【作者】 张相洋；

【导师】 袁渭康； 周兴贵； Gilles Fevotte；

【作者基本信息】 华东理工大学 ， 化学工程， 2010， 博士

【摘要】 溶液结晶作为一种高效、低能耗、少污染的分离和提纯技术,在医药、食品、生物等工业生产中得到了广泛的应用。尤其是其温和的操作条件特别适合于作为提纯手段来生产医药活性组分(APIs)。目前85%以上的医药产品的生产过程中含有结晶单元。溶液结晶过程极其复杂,结晶工艺过程的选择(包括产生过饱和度的方式如溶析、蒸发、冷冻,溶剂的选择)和结晶过程变量的控制(包括混合与搅拌,溶析、蒸发、冷冻速率,晶种的添加等)都将对晶体成核和生长,破碎和聚并产生重要影响,并由此影响结晶过程效率(能耗和收率等)和产品的质量(晶体的纯度、内在结构和外部形态、粒径及分布等)。工业结晶过程中杂质的存在不可避免。在药物生产过程中,产品的纯度往往是衡量产品质量的重要标准。这是由于通常状况下,微量的杂质存在即能大大的降低药物的药效,有时甚至会产生有害的副作用。此外,作为进行结晶器设计和工艺过程优化的关键,晶体成核和生长动力学会显著的受到杂质的影响,并由此改变结晶过程效率和最终产品质量。因此,通过研究杂质影响结晶过程的机理,并在此基础上采用结晶过程控制的方式,从而达到提高产品质量的目的业已成为结晶工艺与过程设计的关键。本文以乳酸锌间歇冷却结晶为研究对象,考察了杂质对其结晶行为的影响。具体的研究内容和结果如下：(1)选取HCl、NaCl以及四种由上游生产过程中引入的有代表性的有机酸或盐为杂质,考察了pH值和杂质对乳酸锌溶解度的影响。并研究了苹果酸杂质对乳酸锌超溶解度的影响。结果显示,乳酸锌的溶解度受溶液pH值、杂质与溶质分子间相互作用力的双重影响,但相比而言,pH值效应对溶解度的影响较小。在此基础上本文提出了不同浓度杂质存在时的乳酸锌溶解度预测模型；苹果酸杂质的引入使得乳酸锌晶体在水中的溶解度和超溶解度均增加,最终使得结晶介稳区的位置上移,且在低温区域结晶介稳区的宽度减小,在高温区域结晶介稳区的宽度变化不大；(2)考察了苹果酸、丁二酸和草酸钠杂质对乳酸锌初次成核动力学以及晶核形态的影响。结果表明：苹果酸和丁二酸杂质的引入增加了乳酸锌溶解度和成核能垒,使得介稳区位置上移宽度增加,而草酸钠杂质的作用相反；晶体成核速率受溶液浓度和分子间结合能双重作用,在实验溶液过饱和度范围内,含苹果酸和丁二酸杂质的乳酸锌晶体成核速率受结合能控制,成核速率低于纯乳酸锌的成核速率。含草酸钠杂质的乳酸锌晶体成核速率在低过饱和度时受结合能控制,成核速率高于纯乳酸锌体系,高过饱和度时受溶液浓度控制,成核速率低于纯乳酸锌体系；(3)提出了一种利用电导率仪在线测量冷却结晶过程中溶液浓度的方法,并根据该方法在实验中对乳酸锌冷却结晶过程进行了溶液浓度的在线测量,通过与离线法测量结果进行对比,证明了模型可靠；在此基础上,提出了一种利用结合改进的Kubota模型和PBE模型实现溶液冷却结晶过程中动力学研究的方法。主要特点是利用Kubota模型考察不同温度和杂质浓度存在条件下的临界过饱和度,由此确定结晶过程中的实际过饱和度,并估算成核和生长动力学。结果显示,该模型能较好的对实验数据进行拟合；苹果酸杂质的引入抑制了乳酸锌晶体的成核和生长过程,且这种抑制作用随苹果酸杂质含量的增加而增强,对此可以解释为苹果酸分子通过与乳酸锌分子形成金属一苹果酸络合物抑制了乳酸锌分子自身在晶体表面的吸附,使得晶体成核和生长速率变慢。(4)以乳酸锌结晶过程为研究对象,通过添加晶种降温结晶的方式考查了苹果酸杂质存在条件下各种结晶过程参数对乳酸锌结晶过程、最终产品纯度和形态的改变。结果表明,在添加晶种的结晶过程中,不同的结晶操作参数如晶种种类、数量、降温速率、添加晶种的温度和最终结晶温度等均会对产品纯度产生一定影响：降温速率和添加晶种温度通过改变结晶过程过饱和度影响晶体成核和生长速率的变化,从而改变最终产品的杂质含量和粒度；晶种种类和数量通过其所提供的总晶种表面积来改变晶体表面可供杂质吸附活性位的多少来影响最终产品的纯度；苹果酸杂质存在条件下各种操作结晶操作参数对乳酸锌晶体结构和形态并没有明显的影响。(5)采用分子模拟技术研究了苹果酸和丁二酸杂质对乳酸锌晶体各晶面的作用。结果显示,苹果酸杂质分子和丁二酸杂质分子在乳酸锌晶体表面的结合能均高于乳酸锌分子在其晶体表面的结合能；苹果酸杂质在乳酸锌晶体(002)和(100)面的结合能相比(110)面稍高,乳酸锌晶体在(110)面方向上生长速率较快,因此变得更为细长；丁二酸杂质在乳酸锌晶体(002)和(110)面的结合能显著强于(100)面,并由此强烈抑制了(002)和(110)面的生长,由此使乳酸锌晶体呈片状。以上结论与实验结果一致。更多还原

