The first physical property information I ask for when I take on a crystallization problem is the solubility curve. My request usually results in a deer-in-the-headlights look or — worse — the presentation of one from 1910 that the plant never has checked. A related issue is the meta-stable zone width and the potential for polymorphism or solvation, especially with solutes that cross at a given temperature. My 2005 article “Don’t let Phase Changes Faze You,” highlights the value of a solubility curve.
You must take eight specific steps to ensure the solubility curve represents your product and operations.
1. Define the solvent and solute. Slight changes in composition can significantly alter the solubility and meta-stable zone width. Your product may not match what’s reported in the chemistry books. You may need to account for batch-to-batch variations in the operation. Consider on-line instrumentation to maintain the correct concentration in the crystallizer.
2. Identify chemical reactions. Crystallization often is confused with precipitation; they do share a similar phase change. However, undesirable reactions can occur due to a shift in solubility where an impurity becomes more concentrated and reacts with either the solute or solvent. When reactions occur, you may need to change the method of generating supersaturation.
3. Identify phase transitions. Polymorphic transformation from a meta-stable polymorph to a stable polymorph will reduce the solubility and alter the particle size. Solvates are less soluble and have slower dissolution characteristics. Both of these changes can cause a crystallizer to produce an undesirable — often poorly filtering — product.
4. Ensure equilibrium. One curve does not represent the system. Some solutes may take years to reach equilibrium. So, knowing that the system is saturated is important. For example, it may be possible to start with a solution saturated at room temperature, cool it and then reheat it to just below the starting temperature to observe the crystallization process. However, if the rate of desupersaturation is slow (because of slow kinetics or not enough crystal surface area), the solution still may be supersaturated. The best way to demonstrate equilibrium is to show that both undersaturation and oversaturation coincide with the same saturation concentration. The rate of approach to equilibrium depends on the available surface area of crystals exposed to the solvent. If you know the mass of excess solute and the particle size distribution, you can estimate the surface area, which gives a crude idea of the kinetics.
5. Maintain temperature control. Perhaps the least appreciated aspect in determining solubility is temperature control. For steep solubility curves, minor temperature fluctuations can damage the results, especially when studying the meta-stable zone width. Most systems can tolerate 0.1°C variation, which is a practical limit for industry.
6. Provide sufficient agitation. During any solubility study, the solution must be fully mixed. However, exercise caution in selecting an agitator; you must avoid secondary nucleation around it.
7. Sample and handle properly. Both the solids phase and solution require analysis to develop the solubility curve. If possible, obtain the particle size distribution to ensure equilibrium. I recommend turning off the agitator and allowing the crystals to settle. Then, take a sample of the crystal-free liquor to determine the solute concentration. Finally, remove the solids by filtration. This last step requires great care to avoid altering the sample through evaporation, crystal growth or dissolution. In-situ sampling, if possible, is best.
8. Determine the meta-stable zone width. Knowing the width helps in selecting a crystallizer as well as in establishing how to operate it. Evaluate at least three temperatures or solute concentrations following this procedure:
• Prepare a saturated solution at a given temperature with a few excess crystals.
• Heat this sample until you observe no solids. (Usually this only requires raising the temperature a couple of degrees.)
• Rapidly cool the sample to observe crystal formation.
• Note the temperature at which crystals form.
• Reheat the sample to the original temperature.
• Finally, record the temperature at which solute dissolves.
The resulting plot of the lower temperature where crystals form defines the meta-stable zone width. The upper temperature verifies the solubility curve.
Solving crystallization problems is a breeze with this information. The secret is recognizing that you need these important physical property details.
TOM BLACKWOOD is a Chemical Processing Contributing Editor. You can email him at TBlackwood@putman.net