Overcoming common freeze drying challenges
In this paper, GEA specialists discuss how freeze drying professionals can overcome common freeze drying challenges by altering tray shape, changing product geometry, and creating more uniform heat distribution. Using test data from freeze-dried raspberries, black currents, mangos, carrots, and chicken, GEA specialists provide further recommendations for optimizing capacity, maintaining quality, and creating a productive and profitable freeze drying process.
Author: Troels Bjerregaard Pedersen, Process Specialist
1. The benefits and challenges of freeze drying.
Freeze drying is an effective way to preserve the shape, color, taste, and nutrients of many types of food and beverages while vastly increasing their shelf life.
The process involves drying a frozen product in a vacuum below the triple point, allowing for a sublimation process that direct converts ice into vapor without an intermediate liquid state. With freeze drying, only vapors are transported in the product, which contrasts to conventional drying methods taking place above the triple point.
Freeze drying can be divided into internal processes inside the product and external processes outside of the product.
-
Internal processes include freezing the product, sublimation, heat transfer within the product and drying hygroscopic-bound water (secondary drying).
-
External processes cover heat transfer to the product and handling released vapors.
Freezing can happen in the freeze dryer or as a separate process outside the freeze dryer, which is the standard in GEA RAY® and CONRAD® concepts. Sublimation is driven by external heat transfer to the product, typically from conduction or infra-red heating.
Heat transfer by conduction requires contact between heating plates and product trays, which yields an uneven heat transfer when there is a small distance between the heating plate and product (or tray).
Radiation is the standard in the GEA RAY® and CONRAD® concepts and is independent of distance. This enables a uniform heat transfer to the product. As the freeze drying process progresses, a dry zone will emerge where ice has been sublimated. The heat will be transferred by conduction through this dry zone and simultaneously cooled by escaping vapors initially at the sublimation temperature.
The vapors created by the sublimation phase change affect the pressure if they are not handled by a vapor trap, which must be sufficiently cold to condense the vapors. A low pressure-drop is expected for vapor passage from the sublimation front to the condenser. This drop is related to product layer thickness and will result in increased local pressure within the product.
Fig.1. Schematic overview of an industrial GEA RAY freeze-dryer with continuous de-icing and a close-up of a schematic representation of a flat tray with granular product midway during freeze drying – Local pressure is indicated for the chamber. This is the measured pressure used for control. Furthermore, local pressure estimates are indicated for the top of the dry product layer, at the sublimation front, in the frozen product and at the coil’s surface in the vapor trap, which is operated at -38°C.
While freeze drying is exceptionally effective for food preservation, there are several limiting factors related to capacity.
Limiting factors on freeze drying capacity include:
- Heat transfer to the product
- Heat transfer within the product
- Transport of sublimated vapors from the product
- Product layering
- Product characteristics and temperature sensitivity
- Service time between batches
- Combined effects
2. Limiting factors on capacity
The foremost requirement for an adequate freeze drying process is that all “wet” parts of the product remain frozen during the entire process. This requires that the pressure in any part of the product is kept below the triple point, which is 6 mbar for pure water. Furthermore, no part of the product may exceed temperatures higher than those defined as acceptable to maintain desired quality.
Limiting factors are directly linked to temperatures or pressure that is too high. By overcoming these obstacles, heat plate temperature can be increased and maintained for an extended period. This leads to a higher sublimation rate and a more efficient high-capacity freeze drying process.
2.1. Vapor removal
As mentioned, vacuum pressure must be maintained below triple point. If pressure rises above that point, the product can collapse or boil.
Within the product vapor removal can be restricted by skin, which functions as a barrier. This could turn blueberries or other skinned fruits or vegetables into pressure cookers. See section 2.4 for more about product characteristics.
The product layer itself can also cause a rise in local pressure, due to pressure drop over the distance in which the vapor must be transported. This is comparable to the pressure drop experienced when liquid is pumped through a pipe. In flat trays this can be alleviated by reducing the tray load. In finned trays, which as a standard are completely filled with product, this act as design limitation of the tray height.
