INSIGHT - Comments
Vol. 1 No. 4 Water Vapor Throughput for A Freeze-Dryer
There are a couple different ways to control pressure at a set point in a freeze dryer, that I am aware of. These include:
1. Turning the vacuum pump on and off
2. Throttling a valve between the vacuum pump and the rest of the system
3. Allowing an outside gas (inert or atmospheric) to be introduced to the chamber through a valve (needle valve or proportional type solenoid valve).
Are there any advantages or disadvantages of these different approaches to protocol/cycle development?
Are there any advantages or disadvantages to the equipment itself?
Ms. Cindy Reiter April 1999
Division Manager
BioPharm Group at FTS/Kinetics
Response by T. A. Jennings
Vacuum Pump: You can certainly control the chamber pressure, within limits, by closing the valve to the vacuum system, assuming that the drying chamber is not isolated from the condenser and that the pump is equipped with a shut-off valve. The increase in pressure in the chamber will be a result of increase in partial pressures of the non-condensable gases. Such gases can stem from the product or from leaks into and from the freeze-dryer. Such a control method would perhaps be best suited for freeze-dryers that have a very large effective volume and a low leak rate. For such systems, the rate of pressure rise (ROR) for non-condensable gases may be in terms of mTorr per hour rather than mTorr per min. The only real advantage I can see with such a system is that gas is not being introduced into the chamber and, thereby, reducing the risk of chemical or biological contamination of the drying chamber.
The major disadvantage of this method would be, especially in the case of dryers having a relatively low effective volume, is that one may see a large ROR as a result of leaks of the non-condensable gases. This increase in pressure could have two effects. The first is to slow down the drying rate, especially during the primary drying. The second effect would be to increase in the product temperature (see INSIGHT Vol. 1 No. 2). If the product temperature increase where to exceed the collapse or eutectic temperature, then repeating such pressure rise cycles could cause collapse of the cake and the product would be vacuum dried rather than lyophilized.
Perhaps the most major drawback to this method is the frequent turning on and off of the pump motor. Such an operation could cause a serious shorenting of the life of the pump motor. When the motor is turned on, it momentarily draws a high current because of the low reactance of the motor windings. These repetitious periods of high current over a relatively long period of time could cause the motor to overheat and eventually fail. Even the repeated stopping and starting of the pump itself may result in undue wear and tear on some parts and thus shorten the life of the pump. The real danger, however, would be that the pump would fail during the primary drying portion of the process. In the absence of a backup or auxiliary vacuum pump, the buildup of excessive chamber pressure would, before a new vacuum pump could be installed, result in a loss of the entire batch. I would certainly be reluctant to use such a method for controlling the chamber pressure.
Throttle Valve: The control of the chamber pressure using a throttle valve, placed in the foreline that is between the vacuum pump and the condenser, will have essentially the same effect on the drying process as that for the on and off of the vacuum pump. The principle difference is that the vacuum pump will continue to operate and thereby eliminate the possibility of overheating of the motor the vacuum pump. Frequent cycling of a pneumatically driven value would not represent a problem with overheating but overheating resulting from frequent cycling could be a factor for an electrical solenoid. It would be recommended that, regardless of the means used to operate the valve, the valve should be one that is designed for continuous service. One difficulty that one would have to be concerned with is a valve malfunction. For example, if the malfunction were to result in a valve not seating properly, then the continuous flow of gas passing the valve seat may cause the chamber pressure to fall below the lower limit for the drying process. If this be the case, then during the primary drying process, assuming a constant shelf-surface temperature, the product temperature could be lowered to such an extent that there would be a significant decrease in the drying rate. A valve that would fail in a closed position would result in an uncontrolled increase in chamber pressure and once again if this were to occur during primary drying then there could be a major change in the cosmetic appearance, potency or stability of the final product.
The throttle valve may also be used to control the chamber pressure by bleeding a gas (e.g. dry sterile nitrogen) into the chamber. In this application, the gas must first increase the condenser to a pressure that is equal or greater than that of the drying chamber. Thus every time that there is a requirement to increase the chamber pressure, the rate of drying would approach zero. But perhaps the most serious impact of this form of pressure control is the potential, especially at low pressures (i.e., < 100 mTorr), to increase the rate of backstreaming to the drying chamber. At such low pressures, hydrocarbon vapors from oil sealed mechanical vacuum pumps are known to cause a film of oil vapors to form on the walls of the foreline where the throttle valve is located. While such hydrocarbon vapors will tend to migrate towards the drying chamber, the introduction of a flow of gas in to the line will only serve to enhance such a migration.
Gas Bleed: The gas bleed directly into the drying chamber is the system that I prefer to use. In this type of system, the pressure is controlled by periodically introducing a gas (generally dry sterile nitrogen) into the drying chamber. The advantage here is that there is a constant flow of gas from the chamber into the condenser. In addition, provided that the collapse temperature is high enough, chamber pressures of the order of 200 mTorr can be used. At such pressures, the flow of non-condensable gases into the mechanical pump is sufficient to prevent backstreaming of hydrocarbon vapors from an oil sealed mechanical pump. A second important advantage would be that during secondary drying one could, if the cake would permit it, use a gas purge to complete the secondary drying process and thereby offer a means of obtaining consistent cake properties from batch to batch. Especially with respect to the residual moisture in the cake.
From an equipment standpoint, this system can be as simple as a solenoid valve in series with a multi-turn throttle valve or a more complex servo-driven valve system that will allow a constant stream of gas into the chamber and hold the chamber pressure to within a few mTorr. In the first case, the throttle valve should be positioned between the chamber and the solenoid valve in order to prevent periodic surges of pressure that could exceed the upper pressure limit of the process. The location of inlet tubing to the chamber will be dependent on the design and construction of the chamber. One should be concerned about the nature of the gas flow from each shelf. The position of the inlet tubing should not cause a non-uniform flow of gas from the shelves. An uneven flow of gas in the drying chamber can generate pressure gradients and if there are pressure gradients then there can also be variation in product temperatures during the drying process. The attempt to increase the efficient usage of the chamber by maximizing the number and area of the shelves can unfortunately lead to an inefficient lyophilization process.
