Drying crystalline active pharmaceutical ingredients (API) is an important operation for the production of consistent, stable, free-flowing materials for formulation, packaging, storage and transport. Generally, the materials being dried are solids that have been isolated by means of filtration following crystallization from water or from one or more organic solvents. As with most operations, drying is best viewed as part of an integrated process that includes these crystallization and isolation steps; changes in these operations can affect particle-size distribution, crystal form and moisture content, and they can have a significant impact on the drying efficiency.

Product drying is not a particularly energy-efficient process. Consider, for example, that it can take 5 to 10 times the amount of energy to remove a kilogram of solvent in a drying operation than in a distillation operation. Thus, it is important to remove as much solvent or moisture from the cake as possible beforehand, and the selection of isolation equipment (pressure filter, product centrifuge, etc.) has a major impact on accomplishing this goal.

It also is important to establish realistic drying specifications for these materials during product development -- for example, drying to “zero” moisture content is not practical -- which will, of course, depend on process requirements and material characteristics. Likewise, having robust analytical methods for monitoring drying or for product release is critical, as is establishing safe drying temperature ranges that will maximize drying efficiency without risking product decomposition, melting or agglomeration.

Equipment Selection

Enlarge this picture Figure 1. A number of dryer types are available for pharmaceutical applications, most of which are conductive (or contact) dryers, in which the product directly contacts the heated dryer surface.

A number of dryer types are available for pharmaceutical applications, most of which are conductive -- or contact -- dryers, in which the product directly contacts the heated dryer surface. These include the smaller, batch-style dryers such as tray, rotary cone, paddle and tumble dryers most often found in pharmaceutical pilot plants (figure 1). A carrier gas often is swept through these dryers to carry off the evaporated solvent.

In these types of systems, drying typically is carried out at reduced pressure. This is particularly effective for temperature-sensitive or air-sensitive products and safer for drying toxic and high-potency substances as well as those containing flammable solvents. Small units usually are heated electrically, but for increased safety in larger units, steam or another heat transfer medium is circulated through the dryer trays or unit jacket.

The most common dryer for laboratory and small-scale pilot work is the vacuum tray dryer (figure 2). It can be scaled up to moderate size, but space and efficiency limitations make further scale-up impractical. It is reliable and has no moving parts, but operation and cleaning are labor intensive. Also, because solvent wicking can cause a crust to form on the cake, the product often requires milling, screening, blending or other post-drying treatment to ensure homogeneity. These problems can be avoided by keeping the cake moving during drying, which is the major advantage of cone, paddle and tumble dryers and agitated combination filter-dryers. Agitation also can reduce cycle time by speeding evaporation.

Enlarge this picture Figure 2. The most common dryer for laboratory and small-scale pilot work is the vacuum tray dryer. Typical equipment associated with a kilo- or pilot scale vacuum tray dryer installation are shown.

As always, it is important to understand your operational needs and capacity requirements before selecting drying equipment. Work only with reputable vendors. They can provide a wealth of information about the advantages and disadvantages of various units for your application. Ease of cleaning is particularly critical in multi-use pilot equipment and current good manufacturing practices (cGMP) plants.

Product Characteristics. The properties of product from pilot drying equipment may be significantly different from that dried in laboratory vacuum tray dryers. This is particularly true of units that agitate the cake mechanically such as orbiting screw conical dryers. Particle attrition or agglomeration can result in major differences in particle size distribution, bulk density, compressibility and flow characteristics. Therefore, it is not valid to base projections on the results of tray dried samples when different equipment will be used on scale-up. Bench or small pilot-sized test units are available for tumble or paddle dryers, but the dynamic similarity to large-scale equipment is poor. The best way to determine what the product will look like is by performing pilot studies in representative drying equipment.

Predicting the Drying Cycle

Enlarge this picture Figure 3. The product experiences several stages during drying. Elucidating these stages can help identify the major resistances to drying and enable better drying optimization.

An accurate prediction of commercial drying cycle time can be valuable for estimating manufacturing costs. The first step in making such a prediction is to determine the major resistances to drying. This involves performing a simple vacuum-oven drying study by monitoring the solvent content of the cake over the course of drying. For completeness, the test should be performed on several samples of varying cake thickness. For the test to be most meaningful, the sample used must be representative of the final process material.

