Temperature distribution effects on micro-CFPCR performance

Abstract

Continuous flow polymerase chain reactors (CFPCRs) are BioMEMS devices that offer unique capabilities for the ultra-fast amplification of target DNA fragments using repeated thermal cycling, typically over the following temperature ranges: 90°C–95°C for denaturation, 50°C–70°C for renaturation, and 70°C–75°C for extension. In CFPCR, DNA cocktail is pumped through the constant temperature zones and reaches thermal equilibrium with the channel walls quickly due to its low thermal capacitance. In previous work, a polycarbonate CFPCR was designed with microchannels 150 μm deep, 50 μm wide, and 1.78 m long—including preheating and post-heating zones, fabricated with LIGA, and demonstrated. The high thermal resistance of the polycarbonate led to a high temperature gradient in the micro-device at steady-state and was partly responsible for the low amplification yield. Several steps were taken to ensure that there were three discrete, uniform temperature zones on the polycarbonate CFPCR device including: reducing the thickness of the CFPCR substrate to decrease thermal capacitance, using copper plates as heating elements to ensure a uniform temperature input, and making grooves between temperature zones to increase the resistance to lateral heat conduction between zones. Finite element analyses (FEA) were used to evaluate the macro temperature distribution in the CFPCR device and the micro temperature distribution along a single microchannel. At steady-state, the simulated CFPCR device had three discrete temperature zones, each with a uniform temperature distribution with a variation of ±0.3°C. An infrared (IR) camera was used to measure the steady-state temperature distribution in the prototype CFPCR and validated the simulation results. The temperature distributions along a microchannel at flow velocities from 0 mm/s to 6 mm/s were used to estimate the resulting temperatures of the DNA reagents in a single microchannel. A 500 bp DNA fragment was generated from a bacteriophage λ-DNA target using 20 cycles of PCR. The amplification efficiencies compared to a commercial thermal cycler were 72.7% (2 mm/s), 44% (3 mm/s), and 29.4% (4 mm/s). The amplification efficiency with the modified CFPCR device increased by 363% at 2 mm/s and 440% at 3 mm/s compared to amplification obtained using a CFPCR device with the same fluidic layout, (Hashimoto et al., Lab Chip 4:638, 2004) strictly due to the improved temperature distribution.