【作者】 龚錾；

【导师】 楼立人； 李银妹；

【作者基本信息】 中国科学技术大学 ， 光学， 2007， 博士

【摘要】 从光镊问世至今已有二十年，它已从最初单纯的捕获操控微粒的装置，发展成了今天集高精度操控和微小力学参数测量于一体的实验系统，已经成为生物学，纳米科学和分散体系等涉及微小粒子的领域的一种理想研究手段。随着这些领域的发展，人们对光镊系统和相关实验方法与技术不断提出新的要求。本研究就是以满足这种日益增长的需求为目标来展开的，包括设计研制一套全新的多功能光镊实验平台——纳米光镊系统，探索更高精度更高效率的实验方法。论文特别着重于对关键的光镊测量技术的分析研究，建立了相匹配的测量系统，并对目前常用的测量方法所存在的问题进行了细致的讨论和探究，以期为完善和发展新的实验方法提供有用的基础。光镊实验系统是利用光镊技术进行实验研究的基础，我们的工作的第一步就是设计研制全新的光镊实验系统。基于对自身研究方向对实验系统的要求，确定了设计的基本考虑和目标。一是要有多个可独立操控的光镊，二是系统应有高的操控精度和测量精度，以适应微小粒子运动学、动力学过程和粒子间相互作用的研究。通过调研国际上已有实验装置的经验，我们选择了三光镊的方案，并针对不同操控维度的具体条件，设计了主动操控和被动操控两种操控模式。在满足纳米精度的操控要求下，选择了不同精度的操控驱动方案(压电驱动和步进电机驱动)，来完成不同维度的操控，既可以实现几个光镊同时对样品进行纳米精度的操控，又可以保证对样品具有较大的操控范围，同时也降低了装置构建的费用。为了使系统兼有高的空间分辨率和时间分辨率，装置中配置了二套探测系统：CCD图像探测系统和四象限光电探测系统(QD)，引进或提出了一些针对不同实验条件的实验方法，诸如CCD图像相关运算分析法和四象限探测器系统结合背散射光照明方法，并开发了相关的数据分析软件。论文还讨论了光阱刚度标定问题，分析比较了不同方法的精度和特点，在此基础上为系统配置了合用的标定方法。实际应用表明，所建立的纳米光镊系统的位置定位精度和探测精度已经达到了亚纳米量级，微小力的测量精度达到了亚皮牛量级，满足了微小粒子个体行为研究的需求。在微小粒子个体行为的研究中常需要测量粒子运动过程中的微小力学量，比如位移和力等物理量等。这时测量精度具有头等重要的意义。论文着重研究了微粒位置和微小力的测量方法，特别是微粒位置的精确测量问题，它不仅是运动学的基本物理量，也是光阱刚度标定的基础。论文围绕这一中心问题，重点研究了采集系统带宽和刚度标定实验中样品的纵向偏移这两个参数对实验测量带来的影响。由于样品在液体环境中所做布朗运动的频率很高，而任何采集系统的带宽都是有限的，都要比布朗运动的带宽低几个量级，因此使用这些采集系统都会损失一定的高频信号，对应在光镊系统中就会导致实验中所测量到的位移具有较大的误差。通过使用Monte-Carlo数值模拟方法，我们从时域的角度研究了不同的实验条件下，采集系统曝光时间的有限长度(或带宽不足)给位置测量和使用热运动分析法进行刚度标定带来的影响。从这一视角进行的分析是对传统的频域分析的补充。通过对数值模拟结果的分析，我们讨论了带来误差的原因，以及使用热运动分析法来标定刚度时所需要的带宽。影响光镊测力精度的重要因素是其刚度标定。在文献中很早就报道过这样的一个现象：当光阱中的粒子受到横向外力(通常为粘滞力)作用时，不仅在横向上会偏离光阱中心，而且在纵向上也会出现一个偏移。但是一直以来，很少有人对这个现象及可能给实验研究带来的影响进行系统的研究。我们首先从实验上测量了光阱捕获的粒子在横向外力下的纵向偏移量与横向偏移量之间的关系。实验结果表明：样品受到横向力作用时，新的平衡位置并不是与光阱中心处在同一水平面，其纵向位置随着横向位移增加而升高。我们进而使用几何光学模型分析光阱中微粒受到的光阱力的分布，由此计算了微粒的纵向平衡位置轨迹，计算结果与实验测量的结果一致。由于这个现象与使用流体力学法进行横向力测量的前提假设相矛盾，因此必然会给测量结果带来误差。然而，分析也表明，在使用流体力学法标定光阱横向刚度时，只要保证样品的横向位置不超过一定的值，其误差可以忽略；但对最大横向阱力的测量，流体力学法所测量到的实际上是逃逸阱力，这个值比实际的最大横向阱力要小很多。上述讨论表明，为了更精确的测量光阱力，必须发展新的刚度标定方法。更多还原

