Solid-Liquid Phase Equilibria and Crystallization of Disubstituted Benzene Derivatives
Total Page:16
File Type:pdf, Size:1020Kb
Royal Institute of Technology School of Chemical Science and Engineering Department of Chemical Engineering and Technology Division of Transport Phenomena Solid-Liquid Phase Equilibria and Crystallization of Disubstituted Benzene Derivatives Fredrik Nordström Doctoral Thesis Akademisk avhandling som med tillstånd av Kungliga Tekniska Högskolan i Stockholm framlägges till offentlig granskning för avläggande av teknologie doktorsexamen den 30:e Maj 2008, kl. 10:00 i sal D3, Lindstedtsvägen 5, Stockholm. Avhandlingen försvaras på engelska. i Cover picture: Crystals of o-hydroxybenzoic acid (salicylic acid) obtained through evaporation crystallization in solutions of ethyl acetate at around room temperature. Solid-Liquid Phase Equilibria and Crystallization of Disubstituted Benzene Derivatives Doctoral Thesis © Fredrik L. Nordström, 2008 TRITA-CHE Report 2008-32 ISSN 1654-1081 ISBN 978-91-7178-949-5 KTH, Royal Institute of Technology School of Chemical Science and Engineering Department of Chemical Engineering and Technology Division of Transport Phenomena SE-100 44 Stockholm Sweden Paper I: Copyright © 2006 by Wiley InterScience Paper II: Copyright © 2006 by Elsevier Science Paper III: Copyright © 2006 by the American Chemical Society Paper IV: Copyright © 2006 by the American Chemical Society ii In loving memory of my grandparents Aina & Vilmar Nordström iii i v Abstract The Ph.D. project compiled in this thesis has focused on the role of the solvent in solid-liquid phase equilibria and in nucleation kinetics. Six organic substances have been selected as model compounds, viz. ortho-, meta- and para-hydroxybenzoic acid, salicylamide, meta- and para-aminobenzoic acid. The different types of crystal phases of these compounds have been explored, and their respective solid-state properties have been determined experimentally. The solubility of these crystal phases has been determined in various solvents between 10 and 50 oC. The kinetics of nucleation has been investigated for salicylamide by measuring the metastable zone width, in five different solvents under different experimental conditions. A total of 15 different crystal phases were identified among the six model compounds. Only one crystal form was found for the ortho-substituted compounds, whereas the meta-isomeric compounds crystallized as two unsolvated polymorphs. The para-substituted isomers crystallized as two unsolvated polymorphs and as several solvates in different solvents. It was discovered that the molar solubility of the different crystal phases was linked to the temperature dependence of solubility. In general, a greater molar solubility corresponds to a smaller temperature dependence of solubility. The generality of this relation for organic compounds was investigated using a test set of 41 organic solutes comprising a total of 115 solubility curves. A semi-empirical solubility model was developed based on how thermodynamic properties relate to concentration and temperature. The model was fitted to the 115 solubility curves and used to predict the temperature dependence of solubility. The model allows for entire solubility curves to be constructed in new solvents based on the melting properties of the solute and the solubility in that solvent at a single temperature. Based on the test set comprising the 115 solubility curves it was also found that the melting temperature of the solute can readily be predicted from solubility data in organic solvents. The activity of the solid phase (or ideal solubility) of four of the investigated crystal phases was determined within a rigorous thermodynamic framework, by combining experimental data at the melting temperature and solubility in different solvents and temperatures. The results show that the assumptions normally used in the literature to determine the activity of the solid phase may give rise to errors up to a factor of 12. An extensive variation in the metastable zone width of salicylamide was obtained during repeated experiments performed under identical experimental conditions. Only small or negligible effects on the onset of nucleation were observed by changing the saturation temperature or increasing the solution volume. The onset of nucleation was instead considerably influenced by different cooling rates and different solvents. A correlation was found between the supersaturation ratio at the average onset of v nucleation and the viscosity of the solvent divided by the solubility of the solute. The trends suggest that an increased molecular mobility and a higher concentration of the solute reduce the metastable zone width of salicylamide. Keywords: Salicylic acid, m-hydroxybenzoic acid, p-hydroxybenzoic acid, salicylamide, m-aminobenzoic acid, p-aminobenzoic acid, solubility, solid-liquid equilibria, thermodynamics, activity of the solid phase, ideal solubility, activity coefficient, van't Hoff enthalpy of solution, solid state, solution properties, metastable zone width, primary nucleation, nucleation kinetics. v i Popular