-
Ultrafast Reservoir Computing based on Nonlinear Nanomechanical Resonators at Ambient Conditions
Authors:
Enise Kartal,
Yunus Selcuk,
Batuhan E. Kaynak,
M. Taha Yildiz,
Cenk Yanik,
M. Selim Hanay
Abstract:
Reservoir computing offers an energy-efficient alternative to deep neural networks (DNNs) by replacing complex hidden layers with a fixed nonlinear system and training only the final layer. This work investigates nanoelectromechanical system (NEMS) resonators for reservoir computing, utilizing inherent nonlinearities and the fading memory effect from NEMS's transient response. This approach transf…
▽ More
Reservoir computing offers an energy-efficient alternative to deep neural networks (DNNs) by replacing complex hidden layers with a fixed nonlinear system and training only the final layer. This work investigates nanoelectromechanical system (NEMS) resonators for reservoir computing, utilizing inherent nonlinearities and the fading memory effect from NEMS's transient response. This approach transforms input data into a higher-dimensional space for effective classification. The smaller size and higher operating frequencies of the NEMS resonators enable faster processing rates than recent demonstrations with micromechanical systems, while their compact footprint and ability to operate under ambient conditions simplify integration into practical applications. Through an MNIST handwritten digit recognition test, this system achieved 90% accuracy with a 3.3-microsecond processing time per pixel, highlighting the potential for various applications that require efficient and fast information processing in resource-constrained environments.
△ Less
Submitted 25 September, 2024;
originally announced September 2024.
-
Multi-mode Brownian Dynamics of a Nanomechanical Resonator in a Viscous Fluid
Authors:
H. Gress,
J. Barbish,
C. Yanik,
I. I. Kaya,
R. T. Erdogan,
M. S. Hanay,
M. González,
O. Svitelskiy,
M. R. Paul,
K. L. Ekinci
Abstract:
Brownian motion imposes a hard limit on the overall precision of a nanomechanical measurement. Here, we present a combined experimental and theoretical study of the Brownian dynamics of a quintessential nanomechanical system, a doubly-clamped nanomechanical beam resonator, in a viscous fluid. Our theoretical approach is based on the fluctuation-dissipation theorem of statistical mechanics: We dete…
▽ More
Brownian motion imposes a hard limit on the overall precision of a nanomechanical measurement. Here, we present a combined experimental and theoretical study of the Brownian dynamics of a quintessential nanomechanical system, a doubly-clamped nanomechanical beam resonator, in a viscous fluid. Our theoretical approach is based on the fluctuation-dissipation theorem of statistical mechanics: We determine the dissipation from fluid dynamics; we incorporate this dissipation into the proper elastic equation to obtain the equation of motion; the fluctuation-dissipation theorem then directly provides an analytical expression for the position-dependent power spectral density (PSD) of the displacement fluctuations of the beam. We compare our theory to experiments on nanomechanical beams immersed in air and water, and obtain excellent agreement. Within our experimental parameter range, the Brownian force noise driving the nanomechanical beam has a colored PSD due to the ``memory" of the fluid; the force noise remains mode-independent and uncorrelated in space. These conclusions are not only important for nanomechanical sensing but also provide insight into the fluctuations of elastic systems at any length scale.
△ Less
Submitted 1 November, 2023;
originally announced November 2023.
