Quantitative Evaluation of SERS-Active Ag Film Nanostructure by Atomic Force Microscopy

The reliability of an image analysis algorithm for atomic force microscopy (AFM) of thin metal films was evaluated by comparison with manual analysis of images and transmission electron micrographs of Ag films deposited on Formvar-coated Cu grids. In order to extract quantitative nanostructural information using the algorithm discussed herein, the optimal fitting parameters were found to be low-pass filtering to reject high-frequency noise, a 5 × 5 point grid for identification of particle maxima, and a linear least-squares fit to a hemispheroidal model of particle shape. Metal particle dimensions were defined from the height and radius of the hemispheroid fit. Due to the close spacing of particles in these Ag films, tip geometry causes the greatest error in the height measurements, rather than width measurements. In addition, the effect of scanning parameters such as scan rate and size, applied load, and humidity on particle count and dimensions was examined. Increasing the scan rate reduced the number of resolvable Ag particles, decreased the apparent particle height, and increased the apparent particle radius. Under conditions of low capillary force, a net repulsive force of ∼19 nN resulted in subtle tip-induced changes in the Ag surface morphology. The Ag film surface was damaged at a net repulsive force of ∼23 nN. At slow scan rates, the moisture layer did not significantly affect the quality of the AFM images obtained over a broad relative humidity range. Finally, the Ag surface structure was found to be very homogeneous over a relatively large area.