Scientific Summary Compounds in nature exist as gases, liquids and solids. These states are separated by temperature and pressure. When the temperature of a liquid decreases at constant pressure, it forms a solid. As an example, ice forms on a lake on a cold winter day. When a solid is placed in contact with a liquid, the solid will start to dissolve into the liquid. This is something we see when we for example add sugar to a cup of coffee. As we continue to add a solid to the liquid the solid will at some point no longer dissolve into the liquid. The solution is then said to be saturated and the concentration of the saturated solution is called solubility. In science, we refer to this condition as a state of equilibrium between the solid and the solution. When the solid is in contact with a liquid we also often call the liquid a solvent. The solubility depends on several factors. When we change the solvent or the solid, the solubility will also change. However, we have an additional factor to consider when looking at the solid. The structure of the solid is sometimes different. The element carbon is for example found in nature in different solid forms such as graphite and diamond. These two forms, or phases, of course are very different and also have different solubility. In the pharmaceutical industry, the drug is often found in different solid forms. When we make a tablet of the drug and administer it to humans it will dissolve when it comes in contact with the fluids in the intestines. The drug then travels into the blood's circulatory system giving its purposed medical effect. Since the solid forms of the drug have different solubility they will also dissolve at different rates. This will affect the medical response of the drug, sometimes dramatically. It is therefore very important that we know what the different solid forms of the drug are and how fast they dissolve. When a drug has been discovered it is produced through a series of chemical reactions. To purify the drug from side-products it goes through several industrial purification and separation processes. One of the last and very important processes is the crystallization process. This is when the drug is formed in a solid form from a solution. Crystallization is almost like the opposite of dissolution. We can perform a crystallization process in different ways, but usually we saturate a solution at a certain temperature and then lower the temperature until the solid starts to appear in the form of crystals. Depending on how we perform the crystallization different solid forms of the crystals may appear. The shape and size of these crystals may also be very different. Just as the solubility is important when the drug is given to humans, the solubility is also the most important factor in the crystallization process. To be able to control a crystallization process we need to know the solubility of the different solid forms and how they are related to the solvent and temperature. vii In this research project, six different solid compounds have been investigated together with up to nine different solvents at different temperature. In the first part of the project, we have looked at what different solid forms, or phases, are found among these compounds. We have found that two of the compounds only crystallize as one solid form, whereas the other four compounds crystallize in two to five different solid forms. The solubility of these different solid forms has been determined in different solvents from 10 to 50 oC. It soon became clear that the solubility of these solid forms is directly related to how solubility is dependent on temperature. This connection has not been explored much in the past. To understand why we see this relationship, we used thermodynamic theory, which describes how solubility is related to the properties of the solid and the properties of the solution. Although the science of this connection is interesting to explore we also obtain several practical applications that we can use as an aid in the crystallization process. We can for example more easily estimate how much of the drug is formed as solid depending on what solvent we use. We can also identify different solid forms just by looking at the solubility data. Since the relationship between the solubility and how it depends on temperature was so strong it was decided to investigate whether this connection is something that appears for organic compounds in general. In order to do that, we used a total of 41 different compounds and their solubility data in different solvents. We could verify this behavior and also construct a model that allows us to estimate what the solubility is at different temperature. In addition, it was observed that the solubility at different temperature was predictable to the extent that we could estimate the melting point of the solid. A common problem in the field of crystallization is not being able to measure how the properties of the solid and the properties of the solution affect the solubility. By using this connection and thermodynamic theory it was possible to separate these two properties from each other.