-
Nanomechanical Measurement of the Brownian Force Noise in a Viscous Liquid
Authors:
Atakan B. Ari,
M. Selim Hanay,
Mark R. Paul,
Kamil L. Ekinci
Abstract:
We study the spectral properties of the thermal force giving rise to the Brownian motion of a continuous mechanical system -- namely, a nanomechanical beam resonator -- in a viscous liquid. To this end, we perform two separate sets of experiments. First, we measure the power spectral density (PSD) of the position fluctuations of the resonator around its fundamental mode at its center. Then, we mea…
▽ More
We study the spectral properties of the thermal force giving rise to the Brownian motion of a continuous mechanical system -- namely, a nanomechanical beam resonator -- in a viscous liquid. To this end, we perform two separate sets of experiments. First, we measure the power spectral density (PSD) of the position fluctuations of the resonator around its fundamental mode at its center. Then, we measure the frequency-dependent linear response of the resonator, again at its center, by driving it with a harmonic force that couples well to the fundamental mode. These two measurements allow us to determine the PSD of the Brownian force noise acting on the structure in its fundamental mode. The PSD of the force noise extracted from multiple resonators spanning a broad frequency range displays a "colored spectrum". Using a single-mode theory, we show that, around the fundamental resonances of the resonators, the PSD of the force noise follows the dissipation of a blade oscillating in a viscous liquid -- by virtue of the fluctuation-dissipation theorem.
△ Less
Submitted 16 December, 2020;
originally announced December 2020.
-
High Resolution Dielectric Characterization of Single Cells and Microparticles Using Integrated Microfluidic Microwave Sensors
Authors:
Arda Secme,
Uzay Tefek,
Burak Sari,
Hadi Sedaghat Pisheh,
H. Dilara Uslu,
Ozge Akbulut,
Berk Kucukoglu,
R. Tufan Erdogan,
Hashim Alhmoud,
Ozgur Sahin,
M. Selim Hanay
Abstract:
Microwave sensors can probe intrinsic material properties of analytes in a microfluidic channel at physiologically relevant ion concentrations. While microwave sensors have been used to detect single cells and microparticles in earlier studies, the synergistic use and comparative analysis of microwave sensors with optical microscopy for material classification and size tracking applications have b…
▽ More
Microwave sensors can probe intrinsic material properties of analytes in a microfluidic channel at physiologically relevant ion concentrations. While microwave sensors have been used to detect single cells and microparticles in earlier studies, the synergistic use and comparative analysis of microwave sensors with optical microscopy for material classification and size tracking applications have been scarcely investigated so far. Here we combined microwave and optical sensing to differentiate microscale objects based on their dielectric properties. We designed and fabricated two types of planar sensor: a Coplanar Waveguide Resonator (CPW) and a Split-Ring Resonator (SRR). Both sensors possessed sensing electrodes with a narrow gap to detect single cells passing through a microfluidic channel integrated on the same chip. We also show that standalone microwave sensors can track the relative changes in cellular size in real-time. In sensing single 20-micron diameter polystyrene particles, Signal-to-Noise ratio values of approximately 100 for CPW and 70 for SRR sensors were obtained. These findings demonstrate that microwave sensing technology can serve as a complementary technique for single-cell biophysical experiments and microscale pollutant screening.
△ Less
Submitted 25 February, 2023; v1 submitted 27 October, 2020;
originally announced October 2020.
-
Full Electrostatic Control of Nanomechanical Buckling
Authors:
Selcuk Oguz Erbil,
Utku Hatipoglu,
Cenk Yanik,
Mahyar Ghavami,
Atakan B. Ari,
Mert Yuksel,
M. Selim Hanay
Abstract:
Buckling at the micro and nanoscale generates distant bistable states which can be beneficial for sensing, shape-reconfiguration and mechanical computation applications. Although different approaches have been developed to access buckling at small scales, such as the use heating or pre-stressing beams, very little attention has been paid so far to dynamically and precisely control all the critical…
▽ More
Buckling at the micro and nanoscale generates distant bistable states which can be beneficial for sensing, shape-reconfiguration and mechanical computation applications. Although different approaches have been developed to access buckling at small scales, such as the use heating or pre-stressing beams, very little attention has been paid so far to dynamically and precisely control all the critical bifurcation parameters, the compressive stress and the lateral force on the beam. Precise and on-demand generation of compressive stress on individually addressable microstructures is especially critical for morphologically reconfigurable devices. Here, we develop an all-electrostatic architecture to control the compressive force, as well as the direction and amount of buckling, without significant heat generation on micro/nano structures. With this architecture, we demonstrated fundamental aspects of device function and dynamics. By applying voltages at any of the digital electronics standards, we have controlled the direction of buckling. Lateral deflections as large as 12% of the beam length were achieved. By modulating the compressive stress and lateral electrostatic force acting on the beam, we tuned the potential energy barrier between the post-bifurcation stable states and characterized snap-through transitions between these states. The proposed architecture opens avenues for further studies that can enable efficient actuators and multiplexed shape-shifting devices.
