A micro-air corona discharge, which is one of the plasmas successfully used for gene transfection in terms of high transfection and cell viability rates, is characterized by optical emission spectroscopy. This non-equilibrium low temperature plasma is generated from the tip of a pulsed high voltage micro-tube (0.2 mm inner diameter and 0.7 mm for outer diameter) placed 2 mm in front of a petri dish containing deionized water and set on a grounded copper plate. The electron temperature, equal to about 6.75 eV near the electrode tip and decreased down to 3.4 eV near the plate, has been estimated, with an error bar of about 30%, from an interesting approach based on the experimental ratio of the closest nitrogen emission spectra of + N 2 (FNS) at 391.4 nm and N 2 (SPS) at 394.3 nm. This is based on one hand on a balance equation between creations and losses of the excited upper levels of these two UV spectra and on the other hand on the electron impact rates of the creation of these upper levels calculated from solution of the multi-term Boltzmann equation. Then using the measured Hα spectrum, electron density n e has been estimated from Stark broadening versus the inter-electrode position with an average error bar of about 50%. n e ≈ 1 × 10 15 cm −3 is near the tip coherent with the usual magnitude of electron density in the streamer head developed near the tip of the corona discharges. Rotational temperatures, estimated from comparison of synthetic and experimental spectra of OH(A − X ), + N 2 (FNS) at 391.4 nm, and N 2 (SPS) at 337 nm are respectively equal to 2350 K, 2000 K and 700 K in the gap space. This clearly underlines a thermal non-equilibrium of the corresponding excited species generated inside the thin streamer filaments. But, due to the high dilution of these species in the background gas, these high rotational temperatures do not affect the mean gas temperature that remains close to 300 K. Then, + N 2 (FNS) for (0,0) and (1,1) head bands spectra at 391.4 nm and 388.4 nm allowed estimation of the vibrational temperature T vib from around 3000 K near the tip electrode up to about 6500 K near the plate. Last, the spatial variation along the z axis of the nitrogen ion density has been determined, with an error bar of about 50%, from the relative intensities of the same close wavelength spectra (N 2 (SPS) at 394.3 nm and + N 2 (FNS) at 391.4 nm) when assuming a prior calibration of 10 15 cm −3 at z = 0 mm taken from literature streamer dynamics simulations. The present experimental plasma characteristics are used to better understand the mechanisms and the processes involved during plasma gene transfections in a Monte Carlo poration model previously developed to simulate the membrane permeabilization and pore formation when the cells are impacted by the present micro-air plasma fluxes.

A Monte Carlo poration model is developed to simulate formation pores of few nanometer of width through cell multilayer membranes when impacted by air plasma species. Each species is assumed as a super-particle involving a large particle number. It is assumed a simple membrane model superposing four layers of phospholipids and proteins. Each layer is composed by super-sites subjected to plasma macro-processes (recombination, reflection, activation, opening). Electrons play a main role on super-site activations and openings but ions and radicals have also their own role on pore formation. Judicious choices probabilities data of macroprocesses lead to sizes of pores compatible with gene transfection. Present model, validated by comparisons with measured rates of transfected and surviving cells, is a tool for understanding the mechanisms of gene transfection.

No abstract