Here we present initial results towards the fabrication of suspended structures between photo-patterned high aspect ratio SU-8 posts.  Preliminary results illustrated in figure 1a and bshow that carbon wires of diameter down to 200 nm can be formed by carbonizing suspended SU-8 bridges. SU-8 is an epoxy-based negative photoresist, commonly used as carbon precursor in carbon MEMS technology.  By controlling the process parameters in the photolithography of different shapes with varying gaps between them, SU-8 structures including suspended wires and bridges, can be manufactured, carbonized and used in application such as biosensors.  The reproducibility of the future biosensor when following the proposed fabrication process is expected to be improved since high quality carbon precursors are used, photolithography can be quite reproducible and most importantly, the complete structure, anchor posts and suspended features, are carbonized together.  We use photolithography to fabricate different high aspect ratio SU-8 shapes of varying size, height and spacing using different exposure doses during the process.  The goal of this work is to characterize the formation of suspended structures between these geometries depending on the photolithography process, shapes and dimensional parameters.   

3D geometries featuring hexagonal, diamond, squared, triangular and circular cross section were fabricated.  Their characteristic dimension, side length in the case of polygons or diameter for circles, varied among the values 10, 20, 30, 40, 80, and 160 µm.  Arrays of 5 by 5 were made for each of the shapes.  The gap between individual features varied among the values 5, 10, 15, 20, 25, 30, and 45 µm.  For a given exposure power, the time of exposure was varied from 10 to 40 s in increments of 10 to achieve a varying exposure dose.  After fabrication, the structures were analyzed using an optical microscope.   

Initial results are summarized in Figure 1 for different cross sections with varying size (x-axis) and gap (y-axis) in between them.  A specific color is used for a given cross section to facilitate interpretation of the graphs.  Figure 1c presents results for shapes closest to each other by a point, as in an array of triangles.  The analysis of shapes with a flat side in between them, as in an array of squares, is shown in figure 1d.  A number of cases are analyzed here: No Bridges (NB), Spikes (S), Broken Bridges (BB), Stable Bridges (SB), Merged posts (M), Irregular posts (IR); depending on the nature of the joint between them or the absence of any contact. The points marked as irregular are those which are not straight or completely missing. Here we are interested on the formation of suspended wires or bridges.  The results show that the formation of suspended bridges follows predictable trends.  As the distance of the posts decreases, bridges form more easily. This can be observed looking at a single column in any of the graphs, and noticing that the type of connection follows the sequence NB-S-BB-SB-M, not necessarily featuring all the steps.  The gap for which the bridge happens, depends on the size and the nature of the 3D shape.  Comparing Figure 1a and 1b, it can be observed that bridges are formed for wider gaps in the case when shapes face each other on their sides rather than on a point.  We hypothesize the reason behind this phenomenon is that shapes facing each other on their long sides have more material interaction than when facing by single points.  Among the shapes facing each other by a point (figure 1c), the hexagonal is the best because it features a softer point, or a broader angle, that leads to improved interaction between neighboring shapes.    Circles are somewhere in between the behavior of the other the interface types detailed above as they feature gradually curved interfaces.

Ongoing work is on characterizing SU-8 posts of different height to assess the impact of shape height in the formation of wires and bridges.  The time of exposure is also being characterized.  The next step will be carbonizing these structures and perform a similar characterization to correlate a photolithography process with the dimensions of a carbon wire that can be used for biosensing. 

Figure 1: (a) Carbonized suspended bridge and wire; (b) SU-8 posts array showing structures connected by a bridge. Characterization results when (c) SU-8 posts are facing by their angular points; (d) SU-8 posts are facing by their sides and circular cross section.