Two-dimensional supercritical fluid chromatography/mass spectrometry for the enantiomeric analysis and purification of pharmaceutical samples

A new analytical two-dimensional supercritical fluid chromatography/mass spectrometry system (2D SFC/SFC/MS) has been designed and implemented to enhance the efficiency and quality of analytical support in drug discovery. The system consists of a Berger analytical SFC pump and a modifier pump, a Waters ZQ 2000 mass spectrometer, a set of switching valves, and a custom software program. The system integrates achiral and chiral separations into a single run to perform enantiomeric analysis and separation of a racemic compound from a complex mixture without prior clean up. The achiral chromatography in the first dimension separates the racemate from all other impurities, such as un-reacted starting materials and by-products. Mass-triggered fractionation is used to selectively fractionate the targeted racemic compound based on its molecular weight. The purified racemate from the achiral chromatography in the first dimension is then transferred to the chiral column in the second dimension to conduct the enantiomeric separation and analysis. A control software program, we coined SFC2D, was developed and integrated with MassLynx to retrieve acquisition status, current sample information, and real time mass spectrometric data as they are acquired. The SFC2D program also monitors the target ion signal to carry out mass-triggered fractionation by switching the valve to fractionate the desired peak. The 2D SFC/SFC/MS system uses one CO(2) pump and one modifier pump for both first and second dimension chromatographic separations using either gradient or isocratic elution. Similarly, a preparative 2D SFC/SFC/MS system has been constructed by modifying an existing Waters preparative LC/MS system. All components except the back pressure regulator are from the original LC/MS system. Applications of the 2D SFC/SFC/MS methods to the separation and the analysis of racemic pharmaceutical samples in complex mixtures demonstrated that an achiral separation (in first dimension) and a chiral separation (in second dimension) can be successfully combined into a single, streamlined process both in analytical and preparative scale.     

Multilinear gradient elution optimization in reversed-phase liquid chromatography based on logarithmic retention models: application to separation of a set of purines, pyrimidines and nucleosides

The analytical solutions of the fundamental equation of the multilinear gradient elution are derived in two cases, when the dependence of the logarithm of the solute retention (lnk) upon the volume fraction of organic modifier (φ) is a three-parameter logarithmic expression, and when a simple linear relationship between lnk and lnφ is adopted. The derived theoretical expressions for retention times under multilinear gradient conditions are embodied to simple algorithms for fitting gradient data and especially for resolution optimization. Their performance was examined by using a mixture of 16 model compounds chosen among purines, pyrimidine and nucleosides in eluting systems modified by acetonitrile. It was found that the accuracy of the predicted gradient retention times is very satisfactory even if the simple logarithmic expression for the retention behavior of solutes, i.e. the linear dependence of lnk upon lnφ, is used.     

基于拓扑优化的声学结构材料分布设计

本文针对结构的声学设计问题进行研究,通过优化两种不同的材料在结构设计域内的拓扑分 布来最小化谐振结构所产生的声场中指定参考面/参考域内的声压。在研究中假定结构为线弹性小变 形结构,材料阻尼为Rayleigh阻尼,声学介质为无粘、可压缩、小扰动流体。对结构响应采用有限 元格式进行计算,对声场采用基于Helmholtz积分的边界元格式进行计算,由于声场在无穷远自由边 界的无反射条件在边界积分中能自动得到满足,该格式特别适合于具有开放边界的声场计算。建立 了结构有限元-声场边界元格式的耦合系统拓扑优化模型,导出了耦合系统敏感度分析的一般格式及 伴随格式。数值算例验证了所提出的结构-声学耦合系统优化方法的有效性和可靠性,并揭示了基于 声学准则的拓扑优化结果的有关特性 关键词边界积分,结构声学耦合系统,拓扑优化,敏感度分析,伴随方法     

杆件分组和网格密度对网壳截面优化结果影响

针对跨度为50\,m, 矢跨比为1/5的单层凯威特K6型球面网壳结构, 对6种杆件分组方式、 7种网格划分密度的42个网壳模型进行截面优化分析, 得出了与以往截面优化不同的、优化 效果更好的杆件分组方式和网格划分密度取值建议, 优化过程充分考虑了网壳结构强 度、刚度和整体稳定性, 使优化结果更具有工程实用价值. 可为工程设计人员提 供有益的参考.     

混凝土复合介质爆炸的不确定优化设计及其高效求解方法

研究不确定因素对炸药在混凝土复合介质中爆炸的影响,并在不确定因素的影响下进行优化设计。用区间数模型来描述系统中的不确定参数,建立相应的区间数优化模型;通过高维代理模型的引入,以及对其中二层嵌套优化问题进行直接解耦,得到了一种高效的求解方法。研究表明：算法经过8个迭代步收敛到最优解,最优解得到的最佳炸药装填深度约为64cm,最佳姿态角约为30°;经过三个迭代步就能迅速定位到最优解附近。算法保证了在最大化毁伤面积的同时,尽量地追求不确定因素影响最小,即综合考虑了毁伤最大和偏差最小。该不确定优化设计方法能有效地推广到相关领域的优化设计,为处理爆炸系统的装药设计和爆炸初始状态参数设计等方面提供了新思路。     

基于混沌粒子群算法的Tikhonov正则化参数选取

Tikhonov正则化方法是求解不适定问题最为有效的方法之一,而正则化参数的最优选取是其关键.本文将混沌粒子群优化算法与Tikhonov正则化方法相结合,基于Morozov偏差原理设计粒子群的适应度函数,利用混沌粒子群优化算法的优点,为正则化参数的选取提供了一条有效的途径.数值实验结果表明,本文方法能有效地处理不适定问题,是一种实用有效的方法.     

无约束优化的一个充分下降共轭梯度算法

在修正PRP共轭梯度法的基础上,提出了求解无约束优化问题的一个充分下降共轭梯度算法,证明了算法在Wolfe线搜索下全局收敛,并用数值实验表明该算法具有较好的数值结果.     

基于两时段的WCVaR风险分析及其应用

针对随机变量的分布信息不完全的情况下,提出了两时段的Worst-Case Conditional Valueat-Risk(WCVaR)指标,并建立了两时段的风险-利润投资组合优化模型,该模型是一高维问题,具有复杂的优化结构.在损失函数为线性以及随机变量为离散界约束分布的假设下,运用最优化对偶理论将具有多层min-ma...     

双侧风险度量方法及其在投资组合优化模型中的应用

现有的资产风险度量方法不能合理的反映收益的向上波动给投资者带来的风险感受,针对这一不足,本文提出了一种新的风险度量方法,这一方法综合考虑了投资者对于损失的规避和对超额收益的偏好,能够更为真实的反映投资者对于资产收益双侧波动的不同风险感受.同时本文结合新的风险度量方法给出了投资组合优化模型,并对模型的解从不同角度进行了分析.研究结果表明,新的风险度量方法可以为投资者提供更有效的投资决策依据,并且投资者的风险态度对于投资组合有效前沿和最优投资组合都有显著的影响.     

需求率为斜坡型的一类变质性物品最优库存策略

本文在考虑需求率服从斜坡型分布的情况下,研究了允许缺货且缺货完全回补、变质率服从威布尔分布、补货率为无穷、有限计划期内的库存模型,证明了最优补货策略的存在性,并给出了求解最优补货策略的算法.