by Timothy Dutill and Wendy Sunderland

When considering the pharmaceutical application of freeze drying, of utmost concern is the ability to manufacture a consistent product with the required critical quality attributes. In order to do so, process uniformity must be ensured. A common and expected approach is to conduct process qualification (PQ), for the freeze drying process. Prior to a product specific PQ, the lyophilizer performance should be verified at the factory, once installed at the site as part of commissioning, and again for operation qualification (OQ). The studies should ensure the critical process parameters are functioning as intended and within a specified allowable range. In addition to tests qualifying the independent and controllable variables of the lyophilizer, including shelf inlet temperature, chamber pressure, and time, it is beneficial to also assess certain dependent variables directly impacting product uniformity. One such important measurement is shelf temperature uniformity, sometimes referred to as shelf temperature mapping.

Shelf temperature is typically controlled and monitored by a resistance temperature detector (RTD) immersed in a thermal well at the heat transfer fluid inlet to the shelves and is often known as the shelf inlet temperature (Shelf Inlet). As heat transfer from the lyophilizer to the product occurs at the shelf surface to vial interface, measuring the actual temperature of the shelf surface is critical for assessing lyophilizer function. The shelf temperature distribution of these measurements should then be compared relative to the control point and the Shelf Inlet to ensure an acceptable range and uniformity.

One of the greatest challenges during such studies is the ability to measure the actual shelf surface temperature without the influence of environmental conditions. Temperature measurement devices can be divided into two distinct groups: direct measurements and indirect measurements. The direct measurement is a device which secures the temperature sensor, usually a thermocouple, directly to the shelf surface. The drawbacks of the direct measurement method are that the temperature sensors are difficult to place, difficult to keep in place at low and high temperature ranges, and typically require extensive cleaning due to residual adhesives or paste. The main advantage of this method is that, when executed correctly, it provides a measurement of the actual surface temperature in a specific location. The indirect measurement is a device with the temperature sensor embedded in a heat conductive material resting on the shelf surface.

Figure 1: Temperature measurement devices – Copper disc (back left), ValProbe (back right), thermocouple setup (front)

There are a number of commercially available devices employing the indirect method of temperature measurement. The main drawback of this type of device is that temperature measurement is more likely to be influenced by the chamber environment; however, these influences can be addressed by the test method. The main advantages of the indirect measurement are the ability to consistently place the devices and the elimination of any potential residues.

Studies were conducted at Lyophilization Technology, Inc (LTI) evaluating the ease of use and effectiveness of a number of different temperature measuring devices on measuring shelf temperature uniformity. Studies were run in a 2ft2 single shelf lyophilizer and a 24ft2 four shelf lyophilizer. Although other methods were investigated, only the most viable methods will be discussed here including direct measurement with thermocouples (thermocouple setup) and indirect measurement with Aluminum discs, Copper discs, and the Kaye1 ValProbe® (ValProbe) (P/N X2534). Figure 1 is a photograph of the thermocouple setup, Copper disc and ValProbe.

Thermocouple setups are often considered the “gold standard” in shelf temperature mapping since the measurement is made directly at the shelf to sensor interface. The method used here to attach the thermocouples to the shelf included two 4” sections of duct tape formed as a cross with the thermocouple and heat conductive paste placed in the center under a 1” x 1” section of refrigeration tape. A variation on this method uses a plastic tube with a spring to hold the thermocouple in place. As described above, the thermocouples are very difficult to place, in particular in the back of the lyophilizer. The pilot scale lyophilizer used for the studies at LTI had a 3 foot depth; using this method for a commercial lyophilizer would only add additional challenge.

Discs or ‘pucks’ of varying conductive material are a simple way to employ the indirect method and are often used in the industry in varying arrangements. A thermocouple or RTD is inserted into, or adhered onto, a flat disc of some heat conductive material such as aluminum or copper which then rests on the shelf surface. The disc can then be easily slid to its proper place on the shelf. This method requires no cleaning of residual paste or tape on the shelves. Studies here used 2 inch diameter cylinders of Aluminum and Copper cut into 0.25” thick discs. Sections of insulation tape secured a thermocouple tip coated in heat conductive paste to the top of the disc.

ValProbes are a wireless device with a stainless steel base and a RTD secured into it. The device is supplemented with a thermal pad that adheres to the base. The thermal pad diminishes the time required to reach steady state and ensures a more accurate measurement especially under vacuum conditions. Similar to the discs, the devices are easily placed in the proper location on the shelves. There is slight inconvenience in comparison to other methods with having to program the ValProbes, ensuring there is adequate battery life, and for some versions, not having real time data and uploading the data into analysis software.

