Brandon K. Chen

Autonomous Robotic Pick-and-Place of Microobjects

This paper presents a robotic system that is capable of both picking up and releasing microobjects with high accuracy, reliability, and speed. Due to force-scaling laws, large adhesion forces at the microscale make rapid, accurate release of microobjects a long-standing challenge in micromanipulation, thus representing a hurdle toward automated robotic pick-and-place of micrometer-sized objects. The system employs a novel microelectromechanical systems (MEMS) microgripper with a controllable plunging structure to impact a microobject that gains sufficient momentum to overcome adhesion forces. The performance was experimentally quantified through the manipulation of 7.5-10.9 ¿m borosilicate glass spheres in an ambient environment. Experimental results demonstrate that the system, for the first time, achieves a 100% success rate in release (which is based on 700 trials) and a release accuracy of 0.45 ± 0.24 ¿m. High-speed, automated microrobotic pick-and-place was realized by visually recognizing the microgripper and microspheres, by visually detecting the contact of the microgripper with the substrate, and by vision-based control. Example patterns were constructed through automated microrobotic pick-and-place of microspheres, achieving a speed of 6 s/sphere, which is an order of magnitude faster than the highest speed that has been reported in the literature.

Active Release of Microobjects Using a MEMS Microgripper to Overcome Adhesion Forces

Due to force scaling laws, large adhesion forces at the microscale make rapid accurate release of microobjects a long-standing challenge in pick-place micromanipulation. This paper presents a new microelectromechanical systems (MEMS) microgripper integrated with a plunging mechanism to impact the microobject for it to gain sufficient momentum to overcome adhesion forces. The performance was experimentally quantified through the manipulation of 7.5-10.9-mum borosilicate glass spheres in an ambient environment under an optical microscope. Experimental results demonstrate that this microgripper, for the first time, achieves a 100% successful release rate (based on 200 trials) and a release accuracy of 0.70 plusmn0.46 mum. Experiments with conductive and nonconductive substrates also confirmed that the release process is not substrate dependent. Theoretical analyses were conducted to understand the release principle. Based on this paper, further scaling down the end structure of this microgripper will possibly provide an effective solution to the manipulation of submicrometer-sized objects.

MEMS microgrippers with thin gripping tips

Gripping small objects requires tool tips of comparable dimensions. Current methods for miniaturizing an MEMS tool entirely down to sub-micrometer in dimensions, however, come with significant tradeoffs in device performance. This paper presents a microfabrication approach to selectively miniaturize gripping tips only to sub-micrometers in thickness. The process involves using the thin buried SiO2 layer of a standard silicon-on-insulator wafer to form gripping tips, and using the thick device silicon layer to construct high-aspect-ratio structures for structural, sensing, and actuation functions. The microgrippers with thin gripping tips (i.e. finger-nail-like) were experimentally characterized and applied to gripping 100 nm gold spheres inside a scanning electron microscope.

Strengthening in Graphene Oxide Nanosheets: Bridging the Gap between Interplanar and Intraplanar Fracture

Graphene oxide (GO) is a layered material comprised of hierarchical features which possess vastly differing characteristic dimensions. GO nanosheets represent the critical hierarchical structure which bridges the length-scale of monolayer and bulk material architectures. In this study, the strength and fracture behavior of GO nanosheets were examined. Under uniaxial loading, the tensile strength of the nanosheets was measured to be as high as 12 ± 4 GPa, which approaches the intrinsic strength of monolayer GO and is orders of magnitude higher than that of bulk GO materials. During mechanical failure, brittle fracture was observed in a highly localized region through the cross-section of the nanosheets without interlayer pull-out. This transition in the failure behavior from interplanar fracture, common for bulk GO, to intraplanar fracture, which dominates failure in monolayer GO, is responsible for the high strength measured in the nanosheets. Molecular dynamics simulations indicate that the elastic release from the propagation of intraplanar cracks initiates global fracture due to interlayer load transmission through hydrogen bond networks within the gallery space of the GO nanosheets. Furthermore, the GO nanosheet strength and stiffness were found to be strongly correlated to the effective volume and thickness of the samples, respectively. These findings help to bridge the understanding of the mechanical behavior of hierarchical GO materials and will ultimately guide the application of this intermediate scale material.

