During the development of lyophilization processes, we are frequently asked how to optimize cycle times and enhance product stability. The common approach is often to focus immediately on adjusting parameters during the primary drying phase—specifically shelf temperature, chamber pressure, and vacuum levels. However, based on over a decade of practical experience at MATTO MACHINERY—guided by the principle of “Quality by Design”—and corroborated by extensive research, we have found that the critical factors determining process bottlenecks and product quality uniformity are often established during the pre-freezing stage. Consequently, our laboratory devotes significantly more effort to researching “pre-freezing parameters” than to any other step in the process. The results speak for themselves: we have achieved increases of over 40°C in collapse temperature and reductions of more than 60% in overall process duration—achievements driven primarily by the precise adjustment of parameters during the pre-freezing phase. Today, let us take a deep dive into this critical step—one that is often oversimplified during development, yet in reality governs the success of every subsequent stage of the lyophilization process.
The Underestimated Stage in the Freeze-Drying Process: Why Freezing Is Far More Than Just “Freezing It Solid”?
Traditionally, freeze-drying has been viewed simply as a two-step process: “first freeze, then dry.” Consequently, many development efforts have focused their energy on optimizing the drying phase, treating pre-freezing merely as a prerequisite step intended solely to “achieve solidification.” This perspective, however, can give rise to numerous uncontrollable risks in subsequent stages. Fundamentally, freeze-drying constitutes a sequential process comprising two equally critical units: pre-freezing and drying. While early research and applications tended to prioritize the drying phase, the profound impact of the pre-freezing step was often overlooked. During pre-freezing, the solution undergoes a complex phase transition from a liquid to a solid state—a process centered on the phenomenon of *supercooling*. The degree of supercooling experienced by the solution directly and fundamentally determines the morphology, size, and numerical distribution of the resulting ice crystals. Furthermore, the porous network left behind by these ice crystals serves as the primary conduit through which water vapor escapes during the subsequent sublimation phase. Consequently, non-uniform supercooling leads to a non-uniform ice crystal structure, which in turn directly results in variations in drying rates between individual vials and discrepancies in the final product’s appearance. This presents a fundamental challenge to the robust development, optimization, and scale-up of the freeze-drying process.
Based on our project experience, the root causes of many issues encountered during pilot-scale scale-up and long-term commercial production—such as fluctuations in drying times, non-uniform residual moisture levels, and an uncontrolled incidence of defective products—can frequently be traced back to insufficient investigation and understanding of the pre-freezing step. Neglecting the scientific design and precise control of the pre-freezing phase is akin to erecting a complex process structure upon an unstable foundation; regardless of how meticulously the subsequent drying parameters are fine-tuned, the fundamental stability of the entire process remains perpetually in doubt.
The Secret of Ice Crystals: How Does Supercooling Define Product Structure?
The key to understanding the pre-freezing stage lies in grasping the concept of supercooling. This is a quantifiable and measurable core physical parameter—essentially the “blueprint” for the resulting ice crystal structure. When a solution is cooled under standard atmospheric pr
essure, it does not immediately freeze upon reaching its theoretical freezing point. Instead, the solution remains in a liquid state even below this freezing point; this specific state is known as supercooling.
Supercooling degree refers to the temperature difference between the actual temperature at which the solution begins to freeze (the nucleation temperature) and its theoretical freezing point. The magnitude of this temperature difference serves as the starting point for every aspect of the ice crystal structure formed during the freeze-drying process. The degree of supercooling is subject to the random influence of numerous factors: the presence of foreign particles, the internal surface area of the container, the cooling rate, sample volume, formulation composition, and the contact interface between the solution and the container, among others. It is precisely this inherent randomness in these factors that introduces uncertainty into the nucleation process, leading to variations in the pre-frozen structure of products within a single batch—a significant challenge in maintaining process uniformity.Simply put, consider an ordinary strawberry as an example: regarding the typical supercooling data for a strawberry, the theoretical freezing point (the ice point) is generally between -1°C and -2°C, as the fruit’s juice contains sugars and organic acids.
However, the actual freezing temperature (the nucleation temperature) may differ significantly; under slow cooling conditions, the liquid within the strawberry might drop to -4.8°C—or even lower—before ice crystals actually begin to form.
During the freeze-drying (or lyophilization) process, the degree of supercooling directly determines the morphology of the ice crystals, which, in turn, dictates the quality of the final dried product and the efficiency of the drying process.