Externally, vapor must be removed to maintain the vacuum. This is done by condensing vapors on the vapor trap coil. Passage of the vapor from product surface to vapor trap can be limited by the connection size between the vacuum chamber and the vapor trap. This is why vapor traps are located inside the drying chamber in GEA RAY® concepts, as shown in Figure 1.
The basis of sublimation acceleration is that the freeze dryer can handle generated vapors. When scaling up such processes, volumes increase massively. One kilogram of steam at 0.8mbar has ~1500m³ volume.
It is important to understand that pressure build-up in relation to freeze drying is a scale of millibars. Small changes in pressure can have a large effect lead and to melting, which destroys the purpose of freeze drying. The product can still be dried, but it does not maintain the original shape or structure.
2.2. Heat transfer
A primary limiting factor that can affect freeze drying is heat transfer.
The driver of phase change is energy (heat), which is required to make water change from ice, to liquid, to steam, and back again. During sublimation, the phase change is directly from ice to vapor.
The heat transfer can be divided into external and internal heat transfer, which refers to the heat transfer from the heat source to the product surface/tray, and transfer from the product surface/tray to the sublimation front inside the product.
Radiation heating is more efficient than contact heating in terms of external heat transfer in high-capacity systems. Radiation allows for uniform heat transfer to the product surface and does not produce hotspots. This is because radiation heating is independent of the distance between object and heat source. In contrast, conduction needs perfect contact between heating source and tray to provide uniform heating. This contact is rarely possible due to small dents and other tray aberrations.
During the GEA freeze drying process, the sublimation process is driven by energy transferred to the product by direct radiation or through the trays. As drying progresses, the dry transport zone is formed, and internal heat transfer takes place. This occurs while the energy is transferred by conduction from the surface, through the dry zone, to the sublimation zone.
Fig.2. Schematic overview of a single particle and multiple particles in a tray midway during freeze drying – Vapors released from sublimation occurring in the bottom of the tray will exit upwards (not shown). They cannot pass through the tray. Mass and heat transfer are indicated with different arrows. Note that exiting vapors have a cooling effect, with an initial temperature that corresponds to the sublimation temperature.
When the dry transport zone becomes too thick or the ice surface becomes too small, sublimation will become insufficient and will not produce enough vapors to cool the outer part of the transport zone.
If radiation (i.e., heating plate temperature) is not reduced, the product’s surface temperature will rise above acceptable limits, resulting in a deterioration of quality. As heating plate temperature is reduced, so is the sublimation and the release of cooling vapors. In the end, the heating plate temperature must be reduced to the maximum acceptable product temperature as the sublimation rate approaches zero. This can be observed in the comparison of a freeze drying trend and the schematic overview of the development in a single product piece during drying (See Figure 3).
Fig.3. Typical process trend for freeze drying in a GEA RAY concept – showing the heating plate temperature, surface and inner-product temperature, product weight, and the pressure measured by a capacitive Pirani sensors. The product temperatures are related to locations indicated on a schematic representation of a single piece of product during freeze drying.
2.3. Product layer thickness
The tray load – or thickness of the product layer – also limits capacity. Thinner layers dry faster than thicker layers, but they add up to more frequent batch changes and downtime between batches. Layers that are too thick can lead to longer drying time due to low sublimation rates when heating plate temperature is reduced.
Another issue with a thin layer is that even small variations in layer thickness can correspond to major deviations. A change from one piece to two pieces is 100 percent, whereas a change from 10 to 11 pieces is only 10%.
Fig.4. Freeze-dried apple slices – Discolored areas due to uneven tray filling.
Discover how Automatic Calf Feeder revolutionises calf feeding routine at the Zweck Family.
Fig.5. Schematic representation of three spots in the same tray – with a difference in layer thickness and the resulting difference by the end of the drying profile.
