In this study, an AgNMPA MEF substrate was proposed to improve the SBR of the MEF substrate. To examine the effects of surface morphology on fluorescence enhancement and background noise reduction, we fabricated the GLAD Ag nanorod structures on the imprinted micropost array at varying deposition angles of 81°, 85°, and 89° with a deposition thicknesses of 500 nm. As the reference substrates for the comparison, a bare glass substrate and GLAD MEF substrate were also prepared. For the qualitative analysis of the performance of the proposed AgNMPA MEF substrate, the measured fluorescence microarray scanning images with a same measurement conditions were compared. Figure 6 shows the measured fluorescence images of the substrates after the MPIF-1 antibody-antigen reaction; (a, b) a bare glass substrate with different brightness condition, (c) GLAD MEF substrate with a Ag nanorods with a thickness of 500 nm and deposition angle of 89°, and AgNMPA MEF substrate with different deposition angles of (d) 81°, (e) 85°, and (f) 89°. For the clear comparison, the fluorescence microarray images of Fig. 6(b-f, except a) were measured at the same PMT gain, contrast, and brightness (selected conditions), where we could avoid the saturation of signals on MEF substrates. Although the fluorescence signals of spots on bare glass substrate were slightly detectable at higher brightness condition as shown in Fig. 6(a), they were not clearly detectable at the selected conditions as shown in Fig. 6(b). However, the spots were clearly observed on GLAD MEF substrate due to MEF effects as shown in Fig. 6 (c~f). Although the high fluorescence signals from spotting area were observed in GLAD substrate we could observe three issues in Fig. 6 (c). First, the shapes of spots were irregular because of the non-uniform spreading of spotted antibody due to the capillary force of GLAD Ag nanostructure. Second, the background signals were higher than the bare glass substrate because the fluorescence signal on non-spotting area was also enhanced by the MEF effect. Third, the nonspecific binding were more severe due to the enhanced surface area of Ag nanostructure on the non-spotting area.

Figure 6 (d-f) shows the fluorescence images of AgNMPA MEF substrates with different deposition angles. In each image, 5 x 5 microposts were shown and 3 x 3 fluorescence signal from the spotting area on micropost were observed because we spotted the antibody every 2 microposts as shown in Fig. 5. The fluorescence signal from the antibody spotting area of AgNMPA MEF substrate were similar to the signal on GLAD MEF substrate due to the MEF effects of Ag nanorods. For the irregular spot shape issue, we could obtain clear circular spot shape on AgNMPA substrates because the spread of the spotted antibody on the top of micropost was restricted in the post area due to the surface tension. For the high background signal issue, we could dramatically reduce the background signal from bottom of micropost on AgNMPA MEF substrates, especially on the substrate with the largest deposition angle of 89° (Fig. 6(f)). Since the fluorescence signal from the non-spotting area on micropost of AgNMPA MEF substrates (Fig. 6(d-f)) were similar to the background signal from non-spotting area of GLAD MEF substrate (Fig. 6(c)), the reduction of background signal from the bottom of micropost were due to the less developed Ag nanorod structure and out-of-focus effect. The slightly increased fluorescence signal was observed at the inter-post area of AgNMPA MEF substrate with a deposition angle of 81° (Fig. 6(d)), because the relatively well developed GLAD Ag nanorods were formed at the inter-post area where the shadowing effect of micropost was minimized during GLAD process. However, this increase was not shown on the AgNMPA MEF substate with a deposition angle of 89°. With the increase in the deposition angle, the background noise signal (signal from the bottom of micropost) and nonspecific bindings tended to decrease, because the shadowing effects were increased as increasing deposition angle.