The practice of chemical engineering underpins chemical processing. Universities teach design methods for unit operations such as distillation and heat transfer. However, colleges seldom cover solids processing operations; if any discussion takes place, it only focuses on basic particle physics issues related to drag and turbulence. Yet, many processors extensively handle solids. For instance, I once worked for a company that made over 70% of its final products in a particulate solid form.
My lack of schooling about solids hindered my troubleshooting. Don’t get me wrong — many of the fundamentals I learned were very useful in analyzing and understanding what was causing a problem with our products. However, I found many exceptions that were hard to explain, because particulate solids always are a two-phase mixture, which gives them an additional degree of freedom. Besides physical changes, chemical interactions may complicate the situation.
One example is agglomeration of particles, often an unwelcome result of handling. Rolling particles over one another can increase the surface energy or impart a charge to the surface. If water is in the fluid or atmosphere around the particles, the surface charge will attract the dipole of the water and draw the particles closer together; Van der Waals forces can continue to hold the solids together even after the water has been removed. Many granulators take advantage of this to convert a fine powder into an attrition-resistant particle. However, I determined many years ago that if you’re trying to fill a bulk bag this may not be a desirable outcome and can promote other undesirable effects.
Our plant produced a very fine particle so it would dissolve rapidly for the customer. However, filling the bag was difficult due to the low bulk density, requiring an extended period of time exposed to the un-air-conditioned plant atmosphere in the southern U.S. The warm solids drove off attracted moisture but allowed the bulk material to clump during storage. Had this been the only problem caused by clumping, the story would end there. However, moisture trapped in the bulk bag condensed and dissolved some of the solids as the bags went though different temperature environments. This enabled particles to recrystallize as a different non-soluble polymorph. We redesigned the process to eliminate the moisture contact, solving the problem. (For information on the causes of clumping and the problems that can occur, see: “Clamp Down on Clumping.”)
Another, quite different example of a solid-state transformation that screamed polymorph involved a drying operation. The prior experience of the company’s research group suggested a process could be scaled up to avoid over-heating of the solids by using a dryer with a fixed inlet temperature and residence time. Occasionally, a batch would have slightly higher total starting solvent content with an imperceptible difference in final total solvent. However, some particles weren’t as dry as the others when the starting solvent was high. Downstream milling released some of the trapped solvent, heated the solids and caused an undesirable polymorph to form through a recrystallization process. Adding a temperature measurement in the dryer permitted starting the fixed drying period only after reaching a specific temperature (close to when free solvent has been removed). The final solvent was uniform throughout the particulate solids, avoiding the polymorph during milling.
So, what are the lessons from these experiences? First, if it screams polymorph, pay attention; fine organic particulate solids are very likely to exhibit this trait. Second, the amount of information needed for a particulate solids operation is more extensive than for others and often involves chemistry in addition to the process mechanics. Third, prior experience may not suffice unless you understand the fundamental details (drying kinetics, particle shape/strength and zeta potential); you must investigate these details during process development.
TOM BLACKWOOD is a Chemical Processing Contributing Editor. You can email him at