【作者】 周权宝；

【导师】 曾德文；

【作者基本信息】 湖南大学 ， 无机化学， 2012， 博士

【摘要】 多元复杂氯化物电解质溶液体系如CuH₂₂–MH₂n–H₂O（M=Li，Na，K，Mg，Ca），由于存在强水化，强缔合，以及强水化和缔合作用竞争，溶液结构与性质复杂，用以往的电解质热力学模型如Pitzer模型等计算时不仅描述能力差、预测困难，且无法解释溶液结构与性质之间的联系。为了解决这一难题，我们在Stokes–Robinson逐级水化模型基础上，引入离子缔合，充分考虑水化、缔合作用竞争影响，建立了以系列水化离子，水化离子对物种为基础，综合考虑了离子间的长程静电作用以及水化、缔合短程作用，并以化学反应为基本特征的化学反应模型。化学反应模型不仅数学形式简单，高度对称，有坚实的理论基础；而且通过推理证明它严格遵守Gibbs–Duhem方程。当电解质溶液达到热力学平衡时，体系的自由能处于最低状态。利用这一原理化学反应模型通过自由能最优化可求得各个物种在化学或相平衡下的浓度分布，进而计算电解质溶液的系列热力学性质。由于直接参数拟合过程具有计算复杂，规模大，变量数众多，结果可靠性低等困难。我们采用嵌套二步优化构架设计，引入分区粒子群最优化算法进行参数拟合，突破了上述困难，模拟测试表明该算法寻优成功率，计算结果的可靠性明显高于原来的算法。作为应用实例，我们用化学反应模型计算了氯化物LiH₂，NaH₂，KH₂，MgH₂₂，CaH₂₂，CuH₂2的六种二元体系以及15种三元体系的热力学性质。计算表明化学反应不仅可以准确描述二元体系的热力学性质，而且补充少量的三元体系参数也可以准确地描述三元体系的热力学性质。若不生成盐盐缔合物种，化学反应模型仅用相应二元体系模型参数就可以直接预测三元体系的热力学性质，例如所预测的LiH₂–CaH₂₂–H2O，LiH₂–MgH₂2–H2O三元体系aw，计算值的平均偏差分别为0.0027和0.0029，最大偏差分别为0.0065和0.0081；所预测的LiH₂–CaH₂₂–H₂O，LiH₂–MgH₂₂–H2O三元体系溶解度等温线也与实验数据高度接近。特别地，化学反应模型所预测的痕量CuH₂₂在LiH₂–H₂O体系中含Cu物种浓度分布曲线与Brugger紫外光谱解析结果保持高度一致，所预测的痕量CuH₂₂在NaH₂–H₂O体系中含Cu物种浓度分布曲线与Haung用热力学模型计算的结果也非常相近。通过大量的计算发现，化学反应模型实际充当了桥梁作用，沟通溶液的微观结构与宏观热力学性质之间的联系。通过它，可以帮助我们理解电解质微观结构对溶液热力学性质影响机制。水化作用，缔合作用以及水化缔合作用竞争在溶液中扮演着重要角色。它不仅在二元体系中决定着水的活度曲线走向和电解质在水溶液中的溶解度；而且在三元体系中，对水的活度，溶解度等温线，以及物种浓度分布曲线造成关键性的影响。例如，在三元体系中，痕量CuH₂₂在MH₂n–H2O（M=Li, Na, K, Ca）体系中的含Cu物种的浓度分布受阴离子与水分子竞争配位作用的影响，这种部分配位作用也是水化缔合作用竞争的一种形式。由于不同氯化物MH₂n供氯离子不同，其阴离子和H₂O分子在Cu–H₂缔合物上的竞争配位作用也存在差别，使痕量CuH₂₂在MH₂n–H₂O（M=Li, Na, K, Ca）体系中的含Cu物种的浓度分布曲线呈有规律的变化。当盐盐缔合作用很强时，三元体系中可能生成盐盐缔合物种，此时，三元体系的热力学性质不能简单地用二元体系的模型参数直接预测，而必须用三元体系溶解度等温数据来拟合盐盐缔合物种的参数，才能较好地描述三元体系溶解度等温线。用水化缔合作用竞争可以很好解释为什么缔合物的形成使得固相溶解度在共晶点附近比预测结果偏大的原因。无论是阴离子与H2O分子竞争配位，还是生成盐盐缔合物种，水化缔合作用竞争都会使得三元体系等水活度线呈不同程度的弯曲。本工作开发的化学反应模型能很好地解释由于上述水化，缔合作用，水化缔合作用竞争的影响所引起的一系列现象，表现出强大的分析能力。更多还原

