Abstract
At low copy number, sequence detection by polymerase chain reaction (PCR) requires up to 30 cycles (amplification by a factor of 109) to produce a reliably detectable concentration of fluorescently-labelled amplicons. The cycle number and hence detection time is determined by the analytical sensitivity of the detector. Hybridisation of complementary DNA strands to oligonucleotide-modified conducting polymer electrodes yields an increase in the charge transfer resistance for the ferri-ferrocyanide redox couple. Sensors using this technology for e-PCR offer a label-free method with detector sensitivity in the pM range, potentially decreasing the required cycle number from 30 to 10 and offering a much simplified instrument construction. We demonstrate sensors using screen-printed carbon electrodes modified with a conducting polymer formed from a monomer pre-functionalised with complementary oligonucleotide. Off-chip pre-functionalisation of the conducting polymer precursor is a key step towards practical manufacture and the method is potentially a general one for sensors which require a capture probe-functionalised surface. We demonstrate reliable sensitivity of the interfacial resistance change at the pM scale for short (20-mer) sequences and at the aM scale for bacterial lysate, with dynamic range extending to μM scale and response time-scale 5 min. Donnan exclusion of the redox couple from the surface, as previously proposed, seems unlikely as a mechanism for such ultra-high sensitivity. We demonstrate that the most important element in the response at the lowest concentrations is due to variation of an electrical resistance within the polymer film. We develop a mechanism based on repulsion from the solution interface of dopant anions and attraction towards and trapping at the interface of radical cations (polarons) by the charge associated with surface-bound DNA. With results for >160 single-use sensors, we formulate a response model based on percolation within a random resistor network and highlight challenges for large-scale manufacture of such sensors. We propose a PCR device concept for rapid use at point-of-sampling.
Content

Supplementary material

VBA code
VBA code for calculation of resistance of a random resistor network on a cubic lattice
spreadsheet implementing calculations
Spreadsheet implementing random resistor calculation. Contains description and embedded code but is not designed to be user-friendly. The authors have used best efforts to track and remove bugs but do not guarantee that the code is bug-free

Supporting Information
Contents
1. Determination of bias potential for EIS measurement
2. Calculation of the real surface area of SPCE
3. Calculation of the thickness of deposited poly(PyPhEG-co-PyPhON) film
4. EIS of bare SPCE in the redox couple solution
5. EIS data of poly(PyPhEG-co-PyPhON) films with different polymerisation time.
6. EIS data after incubation in bacterial lysate
7. Additional files
a. VBA code for calculation of resistance of a random resistor network on a cubic lattice
b. Excel spreadsheet with example results