2.4. Product characteristics – Process impact
Freeze drying is suitable for a broad range of foods. Still, many product characteristics can limit the efficiency of the drying process. These characteristics can be divided into three categories: homogeneity, mass transfer and single-piece geometry.
Homogeneity
Homogeneity within and between product pieces is essential for achieving a uniform drying of the product in terms of required heat input and drying time. Non-homogenous product structure can lead to long drying times, which affects the process efficiency.
Some products have differences in chemical compositions and concentrations in individual parts of the product, resulting in decreased homogeneity. Product pieces can also vary from each other due to differences in e.g. origin, variety or ripeness. This can result in differences in dry matter and sugar, starch and acid composition. Both types of variation may result in longer drying times, since the recipe must cater to various drying speeds within the batch.
Fig.6. Products have diverse characteristics – which may be problematic for capacity optimization.
- Strawberries can have differences in dry matter content in outer and inner parts and can be hollow in the centre. It is difficult to create an even tray layer of large strawberries.
- Broccoli pieces are not very homogeneous. They have an airy flower and a solid stem, and their large awkwardly-shaped bouquets make it difficult to create an even-layered tray.
- Whole blueberries have a tough skin, which function as a barrier for escaping vapors. This effectively turns the fruit into a pressure cooker.
Mass transfer
Peels and skin, which act as a barrier for vapor transport, can negatively affect mass transfer. The result is a build-up of pressure inside the product, which causes collapse and boiling, unless operating pressure and sublimation rates are reduced.
Product geometry
Product geometry can also be a problem. Awkward shapes like apple slices aren’t conducive to producing uniform product layers.
3.Overcoming limitations
Some limitations are inevitable in the freeze drying process, such as the large vapor volumes or energy requirements. Still, it is possible to overcome many of the most significant challenges. Changing a product’s shape and size or performing scarification are some ways to modify products and make them optimal for freeze drying. If product geometry allows, trays can also be divided into smaller compartments, with fins used to create a uniform energy distribution in the product.
Reshaping products
Changing a product’s shape, preferably before freezing creates a geometry that facilitates a uniform and even tray distribution.
Size Reduction
Reducing product piece sizes via cutting prior to freezing or granulation (Producing particles with a size of 2-6mm ) of frozen product can create a more homogenous product layer, break the skin/barrier, and enable the use of finned trays. Optimal size is generally 10-15mm uniform pieces for freeze drying in flat trays.
Scarification
Scarification via small cuts and punctures can alleviate skin issues and limit vapor transfer out of the product.
Fig.7. Size reduction of a broccoli bouquet solves geometry and homogeneity issues – Scarification of blueberries may solve issues relating to the skin acting as a barrier, which limits vapor removal.
3.1.Finned trays
Finned trays should be considered if product size reduction is possible. Finned trays are known from the instant coffee industry and allow for optimal utilization of GEA freeze drying technology.
Fins distribute heat into the product layer, giving a uniform heat distribution. This allows for higher energy transfer over longer time without compromising product quality. Finned trays work for the following product sizes:
- 14 mm fin distance: < Ø 6 mm
- 22 mm fin distance: < Ø 10 mm
The finned trays’ height is restricted for some products with low freezing points, like bacteria concentrates. This is due to the local pressure rise that occurs as vapors are transported through the product layer. (See section 2.1)
Fig.8. Schematic of finned trays with a difference in fin distance – Finned trays greatly increase the heat distribution uniformity within the product. The fins’ distance further affects uniformity of heat distribution, product size demands, and the ease of tray filling.
4.The Raspberry test - design
In May 2020, specialists at GEA designed a test to compare parameters relevant for industrial freeze drying of raspberries. The test aimed to have a maximum product temperature of 60°C for all tests and included groups of the following:
- Whole raspberries dried at 120 and 125 °C heat plate temperature
- Granulated raspberries in finned trays (22mm fin distance and 38mm height) at 125 and 140 °C
- Granulated raspberries in finned trays (14mm fin distance and 50mm height) at 140 °C
Granulation was performed in a GEA GR 305 granulator, and freeze drying was performed in a RAY® 2.
