High surface-area carbon microcantilevers

Abstract

Microscale porous carbon mechanical resonators were formed using carbon nanotube templated microfabrication. These cantilever resonators exhibited nanoscale porosity resulting in a high surface area to volume ratio which could enable sensitive analyte detection in air. These resonators were shown to be mechanically robust and the porosity could be controllably varied resulting in densities from 10² to 10³ kg m^-3, with pore diameters on the order of hundreds of nanometers. Cantilevers with lengths ranging from 500 μm to 5 mm were clamped in a fixture for mechanical resonance testing where quality factors from 10² to 10³ were observed at atmospheric pressure in air.

Fig. 1. Microcantilever fabrication and measurement process consists of (a) photolithographic patterning of a 4 nm iron catalyst film, (b) vertical carbon nanotube growth from iron catalyst, (c) infiltration of the nanotube forest and release from the substrate, and (d) mounting of the device between a clamp and an AC voltage-driven piezoelectric element followed by deflection measurement by means of a reflected laser.

Fig. 2. Scanning electron micrographs (SEM) of CNT-M microcantilevers. (a) a microcantilever extending from a larger base with the entire device resting on a silicon substrate, (b) microcantilever top surface, and (c) microcantilever side wall. (d) an SEM image of a CNT-M microcantilever that was cross sectioned by focused ion beam (FIB) prior to imaging. The samples in (a), (b), and (d) were infiltrated for 2 minutes, and (c) for 4 minutes.

Fig. 3. Quantitative measurements of infiltrated carbon nanotube forest porosity. (a) Coated carbon nanotube sizes at infiltration times of 1 to 6 minutes as indicated in the image. (b) Coating radius vs. infiltration time. Cylindrical radius of coated carbon nanotubes is shown for various infiltration times. (c) Bulk density for various infiltration times is shown along with a model of the coating process. Error bar extends below and above the mean by one standard deviation of the measured data and includes error estimates in measured mass and height along with sample to sample variation.

Fig. 4. Resonance frequency and quality factor at various pressures. (a) Resonance data of a single cantilever (with dimensions of 4.3 mm long, 1 mm wide, ∼200 microns thick, and infiltrated for 10 min) in three partial air environmental pressures. Each data set is fit to a Lorentzian model shown alongside a model fitted to each case that was used to extract the resonance frequency and quality factor. (b) The value of the damping ratio (ζ = 1/(2Q)) for a given cantilever is shown for a range of environmental gas pressures. It can be seen that at atmospheric pressure approximately half of the total damping is due to fluid damping, the other half being due to clamping and other losses.

Fig. 5. Resonant frequency of a cantilever is shown for relative humidity values of 0, 34, and 100 percent (cantilever dimensions are: 3 mm long, 100 μm wide, 64 μm high, and a carbon infiltration time of 6 minutes). This histogram shows observations taken while humidity values were sequentially changed in a randomized order. Shifts on the order of 1 Hz were seen on top of a base resonant frequency of 1.65 kHz.