onto the 5th column. In frame (b) the middle droplet on the 2nd column has been manoeuvred toward the large droplet with which it has spontaneously fused. In frames (c) and (d) the upper and lower droplets have been fused, and in frame (e) the cumulative 0.69 nl droplet has been manoeuvred to the 8th column.

Fig. 4

Metering of small droplet into larger droplet. The image, captured from videotape, shows an 18 μm diameter (3 pl) droplet in the process of being metered into an existing 42 μm diameter (39 pl) droplet. The operational parameters for this experiment: V_DEP = 120_p-p, P/H = 0.72, f_DEP = 150 kHz, d = 2.6 μm, Z = 28 μm, and e = 30 μm.

Fig. 5

Spontaneous droplet fusion with rapid mixing of contents. Frame (a) shows two aqueous droplets suspended in liquid hydrocarbon on a 2 × 8 PFP. Illumination is via a mercury lamp; viewing is through a Texas Red fluorescence filter set. The 87 μm diameter, (0.34 nl), fluorescent droplet on the left contains 53 mM fluorescein in a 16 mM HCl solution. The 91 μm diameter, (0.39 nl), non-fluorescent transparent droplet on the right contains a 16 mM NaOH solution. Frame (b) shows the two droplets after fusion with the fluorescent/non-fluorescent diffusion front moving across the combined droplet.

Fig. 6

Enhanced image of a set of nine droplets distributed on an 8 × 8 PFP. The droplets contained a BSA concentration that increased geometrically from lower right to upper left. The droplet at the lower right contained no added protein and served as a control for determining the background fluorescence. Electrode features are visible as bright areas through the more fluorescent droplets.

Fig. 7

Combined droplet injection data normalised by injector-electrode distance and electrode size. The droplet diameters have been normalised by the sum of Z and e. The square of the applied DEP voltage is scaled by Z and normalised by the hydrostatic pressure as a ratio to the holdoff pressure. In the figure legend, d refers to the injector orifice diameter in μm, e to the electrode edge length in μm, and n to the number of samples in each data set.

Fig. 8

Change in energy vs. radius as a result of droplet fusion: The fractional change in energy, ∂E, of a two-droplet system as a result of the fusion of the droplets is graphed against the change in radius, ∂r, of the larger of the initial droplets. At ∂r = 1 there is no second droplet to fuse, and hence, no change in energy. Fusion of the droplets shown in Fig. 4 results in ∂r = 1.23 and ∂E = 0.83 (dashed line).

Fig. 9

Protein concentration vs. fluorescence intensity: The figure graphs the relationship between the BSA concentration and the integrated fluorescence intensity for the eight droplets shown in Fig. 6 (squares) plus seven additional droplets from a previous experiment (circles). A best-fit linear curve is superimposed over the data points graphed on a log-log scale. The minimum observed fluorescence corresponded to a protein concentration of 0.266 μg ml^-1, (3.9 nM). From the curve fit it is seen that the OPA assay is linear over at least three decades of protein concentration.