Results are presented as the mean sublimation rate and the capacity estimate. The mean sublimation rate is measured from when 1 mBar pressure was established to when vacuum break was initiated.
The capacity estimate is for an industrial RAY® 125. The capacity estimate is calculated directly from the mean sublimation rate, with an adjustment for service time (Change over time = 0.5 hour) and adjustment for scaling up to industrial production (tech transfer = 1 hour).
4.1.The raspberry test - results
The raspberry test resulted in a conclusion that the most effective way to increase capacity in freeze drying is with finned trays.
When it came to increasing heating plate temperature, some effect was seen. Temperature increase was limited, though, since tray-filling uniformity has a greater effect on product quality at higher heat plate temperatures. Granulation also had a slight effect on increasing the raspberries’ sublimation rate. This is ascribed to the increased homogeneity of the product geometry.
The use of finned trays significantly improved results. GEA specialists saw a substantial increase in the sublimation rate when a finned tray with a 22mm fin distance was used with granulated raspberries. The higher sublimation rate is the result of higher heating plate temperature for longer time, which is facilitated by the better heat distribution of the finned trays.
An even greater sublimation increase was seen when GEA specialists used a 14mm finned tray with a higher tray load. The sublimation rate more than doubled from IQF berries in flat trays at 120°C versus granulated berries in 14 mm finned trays at 140°C.
The narrower fins have an even better heat distribution facilitating even longer maintenance of high heat plate temperatures.
Fig.9. Test results – presented as mean sublimation rate (top) and capacity in kg/day input for a RAY 125 (bottom)
5.Extended product test - design
The extended product test was designed to show the difference between granulated products in finned trays (14 mm fin distance and 45 mm height) versus granulated and “whole” products in flat trays, all at 140 °C heat plate temperature. All tests aimed to have a maximum product temperature of 60°C.
Test Specifications
- Individually Quick Frozen (IQF) whole products in flat trays at 140°C heating plate temperature
- Granulated (~5 mm particles) products in flat trays at 140°C heating plate temperature
- Granulated (~5mm particles) products in finned trays (14mm fin distance, 50mm tray height) at 140°C heating plate temperature
Tested Products
- Whole black currants
- 1 x 1 x 1 cm chicken cubes
- 2.5 x 2.5 x 2.5 cm mango cubes
- 1 x 1 x 1 cm carrot cubes
Granulation was performed with frozen product in a GEA GR 305 granulator. Freeze drying was performed in a RAY® 2.
Results are presented as the mean sublimation rate and the capacity estimate. The mean sublimation rate is measured from when 1 mBar pressure was established until vacuum break was initiated. The capacity estimate is for an industrial RAY® 125. The capacity estimate is calculated directly from the mean sublimation rate, with an adjustment for service time (0.5 hour) and tech transfer (1 hour).
6.Extended product test – results
The extended product test shows that finned trays have a substantial benefit over flat trays when it comes to sublimation rates and overall product quality.
To start, the 140°C heating plate temperature for products in flat trays (IQF and granulated) was not feasible. There lacked uniform heat distribution within the product layer to accommodate this temperature, which resulted in discoloration of the top layer.
GEA specialists concluded that a feasible maximum heating plate temperature should be determined experimentally on individual products, with most products having a maximum of 125 - 130°C.
Granulation positively affected the sublimation rate for products with inferior shapes and characteristics to the granulated product. This includes skin and peels that could limit mass transfer, non-uniform composition, and odd shapes. Products with high uniformity and standard shapes, like cubed products that are initially 10-15mm, had a lower mean sublimation rate when granulated and dried in a flat tray.