Figure 2 shows the difference between the average shelf inlet temperature and the average measured surface temperature over a 5 hour interval at -55°C, 5°C, and 50°C. The Aluminum disc showed minimal difference compared to the thermocouple method at -55°C and 5°C. At 50°C, the Aluminum disc was closer to the shelf inlet temperature than the thermocouple which is most likely due to the thermocouple not remaining firmly adhered to the shelf.

 

Figure 2: Difference between the average shelf inlet temperature and the average measured surface temperature for Aluminum disc versus thermocouple (TC) taped to shelf.

Figure 3 shows the difference between the average shelf inlet temperature over a 1 hour interval at -50°C, 0°C, and 50°C and the average measured surface temperature of three thermocouples taped to a Copper disc or three thermocouples taped directly to the shelf. The Copper disc performed as well or better than the thermocouple taped directly to the shelf when the system was at or near atmospheric pressure. However, when the pressure in the chamber was decreased to 200 microns the temperature measurements with the Copper discs were skewed relative to the thermocouples.

 

Figure 3: Difference between the average shelf inlet temperature and the average measured surface temperature for Copper disc versus thermocouple (TC) taped to shelf.

Figure 4 shows the difference between the average shelf inlet temperature over a 1 hour interval at -50°C, 0°C and 50°C and the average measured surface temperature of 20 ValProbes or thermocouples. The ValProbes, in general, performed as well as or better than the thermocouples both at atmospheric pressure and under vacuum. The measured shelf temperatures for the thermocouples and ValProbes were closer to the shelf inlet temperature when the system was under vacuum.

 

Page 4 of 6 March 2014

Figure 3: Difference between the average shelf inlet temperature and the average measured surface temperature for Copper disc versus thermocouple (TC) taped to shelf.

Figure 4 shows the difference between the average shelf inlet temperature over a 1 hour interval at -50°C, 0°C and 50°C and the average measured surface temperature of 20 ValProbes or thermocouples. The ValProbes, in general, performed as well as or better than the thermocouples both at atmospheric pressure and under vacuum. The measured shelf temperatures for the thermocouples and ValProbes were closer to the shelf inlet temperature when the system was under vacuum.

Figure 4: Difference between the average shelf inlet temperature and the average measured surface temperature for ValProbes versus thermocouple (TC) taped to shelf.

 

Figure 5: Average thermocouple and ValProbe temperatures at the end of a ramp to -50°C.

Figure 5 shows the average of 20 thermocouples and 20 ValProbes at the end of a ramp to -50°C and the beginning of a hold at -50°C. The thermocouples level off after about 0.75 to 1 hour into the hold at -50°C while the ValProbes take between 1.5 and 2 hours to reach steady state. This difference does not significantly change the final readings as the average values are between 0.4°C and 0.6°C different in the last half hour of the hold.

Shelf temperature mapping using thermocouples taped directly to the shelf surface has historically been the most effective means of measuring shelf surface temperatures because it provides a direct measurement of the temperature. The direct measurement has typically been more accurate and precise when a good measurement was obtained. However, the main disadvantage of using thermocouples has been the difficulty in obtaining consistently good measurements. This can be seen graphically in Figure 2 where the thermocouple performed much worse at 50°C when compared to the same thermocouple at -55°C and 5°C. The adhesive used did not hold as well at 50°C and the thermocouple rose off the shelf surface slightly causing a shift in the temperature reading.

The alternate methods investigated represent alternative techniques. The Aluminum and Copper discs are easily manufactured, cheap, and require little if any maintenance. The techniques of attaching the thermocouple to the surface or embedding it inside the disc do not alter the resulting temperature measurements. These methods make placing the thermocouples much easier than taping the thermocouple to the shelf; however the wires continue to add difficulty to the method.

The main disadvantage of using discs is that when the system is placed under vacuum, they present an impedance to heat transfer due to the imperfect contact with the shelf surface. In order to get around this limitation, most shelf temperature mapping experiments are run with the system at or near atmospheric pressure.

Tim has been a part of Lyophilization Technology since January 2011. He is involved in managing client projects including product and process development, clinical manufacturing, validation, and troubleshooting. Focused activities include equipment qualification and thermal analysis, including freeze drying microscopy, electrical resistance measurements, differential scanning calorimetry, and thermogravimetric analysis. His career began in 2006, in the vaccine field, as a Product Development Scientist at the Infectious Disease Research Institute. Tim’s educational background consists of a B.S. in Chemical Engineering from Pennsylvania State University. He has been published in the Journal of Colloids and Surfaces B: Biointerfaces, Pharmaceutical Development and Technology and Biopharm International. His professional memberships include the American Association of Pharmaceutical Scientists (AAPS), the International Society of Lyophilization - Freeze-Drying (ISL-FD) and the Parenteral Drug Association (PDA).