Extended Data Figure 1: Impact of channel disruption on channel currents.

Room-temperature transient currents were measured on a DTDCTB-(10 nm C₆₀)-mixed device at L = 3 mm before and after a series of razor blade cuts was made to the organic layers, as shown in the inset. The peak height of the current pulse was greatly reduced and the peak arrival time was delayed for devices with a ‘partial cut’ that was transverse to, and spanned the width of, the ITO anode between the illumination position and the silver cathode, compared with the pristine, uncut device. Charge-diffusion simulations were performed for both geometries, where the only difference was a blocking boundary condition at the position of the partial cut. We find that charge diffusion around the cut accounts for the differences between the cut and uncut device transients, as demonstrated by the agreement between fits (lines) and the data. The partial cut was also extended such that there was no continuous organic path between the illumination position and the cathode (a ‘full cut’ device). This eliminated the response except for a residual current at time t < 200 ms arising from scattered light absorbed in the organic layers between the cathode and cut. This effect was observed in all devices exhibiting channel currents.