Overall, granulation produces smaller pieces and high particle size distribution, which can alleviate problems with geometry and homogeneity but may worsen heat transfer properties if geometry and homogeneity was optimal.
When 14mm finned trays were used, the sublimation rate was approximately double that of flat trays. The expected daily capacities in a RAY® 125 closely follow sublimation rates, with adjustments for service time between batches and tech transfer from lab to industrial scale. These adjustments are related to the expected reduction in diligence when creating uniform product layers across several trays.
In this estimate, the same adjustment for tech transfer has been used for all tests. The adjustment is expected to be smaller for finned trays, which are less prone to error.
Fig.10. Results from the extended product test – presented as the average sublimation rate
Fig.11. Results from the extended product tests – presented as the estimated capacity in a RAY 125
7.Other Comments
A series of other observations regarding optimal utilization of freeze drying equipment should be taken into consideration prior to upgrading and production.
Filling and handling of finned tray
An assisted tray-filling mechanism is necessary to correctly fill the finned trays. Manual filling will either be slow or uneven. Finned trays are heavy, so the process should take place in a small in area or via some form of assisted tray handling. Manual handling of finned trays for a GEA RAY® system is strenuous and can be a burden on personnel.
Optimization of recipe
Some producers prefer having a simple recipe for a variety of products. This is possible, but it can negatively influence capacity. To best utilize equipment, recipes should be optimized for each specific product. The more optimized a recipe, the higher the demands to the uniformity in product and tray-filling quality.
To create an optimal recipe, it’s important to investigate:
- Working hours of the plant and availability of product should be taken into account.
- Maximum acceptable product temperature
- Optimal tray load (if flat trays). Sublimation rates and service time between batches should balance each other.
- Maximum feasible heating plate temperature (140°C is absolute max for aluminum heating plates). The higher the temperature, the higher the energy transfer. This leads to a more efficient process, small variations in tray filling etc. will have a higher impact on product quality.
Conclusion
Both the raspberry and mixed product tests led GEA specialists to recommend finned trays for products where size reduction is possible. Optimal spacing of fins is 14mm, and highest possible heat plate temperature is recommended. Fines should be minimized as much as possible, as these may result in product loss due to fluidization from escaping vapors. The bulk density of the granulated product defines the tray load.
Flat trays are recommended when size reduction is undesirable or impossible. Producers should also consider the importance of their product’s shape. Altering it can significantly improve product layer uniformity and sublimation rates. Carefully consider tray load regarding uniformity of product layer.
With these alterations, freeze drying professionals can overcome common freeze drying challenges and create an optimal end-product.
8.References
- Baker, Christopher G. J. Industrial drying of foods, Chapman & Hall, 1997
- Ciurzynska, A.; Lenart, A., Freeze drying – application in food processing and biotechnology – a review. Pol. J. Food Nutr. Sci., 2011
- Duan, Xu, et al., Technical aspects in freeze-drying of foods, Drying Technology, 2016
- Gutcho, Marcia H., Freeze drying processes for the food industry, Noyes Data Corporation, 1977
- Mitchell, J. R. Developments in food preservation-3, Edited by Stuart Thorne, Elsevier Applied Science Publishers, London, 1985
- Oetjen, Georg-Wilhelm; Haseley, Peter, Freeze-drying. John Wiley & Sons, 2004
- Rey, Louis, Lyophilasation, Hermann, 1964
- Serna-Cock, Liliana; Vargas-Munoz, Diana Patricia; Aponte, Alfredo Ayala, Structural, physical, functional and nutraceutical changes of freeze-dried fruit, African Journal of Biotechnology, 2015
- Shofian, Norshahida Mohamad, et al., Effect of freeze-drying on the antioxidant compounds and antioxidant activity of selected tropical fruits, International Journal of molecular sciences, 2011
- Van Atta, CM, Vacuum science and engineering, McGraw-Hill Book Co, 1966
- Waghmare, Roji Balaji, et al., Recent Developments in Freeze Drying of Foods, 2021
