Transient and steady-state readout of nanowire gas sensors in the presence of low-frequency noise

[Display omitted] •Comparison of transient and steady-state gas sensing in the presence of 1/f noise.•Theoretical analysis of the SNR of chemical sensing in the presence of 1/f noise.•Experimental characterization of transient response of carbon nanotube NO2 sensors.•Criteria are derived for when transient or steady-state sensing is most suitable. Nanowire sensors show great promise in a variety of sensing applications due to their potential for high sensitivities. Practical nanowire sensor systems, however, are often limited by low-frequency, 1/f noise. This work presents theoretical and experimental results comparing the performance metrics of sensing schemes using transient and steady-state parameters in the presence of 1/f noise. Criteria are derived for when the considered transient or steady-state sensing schemes will have a better signal-to-noise ratio (SNR). The theoretical results for the SNR of these sensing schemes are applied to experimental data from carbon nanotube NO2 sensors. These data and theoretical results demonstrate that due to the Langmuir binding behavior of the sensor-analyte system, sensing using the considered transient parameters increases linearity and decreases response time relative to steady-state sensing. Noise analysis further shows that with current devices, transient sensing has a lower SNR relative to steady-state sensing, however this may change if functionalization is considered. The use of transient parameters also has the potential to reduce sensor drift due to 1/f noise, improving system stability. In addition to providing useful considerations towards the design of carbon nanotube gas sensors, these results are relevant towards understanding the SNR of other chemical and biological sensors limited by 1/f noise.