* * * INSIGHT * * *
Role of Other Gases in Lyophilization Process
by
Hiren Shah#
The generation, movement and condensation of water vapour is an actively discussed topic in the field of lyophilization. However the presence and role played by the other non-condensable gases present in the process has not been discussed clearly. This issue of INSIGHT seeks to clarify and highlight the effects of such gases on the process of lyophilization.
The non-condensable gases are atmospheric gases like nitrogen, oxygen and others that form a part of initial atmosphere of chamber and condenser. It is difficult to evacuate the chamber completely and traces of these gases will remain. Such gases may also come from leaks in the construction of the lyophilizer. The leaks are classified as virtual leaks and real leaks. The real leaks penetrate the construction like material, door, piping and welding where atmospheric gases enter the chamber. Virtual leaks are “apparent real” leaks but the gases are originating from the system and are not as a result of any penetration in the dryer. For example, gases desorb continuously from the walls of the chamber.
These gases are also generated during the process from the frozen product itself. This is due to the fact that general solutions have a definite amount of gases present in the dissolved form inside the solution. When this solution is frozen, the gases are also trapped inside the matrix. It is a well reported fact that atmospheric gases are dissolved inside formulation or liquids at room temperature. For example, as a part of the oxygen cycle in the atmosphere the oxygen keeps on slowly dissolving in sea water which is essential for the survival of marine life.
As another example, one often observes gas bubbles inside ice cubes. These bubbles are formed during the freezing process and when the drying gas-surface interface reaches them they would be released. There are papers that have appeared in Nature that describe how they are able to determine the earth’s atmosphere tens of thousands of years ago from the boring of ice in Greenland. The gas content is analyzed at various distances in the bore sample which is indicative of the atmosphere at that time. During the primary drying, such gases are also released along with water vapour.
Apart from this, gases may also come from the chemical reactions of the product during manufacturing. For example, the products that are converted into sodium salt prior to lyophilization are accompanied by formation of carbon dioxide. If this gas is not removed in formulation then this comes out in the drying when water crystals start to sublime. Sometimes the formulation before freezing is sparged with nitrogen (or helium) to reduce the oxygen content in the water. Additionally as an acceptable industrial practice, sterile nitrogen is leaked inside the chamber during the primary and secondary drying processes. This is a separate discussion which will not be discussed here.
So we see that not only non-condensable gases are present at the start of the process but are also generated during the lyophilization process. It is necessary first to understand the role played by these gases in the primary drying. In order to commence and continue the sublimation of ice from the matrix the following conditions are necessary:
1. The pressure inside the chamber must be lower than the equilibrium vapour pressure of the ice at that temperature.
2. Heat must be continuously supplied so as to provide the heat of vaporization of the subliming ice while maintaining the form of the cake by keeping it well below its eutectic/collapse temperature.
Now the important point is that the lower the pressure in the chamber, the better is the rate of heat transfer and subsequently of sublimation given the same product temperature. Given lower pressure, you have lower surface temperature of the (drying) front and hence a better (higher) T for heat transfer. (only referring to the rate of sublimation.) In order to increase the drying rate at a lower pressure one must also increase the shelf-surface temperature while maintaining the product temperature at a given range and the dryer must have sufficient throughput so that the pressure in the chamber pressure is not increased.
When the freezing is over, the chamber atmosphere is consisting of mostly non-condensable gases. When these gases are removed to a value below equilibrium vapour pressure of the ice at this temperature, the sublimation of ice will commence. These non-condensable gases are removed by the vacuum pump by maintaining a low pressure at the suction of pump. The gases will flow from a high-pressure area to a low pressure area. Due to low pressures at the suction of pump the non-condensable gases move continuously from chamber to pump and are discharged to the atmosphere by the vacuum pump. This also means that it is necessary to use the vacuum pump continuously for this purpose. The condenser does not assist in the removal of non-condensable gases.
Case I Presence of Only Water Vapour
The factors controlling transfer of water vapour is different from the transfer of gases from chamber to the condenser. The partial pressure of water vapour in the chamber is high as vapour is being constantly generated. The partial pressure of water vapour in the condenser is very low as the condenser is at a low temperature. The transfer of vapour is due to the different partial pressures of water vapour. The rate of transfer of vapour is dependent on the difference of partial pressures of the vapour and the physical parameters like the construction of the dryer. The rate is not affected by the vacuum pump as the no water vapour is reaching vacuum pump. Theoretically it is not even necessary to have the vacuum pump for the process if the condenser temperature is low enough.
Case II Presence of Water Vapor and Other Gases
Now let us consider the situation where some gases are being generated along with the water vapour. If the vacuum pump is not able to remove all these gases the total pressure in the chamber will go up. This can happen in the case when the amount of gases trapped is high or they start to come out in large amounts. The competing gas release means that the water vapour generation will decrease. This is because the partial pressure of water vapour decreases due to the presence of gas. The freeze-drying front temperature increases as a result of a lower sublimation rate. In case of higher rate of gas release, there can be an increase in the pressure in the chamber. Such a pressure increase will result in an increased product temperature. When the temperature at the ice interface approaches the product temperature, the system will be near equilibrium.
Case III Industrial Situations
There are several industrial situations when solvent residue remains inside the solution coming from either the product being dissolved or the penultimate process. This solvent will also come out of the matrix along with vapour during heating. When present in small quantities this solvent has to be first removed during the start of the heating. If present in large quantities the solvent is released throughout the cycle at varying rates. This poses an additional consideration of false reading of vacuum measurement apart from decrease in heat transfer.
The behaviour of these gases can be different depending on their condensation temperature. If these gases are condensed at condenser temperature then they will compete with the water vapour by the heat putting additional load on condenser. For example tert-butyl alcohol has a melting temperature of about 25 C and a boiling point of 83 C. Thus it will sublimate along with the water vapor. The Pirani gauge determines the pressure by measuring the temperature of a wire. The wire temperature will decrease in the presence of solvent traces. The presence of solvent vapors results in display of a false high pressure in both chamber and condenser.
To sum up, what are the affects of non-condensable gases on the operation of lyophilization? First, we must see that these gases affect the heat transfer across the matrix. Therefore we must control the heat input rate to adjust to the presence of these gases. In many situations they force us to decrease the heat input rate as they decrease the heat transfer across the matrix. Secondly, during formulation or chemical reactions, it is desirable to remove as much as of dissolved gases before freezing so as to have as little effect on the cycle as possible. These gases do not affect the condenser operations as long as the temperature of the condenser remains low. There are also other effects of these gases on the measurement of the vacuum which shows false high reading. This has been discussed in detail in INSIGHT issue Vol . 1 No. 6. Although such errors are not indicated by the normal non-condensable gases which is chiefly nitrogen and some oxygen.
Vol. 2. No. 6 June 1999
# Person to contact with regarding comments or questions about this INSIGHT
Hiren Shah
Eutectic Technologies
Bombay, India
Tel : 91 22 513 9690 Fax: 91 22 515 0744
Email : Hiren.Shah@Mailcity.com