【Abstract】 Strong hydration, association and competition between hydration and association in themultiple metal chlorides aqueous solution of CuH₂2–MH₂n–H2O （M=Li, Na, K, Mg, Ca） leadto complexity of solution structure and properties inlcuding component asctivities, solubility.Developed thermodynamic models up to now for electrolyte solution, such as the Pitzermodel, cannot describe or predict their thermodynamic properties, and cannot explain therelation of the solution structure and properties either.To overcome this difficult, based on the Stokes–Robinson stepwise hydration model, weintroduced ion–association interaction into the stepwise hydration model, develop a newthermodynamic model is called reaction model （RM）. The reaction model consists of thelong–range electrostatic interaction and short interaction of hydration and association. It is notonly simple in mathematical, high symmetrical, but also built on solid theoretical foundationand strictly abides Gibbs–Duhem regulation. There is a minimum value of free Gibbs energyof the electrolyte solution to reach a thermodynamic equilibrium. By this principle, reactionmodel obtains the various species concentration under chemical or phase equilibriumcondition by energy optimization, then further calculate series thermodynamic properties withthem. Because of the large number of variables, complicated, large scale computation and lowreliability of results, many difficults were encountered in direct parameter fitting. To solvethis problem, a framework design of the nested two–step optimization was adopt in this work.An algorithm named Partitioning Quantum–behaver Particle Swarm Optimization （PQPSO） isapplied in parameters evauation. The simulation tests show that the success rate and thereliability of the optimization results have improved significantly than primary method.As examples of application, we use the reaction model to calculate the thermodynamicproperties of six binary aqueous systems and15ternary aqueous systems of metal chloridesLiH₂, NaH₂, KH₂, MgH₂2,CaH₂2and CuH₂2. Calculations results show that the reaction modelcan accurately describe the thermodynamic properties of not only the binary system, but alsothe ternary system if adding small number of the ternary parameters. The reaction model canpredict the thermodynamic properties of the ternary system only with the correspondingbinary parameters, such as the predicted awvalue of LiH₂–CaH₂2–H2O and LiH₂–MgH₂2–H2Osystem, the corresponding calculated average deviations are0.0027and0.0029, the maximumdeviation are0.0065and0.0081, respectively. The predicted isotherm of LiH₂–CaH₂2–H2Oand LiH₂–MgH₂2–H2O system are highly consist with the experimental data. In particular, ourpredicted Cu-containing species concentration distribution curve of trace CuH₂2in LiH₂–H2Osystem is highly consistent with Brugger’s analytical results of UV spectra; at the same time, the predicted same result in the NaH₂–H2O system is very similar to thermodynamic modelcalculation results by Haung.By a large amount of computing, it is found that the reaction model can act as a bridge,which connects the microscopic structure and the thermodynamics properties. It helps usunderstand the mechanism how the microscopic structure characteristic of electrolytesolutions affect their thermodynamics properties.In some electrolyte aqueous solutions, hydration, association and the competition betweenhydration and association play important role to the thermodynamics properties. Theseinteraction not only seriously impact the binary water activities, determine the electrolytesolubility with the different structural characteristics in aqueous solution; but also cause astrong effect on water activities, solubility isotherms in the ternary system, as well as on thespecies concentration distribution curve. For example, the Cu–containing speciesconcentration distribution curve of the trace CuH₂2in MH₂n–H2O（M=Li, Na, K, Ca） systemsis affected by the coordination competition between anions and water molecules which is alsoa form of competition of hydration and association. Different H₂–donation ability of metalchloride MH₂nleads to that the competition coordination of the anions and H2O molecules toCu–H₂complex is also differences, so the concentration distribution curves of Cu–containingspecies of trace CuH₂2in MH₂n–H2O （M=Li Na, K, Ca） systems present variation regularly.When a salt–salt association species will generate in the ternary system due to the strongsalt–salt association, thermodynamic properties of the ternary system can not be directlypredicted simply with the binary parameters. One must evalue the parameters of salt–saltassociation species with fitting solubility isotherm data to better describe ternary solubilityisotherms.With the competition between hydration and association, it can also well explain why theformation of salt–salt association complex makes the solubility of solid phase larger than thepredicted results near the eutectic points. Either the competition coordination of the anion andthe H2O molecule, or the formation of salt–salt association species, the competition ofhydration and association will make iso awlines in the ternary system bending in differentdegrees.The reaction model developed in this work can well explain above described seriesphenomena caused by the impaction of hydration, association and the competition ofhydration and association, exhibits a powerful analytical ability.更多还原