Deep Freeze
Deep Freeze
August/September 2003
By Narlin B. Beaty, PhD
If Mr. Freeze had current lyophilization processes at his disposal, he might have triumphed over Batman. Fortunately, pharma scientists are well-acquainted with the art of freeze drying drug preparations to prolong shelf life of active ingredients. Much attention recently has been directed at molecular protective properties of excipients and selection of improved formulations. The three common physical lyo processes are freezing, sublimation (primary drying) and desorption (secondary drying).
Freezing. Before sublimation, aqueous products are filled into vials and placed on a lyophilizer shelf for freezing. As shelf temperature is decreased from ambient to about ?50?C, sterile, filtered product dissolved in Water for Injection (WFI)?filled into WFI-cleaned vials in a low-particle environment?will cool significantly below zero before nucleation causes crystallization. Since crystallization is stochastic, time or temperature cannot accurately be predicted for any two vials. Yet, nucleation temperature determines the size of ice crystals and subsequently, the speed of sublimation. Larger ice crystals, obtained through slow crystalline growth, sublime more readily than small crystals. However, slow freezing is not typically achieved by merely lowering the temperature slowly. Low particulate water super cools to well below its freezing point and activity is apparently suspended until some unaccountable event nucleates the solution, causing rapid crystallization throughout the vial.
Thus, most pharmaceutical preparations contain vials with varying, but mostly small, crystal sizes. Annealing?ice crystals can be encouraged to reorganize and favor larger crystal sizes?can speed up sublimation. Of course, solute will become more concentrated if the degree of crystallinity is increased, which may be undesirable for the product. One alternative to annealing?s propagation of larger ice crystals is to decrease temperature very rapidly and uniformly produce ice with many small crystals. However, commercial dryers are largely limited to shelf-temperature decrement of about 1?C per minute. Enclosed in glass insulation on the shelf, product does not cool fast enough to achieve crystal uniformity.
Sublimation. Frozen product consists of ice crystals?crystals of a bulking agent or excipient such as mannitol?and a glass interstitial region that must be maintained critically in the glassy state. If the interstitial region softens, macroscopic changes in the lyophile, or shrinkage, occurs. This is common in lyophilized pharmaceuticals, but mostly inconsequential. Should the glass liquefy, then crystalline phase (ice) dissolution will occur and lead to ?collapse.? For many products, collapse is a reject state for the vial, due to cosmetic or product degradation. Both shrinkage and collapse must be controlled during sublimation because when the extent of adsorption surface area changes, desorption is affected. If the shrinkage/collapse is non-uniform from vial to vial, uniform moisture content after desorption cannot be assured. The glassy state is maintained by performing sublimation when the ice temperature is lower than the collapse temperature. Collapse temperature for a formulation can be measured by differential scanning calorimetry (DSC) or freeze drying microscopy. Often, the collapse parameter is unknown for existing approved products. Such work should be performed before moving a product to a different dryer.
During primary drying, the ice (product) temperature must be controlled. Too often, excessive attention is placed on the precisely controlled shelf temperature. However, shelf heating should be recognized as nothing more than an energy source for sublimation. Most importantly, during sublimation, the ice temperature and the shelf temperature are very different. Within the ice cylinder, a temperature gradient exists from the vial bottom to top; temperature at the sublimation interface (top) is very close to a theoretical temperature predicted for saturated water vapor over ice at a given pressure. The extent of the temperature distribution is variable and depends on the depth of the ice, chamber pressure and the temperature of the underlying shelf. At most, the difference from top to bottom should be a few degrees. Recognizing that a difference exists, and knowing that collapse must be avoided, the product thermocouples must be positioned into the bottom center of each vial.
Chamber pressure determines the temperature of the sublimation interface. For example, suppose that a sucrose formulation has been chosen with a collapse temperature of ?32?C. Assume a safety margin of four degrees below the collapse temperature and a three-degree difference between the bottom and the top of the ice. The vapor pressure of ice at ?38?C is 121mTorr. Such data would justify choosing 121mTorr for the chamber pressure, or even slightly lower if the dryer varies by 2 to 5mTorr during the control cycle. Pressure should be as high as possible. Much of the heat absorbed by the product is transported by convection through the rarified atmosphere. As the pressure is lowered, convection is inhibited and sublimation is slowed.
Desorption. During secondary drying, heat is applied to desorb water from a surface. Generally, more heat is faster and better, reducing moisture content in less time. A moisture-dependent glass transition of the solid lyophile, when exceeded, causes the product to soften and fall apart or melt. Because the transition temperature is moisture dependent, the transition temperature is a moving target. As the moisture decreases (from <15% at the end of primary), the transition temperature rises. Final secondary drying temperature may have to be approached slowly. The exact relationship between glass-transition temperature and moisture content can be determined by temperature-modulated differential-scanning calorimetry, but is not measured commonly.
The absence of ice during secondary drying means pressure does not control temperature. Consequently, pressure might be raised to between 200 and 400mTorr to assist in convective heat transfer. The rationale for choosing pressure during secondary drying should be further researched. ?PFQ
Comments: 0