△ Less
Submitted 13 February, 2019;
originally announced February 2019.
-
Nonlinear Nanomechanical Mass Spectrometry at the Single-Nanoparticle Level
Authors:
Mert Yuksel,
Ezgi Orhan,
Cenk Yanik,
Atakan B. Ari,
Alper Demir,
M. Selim Hanay
Abstract:
Nanoelectromechanical Systems (NEMS) have emerged as a promising technology for performing the mass spectrometry of large biomolecules and nanoparticles. As nanoscale objects land on NEMS sensor one by one, they induce resolvable shifts in the resonance frequency of the sensor proportional to their weight. The operational regime of NEMS sensors is often limited by the onset-of-nonlinearity, beyond…
▽ More
Nanoelectromechanical Systems (NEMS) have emerged as a promising technology for performing the mass spectrometry of large biomolecules and nanoparticles. As nanoscale objects land on NEMS sensor one by one, they induce resolvable shifts in the resonance frequency of the sensor proportional to their weight. The operational regime of NEMS sensors is often limited by the onset-of-nonlinearity, beyond which the highly sensitive schemes based on frequency tracking by phase-locked loops cannot be readily used. Here, we develop a measurement architecture to operate at the nonlinear regime and measure frequency shifts induced by analytes in a rapid and sensitive manner. We used this architecture to individually characterize the mass of gold nanoparticles and verified the results by performing independent measurements of the same nanoparticles based on linear mass sensing. Once the feasibility of the technique is established, we have obtained the mass spectrum of a 20 nm gold nanoparticle sample by individually recording about five hundred single particle events using two modes working sequentially in the nonlinear regime. The technique obtained here can be used for thin nanomechanical structures which possess a limited dynamic range.
△ Less
Submitted 7 February, 2019;
originally announced February 2019.
-
Spatially Resolved Sensing in Microfluidics with Multimode Microwave Resonators
Authors:
Mehmet Kelleci,
Hande Aydogmus,
Levent Aslanbas,
Selcuk Oguz Erbil,
M. Selim Hanay
Abstract:
The analogy between mechanical and electromagnetic resonators has been a celebrated paradigm of science and engineering. Exploration of this analogy in recent years has resulted in several exciting research directions, including cavity optomechanics[1], phononic bandgap materials[2] and phononic metamaterials[3-5]. In these examples, progress in electromagnetic research has usually led the way for…
▽ More
The analogy between mechanical and electromagnetic resonators has been a celebrated paradigm of science and engineering. Exploration of this analogy in recent years has resulted in several exciting research directions, including cavity optomechanics[1], phononic bandgap materials[2] and phononic metamaterials[3-5]. In these examples, progress in electromagnetic research has usually led the way for their mechanical counterparts. Here, we contribute to this analogy from a different perspective by adapting a sensing technique originally developed for mechanical devices to increase the capabilities of sensors based on electromagnetic fields. More specifically, multimode resonance techniques, which enable spatial resolution in inertial mass sensing experiments with nanoelectromechanical systems (NEMS), are tailored for use in microwave resonant sensing, which is commonly employed in microfluidics. We show that the use of higher-order modes of such sensors can provide electrical volume, position and geometric size data. The combination of such spatial features implies the potential for image reconstruction when a large number of modes are used. With the analytical and experimental framework presented here, we can move beyond simple counting and achieve the sizing and imaging of analytes with impedance spectroscopy.
△ Less
Submitted 20 February, 2017;
originally announced February 2017.
