Techniques to measure the trapping force in an optical tweezers without any prior assumptions about the trap
shape have been developed. The response of a trapped micro or nanoparticle to a step input is measured and
then used to calculate the trapping force experienced by the particle as a function of it's position in the trap. This
method will provide new insight into the trapping behavior of nanoparticles, which are more weakly bound than
microparticles and thereby explore larger regions of the trapping potential due to Brownian motion. Langevin
dynamics simulations are presented to model the system and are used to demonstrate this technique. Preliminary
experimental results are then presented to validate the simulations. Finally, the measured trapping forces, from
simulations and laboratory experiments, are integrated to recover the trapping potential.
There are several new tools for manipulating microscopic objects. Among them, optical tweezers (OT) has two distinguishing advantages. Firstly, OT can easily release an object without the need of a complicated detaching scheme. Secondly, it is anticipated to manipulate an object with six degrees of freedom. OT is realized by tightly focusing a laser beam on microscopic objects. Grabbing and releasing is easily done by turning a laser beam on and off. For doing a dexterous manipulation on an object, a complicated potential trap must be calculated and applied. We foresee that such calculation method will be developed in the near future. One of the candidates for implementing the calculated trap is scanning optical tweezers (SOT). SOT can be built by using actuators with a scanning frequency in the order of a hundred Hertz. We need fast scanners to stably trap an object. In this study, we present our design of such SOT. The SOT uses piezo-actuated tilt mirror and objective positioner to scan full three-dimensional workspace.
In order to realize the flexibility optical trapping offers as a nanoassembly tool, we need to develop natural and intuitive interfaces to assemble large quantities of nanocomponents quickly and cheaply. We propose a system to create such an interface that is scalable, inter-changeable and modular. Several prototypes are described, starting with simple interfaces that control a single trap in the optical tweezers instrument using a 3-dimensional Phantom haptic device. A networkbased approach is adopted early on, and a modular prototype is then described in detail. In such a design, individual modules developed on different platforms work independently and communicate with each other through a common language interface using the Neutral Messaging Language (NML) communication protocol. A natural user interface is implemented that can be used to create and manipulate traps interactively like in a CAD program. Modules such as image processing and automatic assembly are also added to help simplify routine assembly tasks. Drawing on lessons learned from the prototypes, a new system specification is formulated to better integrate the modules. Finally, conclusions are drawn on the overall viability and future of network-based systems for nanoassembly using optical tweezers.
System identification methods are presented for the estimation of the characteristic frequency of an optically trapped particle. These methods are more amenable to automated on-line measurements and are believed to be less prone to erroneous results compared to techniques based on thermal noise analysis. Optical tweezers have been shown to be an effective tool in measuring the complex interactions of micro-scale particles with piconewton resolution. However, the accuracy of the measurements depends heavily on knowledge of the trap stiffness and the viscous drag coefficient for the trapped particle. The most commonly referenced approach to measuring the trap stiffness is the power spectrum method, which provides the characteristic frequency for the trap based on the roll-off of the frequency response of a trapped particle excited by thermal fluctuations. However, the reliance on thermal fluctuations to excite the trapping dynamics results in a large degree of uncertainty in the estimated characteristic frequency. These issues are addressed by two parameter estimation methods which can be implemented on-line for fast trap characterization. The first is a frequency domain system identification approach which combines swept-sine frequency testing with a least-squares transfer function fitting algorithm. The second is a recursive least-squares parameter estimation scheme. The algorithms and results from simulation studies are discussed in detail.