Nonlinear optical spectroscopy and super-resolution microscopy

水表面化学的非线性光谱研究 该提案应用非线性光学光谱学来研究与环境和工业相关的水表面化学。非线性光学光谱学，如和频产生，是少数几种可用于测量掩埋的液体/固体界面的振动光谱的技术之一。振动光谱提供了有价值的信息的分子组成和组织的水界面。该提案将研究水/固体界面处的非均质冰形成，旨在了解大气中的冰云形成，以及水/矿物界面处的分子吸附动力学，旨在减少加拿大油砂开发的资源消耗和环境影响。 超分辨率显微镜：从单分子光谱到整个细胞中的蛋白质组装 在过去的几年里，科学家们已经开发出几种方法来克服衍射极限，并彻底改变了光学显微镜。单分子定位显微镜（SMLM）提供了前所未有的10 nm的空间分辨率。SMLM技术在任何给定时间仅激活和定位荧光团的稀疏子集。因为每个荧光团的图像在空间上被很好地分离，所以可以以非常高的精度确定每个荧光团的位置。可以使用从一系列衍射极限图像获得的荧光团的x、y和z坐标来重建超分辨率图像。这些3D点云无法使用传统的图像分析工具进行分析。我们将采用机器学习来模拟人眼的调制传递函数，以分析SMLM产生的3D点云。由于在许多方面人类优于任何模式识别机器，这种方法将导致一个有效的蛋白质聚类分析工具，而无需任何用户输入。 此外，我们建议将样品稳定系统植入到结构照明显微镜（SIM）中，以消除由于样品漂移而产生的伪影。与SMLM相比，SIM在3D成像方面有几个优势。它具有更高的速度，使用相对较低的照明功率，并且易于用于数字成像。拟议的SIM将使我们能够获得大细胞的可靠超分辨率图像，例如心肌细胞，其平均长度为100 µm，宽度为18 µm。显微镜将使我们能够对整个细胞中的蛋白质组织进行成像，并做出令人兴奋的发现。

Studies of Water Surface Chemistry Using Nonlinear Optical Spectroscopy       This proposal applies nonlinear optical spectroscopy to study environmentally- and industrially-relevant surface chemistry of water. Nonlinear optical spectroscopy, such as sum-frequency generation, is one of the few techniques available to measure the vibrational spectra of a buried liquid/solid interface. The vibrational spectra provide valuable information on the molecular composition and organization of water interfaces. This proposal will study heterogeneous ice formation at water/solid interfaces, aiming to understand ice cloud formation in the atmosphere, and the molecular adsorption kinetics at water/mineral interfaces, aiming to reduce the consumption of resources and environmental impacts of the oil sands development in Canada. Super-Resolution Microscopy: From Single-Molecule Spectroscopy to Protein Assemblies in Whole Cells       In the past several years, scientists have developed several approaches to overcome the diffraction limit and revolutionized optical microscopy. Single-molecule localization microscopy (SMLM) has provided an unprecedented spatial resolution of 10 nm. SMLM techniques activate and locate only a sparse subset of fluorophores at any given time. Because the images of each fluorophore is well separated spatially, the location of each fluorophore can be determined with a very high accuracy. A super-resolution image can be reconstructed using the x-, y-, and z- coordinates of fluorophores obtained from a series of diffraction-limited images. These 3D point clouds cannot be analyzed using traditional image analysis tools. We will employ machine learning to mimic the modulation transfer function of human eyes to analyze the 3D point clouds produced by SMLM. Since in many aspects humans outperform any pattern-recognition machines, this approach will lead to an effective protein cluster analysis tool without any user input.       Additionally, we propose to implant a sample stabilization system into a structured illumination microscope (SIM) to remove artifacts due to sample drift. Compared to SMLM, SIM has several advantages for 3D imaging. It has a much higher speed, uses a relatively low illumination power, and is easily available for multicolor imaging. The proposed SIM will enable us to obtain reliable super-resolution images for large cells, such as cardiac myocytes, which has a mean length of 100 µm and a width of 18 µm. The microscope will allow us to image protein organizations in whole cells and make exciting discoveries.