Figure 1: Schematic of a force spectroscopy probe membrane with self-sensing and self-actuation capabilities coupled with an AFM cantilever in a typical force spectroscopy experiment setup.

Figure 2: Displacement noise spectrum of the 200 μm diameter membrane probe.

Due to its low elastic modulus parylene, a biocompatible polymer makes the fabrication of softer membranes possible. Moreover the hydrophilicity of parylene can be altered by oxygen plasma treatment, a key feature for their functionalization with biomolecules. These probes, functionalized with biotinylayted BSA and tested against streptavidin-coated cantilevers. While biotin could interact with streptavidin, BSA helped reduce nonspecific binding by blocking all exposed surfaces. Figure 3 shows the simultaneous data captured by both sensors.

Figure 3: Comparison of force data simultaneously captured by a coupled AFM cantilever and a membrane probe functionalized by biotin/streptavidin.

The stiffness of the probes can be tuned using the parallel-plate type integrated electrostatic actuator which moves the membrane. An important implication of adjustable spring constant for force spectroscopy is that, a small, relatively stiff membrane with smaller coefficient of viscous damping can be used for low force noise experiments. The large spring constant of such a membrane can be reduced electrically to obtain a larger signal for unit force, i.e. higher detection sensitivity, while still being limited by thermal mechanical noise dominated by the interaction of the probe structure with the surrounding liquid. A proof of principle experiment showing sensitivity improvement was carried out by taking force curves with a single probe at two different bias points as shown in figure 4.

Figure 4  (a) Measured and predicted variation of the spring constant of the membrane with increasing DC bias voltage. The inset shows the measured and predicted membrane displacement as a function of DC bias voltage. (b) Force curves obtained by bringing the AFM tip in and out of contact with the probe membrane biased at 35V. The top curve recorded from the AFM cantilever shows a peak force of 400 nN. The bottom curve is recorded from the membrane with a spring constant of 22 N/m. (c) The peak force applied by the cantilever is kept constant at 400 nN as shown in top trace. The bottom trace is from the same membrane with an electrically reduced spring constant of 11 N/m.

The membrane architecture allows array formation for parallel operation. The parallel AFM setup we have developed uses the array of membranes that can be coupled to an array of cantilevers to perform parallel experiments. In this setup, the membranes are used to actuate the cantilevers that serve as force sensors. It is possible to control the force on the individual cantilevers precisely since they are surface actuated in this scheme. Figure 5 shows simultaneous reading we obtained from two different cantilevers  to demonstrate the feasibility of the method.

Figure 5. Cantilever 1 and 8 are engaged to the membrane probe and the substrate, respectively. The drive signals applied to the membrane probe and the piezo actuator that carries the microcantilevers are shown. The readout signals obtained from cantilever-8 shows that this cantilever follows the piezo movement whereas cantilever-1 movement is a combination of piezo and membrane movements.

Biological force studies using AFM typically involve coating microcantilevers and an apposing surface with different biomolecules of interest and retracting the cantilevers at set speeds that span a wide range. At very high speeds of pulling, the cantilevers are subjected to viscous drag forces, whose magnitudes often match or exceed the unbinding forces between the biomolecules. This introduces ambiguity in the forces being measured. We are currently modeling the hydrodynamic effects on cantilevers and membrane probes using FLUENT (CFD software) under conditions that simulate high speed pulling. Preliminary investigations have suggested that membrane-based probes are much less susceptible to drag forces than cantilevers for the same speed of pulling.

H Torun, O Finkler, F L Degertekin, "Athermalization in atomic force microscope based force spectroscopy using matched microstructure coupling", Review of Scientific Instruments, 80, 076103, 2009

H Torun, K K Sarangapani, F L Degertekin, “Fabrication and Characterization of Micromachined Active Probes with Polymer Membranes for Biomolecular Force Spectroscopy”, under review, JMEMS, 2008

Opto–mechanical thermal detector with optical readout is designed and realized for infrared (IR) imaging. The detector pixels are membranes that are connected to bimaterial legs, which are connected to a substrate through thermal isolation legs. The conversion of IR radiation into temperature difference causes deflection along bimaterial legs and the deflected membranes as a result of this phenomenon are detected by optical means. Since the readout method proposed is by optical means, no electrically (so thermally) conductive paths are needed on the pixels. Therefore the thermally isolated pixels offer higher sensitivity compared to cooled detectors. The readout method proposed for the project is to detect the deflection of membranes by pixel level micro-interferometers with diffraction gratings. That architecture offers operation at shot noise level of the coherent light source for the readout illumination. By this way, it is possible to prevent the FPA be readout noise limited. It is shown that noise equivalent temperature difference (NETD) for our design is <10 mK including the readout noise sources.

Figure 6: (a) Fabricated thermal detector pixels (b) FEM result of thermomechanical deflection for the selected pixel geometry (c) Analytical calculation for the deflection as a function of layer thicknesses per 1K temperature difference.

M F Toy, O Ferhanoglu, H Torun, H Urey, “Uncooled Infrared Thermomechanical Detector Array: Design, Fabrication and Testing”, to appear, Sensors and Actuators A, 2008