Several types of drying may occur during a drying cycle. As the conditions in the dryer approach the boiling point of the solvent, the first solvent to be removed is the free, unbound solvent wetting the surface of the solids. This is removed at a fairly constant rate, limited primarily by the heat transfer rate, until the surface is no longer saturated. Once surface solvent is removed, then solvent that is trapped in interstitial spaces and micro-capillaries in the crystals is removed. Evaporation of this solvent is slower because additional energy is required to overcome capillary-attractive forces. Next, solvent that is completely trapped in vacuoles may be removed, but the rate of removal is very slow and limited by diffusion. It is best to assume that this solvent will not be completely removed, nor will solvent that is part of the molecular crystal lattice.

The drying study mentioned above can give a good indication of where the major resistances to drying occur. The solvent content vs. time data can be plotted directly, but it is more useful plotted as drying rate vs. time, or as drying rate vs. solvent content (figure 3). The period where the surface moisture is removed is called the constant-rate period, up to the point where there is no longer sufficient solvent to make a continuous layer over the surface (called the critical moisture content). This is followed by the falling-rate period, characterized by an ever-changing drying rate as, first, the solvent from the unsaturated surface is removed, followed, in turn, by the various components of the internal solvent. Often the falling-rate period dominates the drying cycle. It is a case of diminishing returns, which is why a reasonable drying specification is so important.

Many actual drying curves may not appear to fit this model well. Some product wet cakes may come out of the filter already below the critical moisture content, and then the entire drying cycle will consist of falling-rate drying. This is one reason why it is valuable to know the critical moisture content. In other cases, a short period of rapid solvent removal may occur, followed by a settling into the falling-rate period. This is most likely to occur if a product cake is placed in a preheated dryer and then the vacuum applied.

The results of laboratory drying studies can provide a great deal of useful information. For example, if the constant-rate period predominates in the bench test, then it will likely predominate at scale. Agitated dryers may then offer the shortest drying times by increasing the effective surface area of the cake exposed for heating and drying. Tray dryers may not be as advantageous because increased cake depth in larger-scale units means decreased heating surface area per unit mass.

If the falling-rate period dominates, it means that the process is diffusion limited, and this period will likely dominate at scale. In such cases, agitated dryers such as orbiting screw cones or combination filter dryers may offer advantages by increasing the surface area for diffusion by particle attrition.

More rigorous treatment of the theoretical aspects of drying and calculations of energy requirements, etc., can be found in standard chemical engineering references such as Unit Operations (McCabe & Smith) or Perry’s Chemical Engineering Handbook.

Tips for Operating Dryers

As with any operation, it is important to have a startup checklist to ensure that no important detail is overlooked in starting a drying cycle. Also, prepare detailed cleaning procedures and establish cleaning criteria as necessary. Record all operations, including cleaning and maintenance, in the equipment log book.

Ensure chemical compatibility of seals and gaskets with the solvents in use, as well as the compatibility of the tray and dryer materials with the product. Make sure that no corrosive vapors will attack dryer surfaces. Polyethylene tray liners can be used if desired, depending on the drying temperature. Make sure that the solvent trap is empty and operational. Cryogenically cooled traps can be a convenient substitute for dry-ice/alcohol traps but make sure that the heat-removal capacity will match the peak solvent-removal rate. To ensure sufficient capacity, determine roughly the amount of solvent to be trapped or plan on monitoring the solvent trap level.

Lines and connections must offer no restrictions that could be clogged with ice or other crystals (often a finite amount of water is removed from cakes wet with even hydrophobic solvents). Include a dust trap upstream of the vacuum supply and a solvent knockout trap at the vacuum exhaust. Include an inert gas supply line to release vacuum, especially for air-sensitive or hydroscopic compounds. It is also common to leave on a slight nitrogen bleed during drying to increase the convective removal of evaporated solvent. Remember that if the gas flow rate is too high, however, uncondensed vapors can pass through the solvent trap. The nitrogen bleed also works well to prevent condensation buildup on the dryer window glass.

If covered drying trays are used, make sure to leave an opening to allow unrestricted removal of the solvent. Place the product in the dryer, apply vacuum, and then apply heat. This can help prevent a bolus of solvent from overwhelming the solvent trap condenser at startup.

The temperature stability of the product must be considered in setting the operating conditions to ensure that it will not be heated above the allowable limit. Generally speaking, the higher the vacuum the better in order to keep evaporation temperature low and maximize the temperature difference with dryer surfaces. Of course, keep in mind that a certain minimum temperature may be necessary to reach the specification. Some compounds are prone to polymorphic shifts well below their melting point, and this needs to be understood as well. Finally, by monitoring the actual product temperature during the drying cycle and not just the oven temperature, it is usually much easier to determine exactly when drying is complete.