Novikov E, Hofkens J, Cotlet M, Maus M, De Schryver F C, Boens N
Department of Chemistry, Katholieke Universiteit Leuven, Belgium.
Spectrochim Acta A Mol Biomol Spectrosc. 2001 Sep 14;57(11):2109-33. doi: 10.1016/s1386-1425(01)00500-5.
Up to now, single molecule fluorescence experiments were performed by dividing the time into a set of intervals and to observe the number of fluorescence photons arriving in each interval. It is obvious that the detected photons carry less information than the arrival times of the photons themselves. From the arrival times, one can still calculate the number of photons in any user-defined interval; whereas, when only the number of photons in an interval are recorded, information about their positions in time is lost. Therefore, we present a new analysis method of single molecule fluorescence data based on the positions in time of the detected fluorescence photons. We derive mathematically different statistical characteristics describing the single molecule fluorescence experiment assuming an immobilized molecule. The theory of point processes using the generating functionals formalism is ideally suited for a consistent description, linking the statistical characteristics of the excitation and detected photons to the statistical characteristics of the single motionless molecule. We then use computer-generated data sets mimicking the single molecule fluorescence experiment to explore the parametric estimation of mono- and bi-exponential single molecule impulse response functions (SMIRFs) via the following statistical characteristics: the probability density distributions (pdd) of the single and first photocount time positions in a user-defined detection interval, the probability distribution of the number of photocounts per user-defined detection interval, the time correlation function and the pdd of the time interval between two consecutive photocounts. It is shown that all of the above characteristics ensure a satisfactory recovery of the decay time of mono-exponential SMIRFs for a broad range of excitation intensities and widths of user-defined detection intervals. For bi-exponential SMIRFs, the selection of the experimental conditions is more critical and dependent on the detection procedure. At lower excitation intensities it is advantageous to use the pdds of the single and first photocount time occurrences in the user-defined detection interval. To show the practical usefulness of the new analysis method, series of photon arrival times from immobilized single molecules of DiI and rhodamine 6G were analyzed to estimate triplet lifetimes and intersystem crossing yields.
到目前为止，单分子荧光实验是通过将时间划分为一组间隔并观察每个间隔内到达的荧光光子数量来进行的。显然，检测到的光子携带的信息比光子本身的到达时间要少。从到达时间，人们仍然可以计算任何用户定义间隔内的光子数量；然而，当只记录一个间隔内的光子数量时，关于它们在时间上的位置信息就丢失了。因此，我们提出了一种基于检测到的荧光光子时间位置的单分子荧光数据分析新方法。我们从数学上推导了描述固定分子单分子荧光实验的不同统计特征。使用生成泛函形式的点过程理论非常适合进行一致的描述，将激发光子和检测光子的统计特征与单个静止分子的统计特征联系起来。然后，我们使用模拟单分子荧光实验的计算机生成数据集，通过以下统计特征来探索单指数和双指数单分子脉冲响应函数（SMIRF）的参数估计：用户定义检测间隔内单个和第一个光计数时间位置的概率密度分布（pdd）、每个用户定义检测间隔内光计数数量的概率分布、时间相关函数以及两个连续光计数之间时间间隔的pdd。结果表明，对于广泛的激发强度和用户定义检测间隔宽度，上述所有特征都能确保单指数SMIRF衰减时间的满意恢复。对于双指数SMIRF，实验条件的选择更为关键且取决于检测程序。在较低激发强度下，使用用户定义检测间隔内单个和第一个光计数时间出现的pdd是有利的。为了展示新分析方法的实际实用性，我们分析了来自固定化的DiI和罗丹明6G单分子的一系列光子到达时间，以估计三重态寿命和系间窜越产率。
