Rapid progress in nanofabrication techniques resulted in the emergence of ultrasensitive nanomechanical sensors, commonly consisting of simple vibrating structures such as cantilevers, strings or membranes that exhibit resonance behavior. The principle of operation is based on monitoring the resonance frequency shift due to various external factors, including mass, force and temperature change. The high sensitivity of nanomechanical resonators has already been exploited to create a group of photothermal spectroscopy devices capable of exceptionally fast chemical analysis of compounds on the femtogram level.Nanomechanical infrared spectroscopy (NAM-IR) originated from photothermal bilayer cantilever deflection spectroscopy and is based on the photothermal response of a nanomechanical resonator. It has already been presented with a string resonator, which was acting as sampling element and temperature sensor. The string could be considered as a single filter-fiber and guaranteed relatively high overall sampling efficiency through impaction of airborne nanoparticles on the resonator surface. When the analyte, collected by the sensor, is exposed to IR radiation it absorbs light at a certain wavelength corresponding to its specific molecular vibrations. This thermal energy heats the resonator and leads to its thermal expansion followed by a decrease of the tensile stress in the resonator. In turn it eventually causes the resonance frequency to shift towards lower values.However, further development of this approach was curbed by difficult and inefficient coupling of the IR light beam to a nanometer-sized resonator. In addition, readout of vibration was done by laser Doppler vibrometer, a precise but bulky and expensive instrument. These issues hindered a realworld application of the NAM-IR method. In order to overcome them, string resonators were replaced by membranes. A reliable sampling technique was maintained by adding perforation to membranes and thereby essentially getting membrane porous filters. Membranes gave also access to fully integrated magnetic transduction that allowed for significant shrinkage and simplification of the system.An analytical model of a locally heated membrane was developed and confirmed through FEM simulations. Then, low stress silicon nitride perforated membranes were fabricated and characterized using two different experimental setups that employed optical and magnetomotive readout. Finally, spectroscopic measurements and crystallization study of about 100 pg of the model drug ‘indomethacin’ were performed. Obtained IR spectra were in good agreement both with conventional Fourier transform IR spectroscopy (FTIR) reference and literature reports. The performance of the magnetic transduction scheme was found to be comparable to the traditionally used optical detection with a minimum sample mass required for analysis of roughly 100 fg. NAM-IR technique requires exceptionally small amount of sample and does not involve timeconsuming sample preparation. Therefore, it is a promising alternative to standard IR spectroscopy with vast possible applications for example in the pharmaceutical industry.
展开▼