【Abstract】 It has been over twenty years since optical tweezers are invented. At first, Optical tweezers are just a simple facility for trap and manipulate micro sized particles. But now, it has become a complex experiment setup consisting of the high precision manipulation and the measurement of small mechanical quantities. Optical tweezers can manipulate many kinds of particles with submicron to tens of microns diameter, including cells, organelles, and polymer beads. Such particles are usually the subject investigated in biology, nanophase materials, and disperse system fields. At present, optical tweezers technique can position and detect the trapped particles in sub nanometer magnitude, and it can precisely measure the small force between sub piconewton to hundreds of piconewtons. The displacements and small forces studied in biology and disperse system are just in such magnitudes. Additionally, because the specimen manipulation by optical tweezers is non-contact, there’s mechanical damage on the specimen. All the characters of optical tweezers make it very suitable for being applied in biology and disperse system fields.Along with the development of biology studies, especially in the biological macromolecules researches, there are new requests to optical tweezers technique because it is a very important means in these fields. Our study is just to fulfill these increasing requests. In order to follow this situation, we have begun to upgrade the optical tweezers system in our lab and seeking for more precise and efficient experiment methods. This thesis is mainly about comprehensively upgrading the study ability of optical tweezers technique in our lab. We have designed a new multi-functional experimental setup: nanometer optical tweezers system, introduced or brought forward some experiment methods for different experiment conditions, analyzed the measurement accuracy of these methods, and studied the influence of two parameters, the bandwidth of the acquisition system and the axial deviation of specimen in the experiments, on the measurements of the mechanical quantities.Because optical tweezers setup is the base of the experiment researches using optical tweezers technique, the first step of our job is the design of a new optical tweezers setup. Through the surveys of the setups in other labs and the analysis of our own needs, our design emphasis is finally put on the combination of multi optical tweezers and the upgrade of the accuracies of the manipulation and measurement. In order to study the biological macromolecules and the interaction between the molecules, we have to manipulate several molecules at the same time. We finally choose the project of treble optical tweezers. In the setup, we use several piezoelectric and stepping motor devices. Based on these devices, we design two manipulation modes: initiative manipulation mode and passive manipulation mode. In the experiments, combining these two manipulation modes, we can manipulate three specimens with very high precision and rather big area at the same time.In the experiments, we usually need to measure some small mechanical quantities, for example, the displacements and the forces. Force measurement needs the stiffness of the optical tweezers first, and the stiffness calibration is generally based on the analysis of the specimen movements, so the displacement measurement is actually the foundation of all the mechanical quantities measurements. In our experiment setup, we use two projects to measure the specimen displacements in high precision. One is CCD system associated with correlation calculation of images; the other one is quadrant detector system associated with back scattering illumination. Each of these two projects has some merits and shortcomings, but they can reinforce each other. Combining these projects, the accuracy of the displacement measurement can reach sub nanometer, and the time resolution can reach sub milliseconds. Based on the high precision displacement measurement, we introduce four methods, thermal motion analysis method, drag force method, power spectrum method and periodic driving force method to calibrate the optical trap stiffness. We measure the accuracies of these methods by experiments and analyses, compare the errors of them, discuss the sources of the errors, and finally get the respective appropriate experiment condition of these four methods.The frequency of the Brownian motion of the specimen is quite high; however the bandwidth of any acquisition system is limited and much smaller than that of the Brownian motion. So when an acquisition system is used to measure the track of the Brownian motion, the high frequency signal of the motion will be dropped out. This causes the experimentally measured displacement has a large error. By using Monte-Carlo numerical modeling method, the influence of the insufficient bandwidth of the acquisition system on the displacement measurement and the stiffness calibration using the thermal motion analysis method is studied in the time domain. Through the analysis of the numerical modeling results, we discuss the source that brings on the error and the essential bandwidth when thermal motion method is used.It has been reported a few years ago that when the specimen endures a lateral force, it deviates from the optical trap center not only laterally but also axially. However, this phenomenon and the influence of it on the experiment measurements haven’t been studied in detail. We first measure the relation between the axial deviation and the lateral deviation. The result shows that the equilibrium track of the specimen is not in the horizon plane that passes the trap center and the axial position lifts along with the increase of the lateral displacement. Then we calculate the equilibrium track of the specimen by using a ray optics model, the result is consistent with the experiment result. This phenomenon conflicts with the hypothetic premises of drag force method, so it surely brings on errors to lateral force measurement using drag force method. Through further analysis of force field of the optical trap, we can figure out the detailed effect of the axial deviation. When the drag force method is used to calibrate the lateral trap stiffness, the error brought by the axial deviation can be neglect as long as the lateral displacement of the specimen is less than a special value. However, when the drag force method is used to measure the max lateral trapping force, the measured force is actually the escape force, which is much less than the true max lateral trapping force.更多还原