The samples were illuminated using a 405-nm laser diode source (Coherent Cube) using an inverted microscope fitted with a TIRF illuminator (Nikon) and a × 100 oil immersion objective (NA 1.49, Nikon). The laser light was filtered using dichroic (Z405rdc Chroma) and emission filters (ET420LP Chroma). Single molecule fluorescence was collected using an EMCCD camera (Photometric Evolve 512). Each frame had a 100 ms exposure time, with ∼30 ms of dead time between acquisitions. SR maps were constructed from a minimum of 60,000 images and, on average, contain ∼4,000 successful localization points after suitable filtering. In order to account for sample drift scattered laser light from the sample was reflected by the dichroic mirror and collected via a second camera (QICam). A disk in each array of structures was used as a reference point and the scattered laser light was localized and used to correct the sample position.
During the measurement process, the illuminating laser was kept at low power (on the order of ∼10−1 W cm−2) to ensure that in the presence of an enhanced EM field around our plasmonic structures, our dyes continued to operate in a linear response regime. The reader should note that this is one to several orders of magnitude less than conventional super-resolution microscopy techniques and as a result unenhanced molecules at the glass/sample interface are not observed.
During the measurement process, the illuminating laser was kept at low power (on the order of ∼10−1 W cm−2) to ensure that in the presence of an enhanced EM field around our plasmonic structures, our dyes continued to operate in a linear response regime. The reader should note that this is one to several orders of magnitude less than conventional super-resolution microscopy techniques and as a result unenhanced molecules at the glass/sample interface are not observed.