Recently, there has been a strong interest in implementing nanomechanical devices as mass sensors. In this type of experiment, an important question to address is to know to what extent the observed frequency shift is exclusively due to the mass loading. We present a device with an innovative design that allows the direct determination of the measurement uncertainty. Two almost identical nanomechanical resonators are simultaneously operated: one serves as sensor and the other as reference. In this way, rapid and reliable measurements in air are made possible.
The nanomechanical device is a polysilicon double cantilever (DBC) defined by electron-beam lithography on prefabricated CMOS substrates. Both resonators are orthogonally orientated and have same anchor and same readout electrode. DBC are operated in their lateral flexion fundamental mode. The resonance spectrum is detected using a capacitive scheme. Therefore, DBC are monolithically integrated with CMOS circuitry for signal amplification and parasitic capacitances reduction. Details on fabrication process and circuit design can be found elsewhere. Typically, these cantilevers are 14 µm long, 300 nm wide, 500 nm thick and have a 600 nm gap respect to the readout electrode. The theoretical mass sensitivity of the cantilever is 1.5 attog/Hz.
First, the device is measured in air. Its resonance spectrum exhibits two different peaks related to each cantilever. Then, the chip is placed into the vacuum chamber of a focused ion beam (FIB) and a certain amount of material (a Ga-Pt-C alloy, density of 12 g/cm3) is locally deposited at the tip of only one of the two cantilevers. A typical deposit is 500 nm long, 250 nm wide and 160 nm thick. Then the DBC is measured again in air. The cantilever on which the material was deposited is used to determine with high precision the amount of deposited mass by measuring its resonance frequency shift downwards. The other cantilever also exhibits a shift, whether positive or negative, but much smaller: this deviation provides a straightforward value of the measurement uncertainty.
Following this procedure, we have measured masses in the order of 200 fg with an uncertainty of 60 fg. These measurements are corroborated by the determination of the deposit size with SEM images.
|