A plasmonic gold nanofilm-based microfluidic chip for rapid and inexpensive droplet-based photonic PCR

Abstract

Polymerase chain reaction (PCR) is a powerful tool for nucleic acid amplification and quantification. However, long thermocycling time is a major limitation of the commercial PCR devices in the point-of-care (POC). Herein, we have developed a rapid droplet-based photonic PCR (dpPCR) system, including a gold (Au) nanofilm-based microfluidic chip and a plasmonic photothermal cycler. The chip is fabricated by adding mineral oil to uncured polydimethylsiloxane (PDMS) to suppress droplet evaporation in PDMS microfluidic chips during PCR thermocycling. A PDMS to gold bonding technique using a double-sided adhesive tape is applied to enhance the bonding strength between the oil-added PDMS and the gold nanofilm. Moreover, the gold nanofilm excited by two light-emitting diodes (LEDs) from the top and bottom sides of the chip provides fast heating of the PCR sample to 230 °C within 100 s. Such a design enables 30 thermal cycles from 60 to 95 °C within 13 min with the average heating and cooling rates of 7.37 ± 0.27 °C/s and 1.91 ± 0.03 °C/s, respectively. The experimental results demonstrate successful PCR amplification of the alcohol oxidase (AOX) gene using the rapid plasmonic photothermal cycler and exhibit the great performance of the microfluidic chip for droplet-based PCR.

Figure 1

(a) Layout of the manufactured microfluidic chip for droplet generation and trapping: two syringe pumps are applied at the inlets, which control sample and oil flow rates to generate monodisperse droplets. (b) Illustration of the layered chip structure composed of a patterned PDMS layer, a partially cured PDMS thin layer, a double-sided tape layer, an AuNF layer, and a glass substrate layer.

Figure 2

(a) Schematic illustration of the fundamental principle of plasmonic photothermal light-to-heat conversion for droplet-based photonic PCR (dpPCR): when a light is turned on, fast heating of sample droplets can be achieved by fast heat diffusion of hot electrons throughout the AuNF. When a light is turned off, the heat dissipation through the AuNF can occur, leading to the cooling of the heated droplets. After thermal cycling, the success of the amplification can be identified by analyzing the fluorescence intensity of droplets. (b) Circuit diagram of the plasmonic photothermal cycler. (c) An enlarged section of (b), indicating the position of the LED, heatsink, chips, and thermocouple.

Figure 3

Comparison of droplets behavior in oil-blended PDMS chip with (a) 0, (b) 3, and (c) 7% mineral oil after amplification (bright-field microscopy images).

Figure 4

(a) Schematic depiction of the photonic PCR thermal cycling, consisting of three successive temperatures for denaturation, annealing, and extension, using the Au nanofilm (AuNF) excited by the LEDs. Light that encounters AuNF (I₀ and I₁) can be reflected (R₀ and R₁), transmitted (T₀ and T₁), or absorbed (A₀ and A₁). (b) Comparison of temperature profiles of the parallel reaction chamber (PRC) between position P and position M to ensure that the temperature ramping rate of both chambers is the same. Inset shows a schematic of positions P and M. (c) Comparison of heating and cooling rates of the sample using the constructed chip with (stable bonding) and without (unstable bonding) tape layer.

Figure 5

(a) Temperature profiles of the single-sided plasmonic heating mechanism (ssPHM) and the dual-sided plasmonic heating mechanism (dsPHM) for the blue light. (b) Temperature profiles of the sample in the ssPHM and the dsPHM for denaturation-to-annealing ramp and vice versa (30 PCR thermal cycles, 95 °C for 0 s and 60 °C for 1 s). (c) Heating and cooling rates obtained from the dsPHM thermal cycling (average heating and cooling rates of 7.37 ± 0.27 °C/s and 1.91 ± 0.03 °C/s, respectively). Solid lines indicate average values. Dashed lines are placed one standard deviation away from averages.

Figure 6

Illustration of the plasmonic photothermal cycling results with the dsPHM. (a) PCR thermal profile of a complete PCR reaction. The thermocycling program settings are 95 °C pre-denaturation for 8 min followed by 30 cycles of 95 °C (denaturation), 60 °C (annealing), and 72 °C (extension) for 15 s, 20 s, and 20 s, respectively. (b) Evaluation of temperature accuracy and stability. The average values with standard deviation at 95 °C, 60 °C and 72 °C are 94.99 ± 0.41 °C, 60.02 ± 0.37 °C and 72.01 ± 0.39 °C, respectively.

Figure 7

Representative fluorescence microscopy images of the droplets after 30 thermal cycles for (a) positive control reaction and (b) no template control (NTC) reaction. (c) Normalized fluorescence intensity of the positive (right) and negative (left) end-point signal. The results of the positive droplets were normalized to that of the NTC droplets. The average green pixel intensity of the positive droplets was significantly higher than that of the NTC ones. The bottom and top hinges of the boxplot show the first and third quartiles of a sample of 300 randomly selected droplets represented in each box, respectively. The medians are depicted by the horizontal lines in the middle, and the whiskers indicate min and max values. Outliers and the p-value are also shown. The p-value is calculated using two-tailed t-test for independent samples.

Figure 8

Histograms of fluorescence intensity values for over 1000 droplets at (a) λ = 0.2, (b) λ = 0.5, and (c) λ = 1.5.