【Abstract】 Crystallization is an important separation and purification process used widely in food, fine chemical and pharmaceutical industries. Especially in the pharmaceutical industry, crystallization processes are often carried out as the final purification step in the manufacturing of active pharmaceutical ingredients (APIs), enabling one to obtain solid products with high purity at low costs.Crystallization is a very complicated process as it involves many concomitant phenomena such as nucleation, growth, breakage and agglomeration of crystals, and the purity and physical properties (morphology, crystal size distribution, bulk density, product filterability, and dry solid flow properties) of the crystals are significantly influenced by the selection of the crystallization process (e.g. cooling, evaporation, or antisolvent addition). The control of the manipulated "input" variables during the crystallization process (e.g. stirring, cooling or evaporation rates, seed amount, rate of antisolvent addition, etc.) also plays a key-role in the determination of the final dispersed solid properties. Despite the long history and widespread application of crystallization, a large number of unsolved problems remain concerning modeling, design and control of the industrial processes, which requires an in-depth and improved understanding of the mechanisms of nucleation, growth, breakage and agglomeration, and of the influence of the process variables on these mechanisms.Industrial solutions are almost invariably impure, by any definition of the term, and in many cases, small amounts of impurities present in the solution can have a dramatic effect on solubility, nucleation, crystal growth, and morphology. Furthermore, product purity is probably the most important index of product quality, and is obviously of paramount concern for pharmaceuticals because small amounts of impurities are likely to result in reduced drug efficacy and harmful side effects. Understanding the effect of impurity on the crystallization and controlling the process to reduce as much as possible the impurity content of the main crystallizing product is therefore of significant engineering importance.In this dissertation, crystallization behaviour of zinc lactate in presence of impurities was investigated. The main work and results are listed as follows:(1) NaCl, HC1 and four other generic organic acid/salt including succinic acid, citric acid, malic acid and sodium oxalate are selected as impurities to investigate the effects of pH and impurities on the solubility of zinc lactate. Moreover, effect of malic acid on the position and width of zinc lactate metastable zone was also studied. The results show that the solubility of zinc lactate is a combined effect of pH and intermolecular interaction between impurity and host species which is mainly due to the formation of metal-organic complexes. In our system, the solubility of zinc lactate slightly increases with decreasing pH. And the solubility of zinc lactate increases in presence of organic acid, while decreases in presence of sodium oxalate. On this basis, a model is proposed for the prediction of effects of impurities on solubility at different initial concentration, which is able to fit the experimental data satisfactorily. Besides, the presence of malic acid increases both the solution solubility and supersaturation limit, and consequently, changes the position and width of metastable zone.(2) Primary nucleation kinetics of zinc lactate in presence of impurities including malic acid, succinic acid and sodium oxalate were investigated. The results show that the solubility, energy barrier for the formation of nuclei and intermolecular binding energy are influenced in presence of impurities. The position and width of metastable zone are consequently changed. The primary nucleation rate depends on the combination effects of molecular binding energy and solute concentration, which will transform with increasing the supersaturation in aqueous solution.(3) A method concerning on-line monitoring of solution concentration during cooling crystallization process by using conductivity meter was proposed. Furthermore, four cooling crystallization processes of zinc lactate were performed, and the variations of concentration were measured on-line based on this method. And the accuracy was verified by comparing with the experimental results measured off-line by HPLC. On this basis, a simple numerical simulation model, which can be used to predict the nucleation and crystal growth kinetics during crystallization process, was proposed by combining the revised Kubota and PBE model. The results show that this model was successfully adopted to obtain kinetic expressions for both nucleation and crystal growth as a function of supersaturation. Moreover, the presence of malic acid led to a reduction in the overall nucleation and growth kinetics of zinc lactate. And the reductions to the nucleation and growth kinetics were more pronounced when the malic acid concentration increased.(4) Batch cooling crystallization of zinc lactate in the presence of malic acid was carried out at different operating conditions to show the influence of secondary nucleation and the rate of crystal growth on the purity, crystal size, yield and crystal structure of the final products. The results show that appropriate amount and size of seed crystals should be chosen to get the final product with high average volume crystal size and purity. Moreover, other operating parameters such as seeding temperature, cooling rate and terminated temperature were also shown to exert some influence on the crystallization process. Concerning the crystal shape, crystals grown in impure media appeared to be substantially different.(5) Adsorption of impurities including malic acid and succinic acid on the crystal surface of zinc lactate was investigated using molecular modeling. The results show that the binding energies of malic acid and succinic acid molecules on the surface of zinc lactate are higher than that of zinc lactate molecules. With respect to malic acid, the binding energies of which on surface (002) and (100) of zinc lactate are slightly higher than that on surface (110), the crystal then mainly growth in the direction of surface (110), and the crystal habit exhibits rather thin anisotropic shape. While the binding energies of succinic acid on surface (002) and (110) of zinc lactate are significantly higher than that on the surface (100) of zinc lactate. The crystal habit therefore exhibits a laminar shape. All these predictions are conincident with the results obtained from experiments.更多还原