Robotic Probing of Nanostructures inside Scanning Electron Microscopy

Probing nanometer-sized structures to evaluate the performance of integrated circuits (IC) for design verification and manufacturing quality monitoring demands precision nanomanipulation technologies. To minimize electron-induced damage and improve measurement accuracy, scanning electron microscopy (SEM) imaging parameters must be cautiously chosen to ensure low electron energy and dosage. This results in significant image noise and drift. This paper presents automated nanoprobing with a nanomanipulation system inside a standard SEM. We achieved SEM image denoising and drift compensation in real time. This capability is necessary for achieving robust visual tracking and servo control of nanomanipulators for probing nanostructures in automated operation. This capability also proves highly useful to conventional manual operation by rendering real-time SEM images that have little noise and drift. The automated system probed nanostructures on an SEM metrology chip as surrogates of electronic features on IC chips. Success rates in visual tracking and Z-contact detection under various imaging conditions were quantitatively discussed. The experimental results demonstrate the systems capability for automated probing of nanostructures under IC-chip-probing relevant electron microscope imaging conditions.

A Closed-Loop Controlled Nanomanipulation System for Probing Nanostructures Inside Scanning Electron Microscopes

Probing nanostructures (e.g., nanoelectronics) requires accurate and precise nanopositioning. Furthermore, since measuring I-V data from dc to GHz typically takes more than a minute, tolerance for position drift is stringent during the data collection process. This paper reports a closed-loop controlled nanomanipulation system for operation inside a scanning electron microscope. A new position sensing method with low power consumption is used to achieve nanometer sensing resolution and effective heat dissipation management. For automated probing of nanostructures, the position sensor-based closed-loop probing approach was found to be four times faster than visually servoed probing, and ten times faster compared to manual operation. Probing accuracy was determined to be better than 3 nm and a drift rate lower than 1 nm/min.

Overcoming adhesion forces: Active release of micro objects in micromanipulation

Due to force scaling laws, rapid, accurate release of micro objects has been a long-standing challenge for microrobotic manipulation. This paper presents an active release technique that for the first time, achieves 100% repeatability and a release accuracy of 0.70±0.46µm, experimentally quantified through the manipulation of 10µm glass spheres under an optical microscope. Using a new MEMS (microelectromechanical systems) microgripper, this technique employs a controllable plunging mechanism for the micro object to gain sufficient momentum to overcome adhesion forces. Experimental results also confirmed that this technique is not substrate dependent. Theoretical analyses were conducted to understand the release principle. Based on this preliminary study, the technique may also prove to be an effective solution to active release of sub-micron objects in robotic pick-place.

Mechanical characterization of thin films using a MEMS device inside SEM

A MEMS device was developedfor mechanical characterization of 2D ultra-thin films. The device utilizes electrothermal actuators to apply uniaxial tension. The robust design makes the device capable of withstanding both dry and wet transfer of 2D ultra-thin film materials onto the suspended structures of the device. Fracture stress of thin graphene oxide (GO) films was measured.

Correlative microscopy for nanomanipulation of sub-cellular structures

Nanomanipulation under scanning electron microscopy (SEM) has been demonstrated as an enabling technique for the manipulation and characterization of nanomaterials. We recently developed nanomanipulation techniques for the extraction and identification of DNA contained within sub-nuclear locations of a single cell nucleus. In nanomanipulation of DNA, a key step is target identification through SEM-fluorescence correlative imaging. Existing image correlation techniques often require fiducial marks and/or manual feature selection or data training, which are unsuitable for DNA nanomanipulation. This paper presents an approach for correlating SEM-fluorescence microscopy images, proven effective in processing images taken under poor SEM imaging conditions imposed by the necessity of preserving DNAs biochemical integrity. The performance of the image correlation approach under different imaging conditions was quantitatively evaluated. Compared to manual correlation by skilled operators, the automated correlation approach demonstrated an order of magnitude higher speed. The SEM-fluorescence correlation approach enables targeted nanomanipulation of sub-cellular structures under SEM.

Automated nanoprobing under scanning electron microscopy

Nanomanipulation inside electron microscopes enables a multitude of precision applications. The semiconductor industry employs this capability to probe sub-micrometer-sized features to evaluate the performance of integrated circuits (IC) for design/quality monitoring. In electron microscopy imaging, the use of low accelerating voltages and high magnifications, as required for IC nanoprobing tasks, results in significant image noise and drift. This paper presents automated nanoprobing with a nanomanipulation system inside a standard scanning electron microscope (SEM). We achieved SEM image denoising and drift compensation in real time. This capability is necessary for achieving robust visual tracking and servo control of nanoprobes for probing nanostructures in automated operation. This capability also proves highly useful to conventional manual operation by rendering real-time SEM images that have little noise and drift. The automated system probed nanostructures on an SEM metrology chip as surrogates of electronic features on IC chips. Success rates in visual tracking and Z-contact detection under various imaging conditions were quantitatively discussed. The experimental results demonstrate the systems capability for automated probing of nanostructures under IC-chip-probing relevant EM imaging conditions.