Relationship Between Supercooling and Freeze-Drying
Thus, the degree of supercooling acts as the “master conductor” of the freeze-drying process, exerting its influence across the following three dimensions:
1. Determining Ice Crystal Size (The Core Impact)
High Supercooling Degree (Rapid Cooling):When a strawberry is cooled rapidly—dropping significantly below its freezing point before freezing occurs—it generates a multitude of tiny, uniformly distributed ice crystals.
Result: This causes minimal damage to the cellular structure and ensures excellent rehydration properties; however, the resulting pores left behind in the dried product are extremely fine.
Low Supercooling Degree (Slow Cooling): When freezing begins at a temperature very close to the theoretical freezing point, the ice crystals have ample time to “grow,” resulting in the formation of large, irregular crystals. Result: Ice crystals puncture the cell walls (resulting in a softer texture), but the “tunnels” (micropores) they leave behind are relatively large.
2. Impact on Drying Rate (Efficiency Considerations)
This represents the most paradoxical aspect of freeze-drying:
Large ice crystals (formed under low supercooling) are easier to dry: During sublimation, the channels left by large crystals act like “highways,” offering minimal resistance to escaping water vapor and resulting in a rapid drying rate.
Small ice crystals (formed under high supercooling) are more difficult to dry: The minute channels resemble “narrow winding paths,” making it difficult for water vapor to escape; this significantly prolongs the sublimation time and increases electricity costs.
3. Determining Appearance and Rehydratability
Well-controlled supercooling: When freeze-dried, strawberries retain their original structural framework (without collapsing) and vibrant color. Conversely, if insufficient supercooling leads to incomplete pre-freezing, “foaming” or “concentration” may occur during the vacuuming stage, causing the strawberries to shrivel into a dense mass.
Rehydratability: If you want the strawberries to soften rapidly when added to milk, an intermediate level of supercooling is typically required to strike a balance between porosity and structural integrity.
| Degree of Supercooling | Ice Crystal Characteristics | Drying Rate (Sublimation) | Final Product Quality |
| High (Rapid Cooling) | Small & Numerous (Fine distribution) | Slow (High resistance in tiny pores) | Excellent structure, fine texture, good rehydration |
| Low (Slow Cooling) | Large & Sparse (Coarse structure) | Fast (Large pores act as “highways”) | Risk of collapse, coarser texture, potential cell damage |
From Structure to Performance: How Does Pre-freezing Govern the Entire Drying Process?
The ice crystal structure formed during pre-freezing is not merely a static “relic,” but rather a dynamic “scaffold” that dominates the entire drying process. Its influence is primarily manifested through two core engineering parameters: drying layer resistance and product temperature.
1. Decisive Impact on Drying Layer Resistance
The most direct effect of pre-freezing is the shaping of ice crystals. Once these ice crystals sublimate, the space they originally occupied forms the pores within the dried cake: large ice crystals create large pores, offering low resistance to water vapor flow; conversely, small ice crystals create small pores, resulting in high resistance. Drying layer resistance is inversely proportional to ice crystal size. This principle has been repeatedly validated through the use of freeze-drying microscopy and extensive batch data analysis.
2. Complex Regulation of Product Temperature and Sublimation Kinetics
Product temperature represents the critical equilibrium point between process safety and efficiency, and it is profoundly influenced by the structural characteristics established during pre-freezing. Analysis using heat and mass transfer equilibrium models reveals the following:
For products characterized by large ice crystals and low resistance, the sublimation rate is rapid, resulting in significant heat absorption. If the heat input (shelf temperature) is insufficient, the product temperature will drop; this leads to a reduction in vapor pressure at the ice front, thereby diminishing the driving force for sublimation. In such instances, it may be necessary to increase the shelf temperature to maintain the target product temperature and fully capitalize on the advantage of low resistance.
For products characterized by small ice crystals and high resistance, the sublimation process itself becomes the rate-limiting step. If the heat input is excessive, the heat that cannot be dissipated in a timely manner will cause the product temperature to rise. Should the product temperature exceed its eutectic temperature (for crystalline systems) or collapse temperature (for amorphous systems), it will trigger the destruction of the product’s microstructure (manifesting as melt-back or collapse), resulting in unacceptable product appearance or even compromised stability. Consequently, to safeguard the product, it is often necessary to lower the shelf temperature—albeit at the cost of an extended drying duration.
In essence, the pre-freezing step defines the operable safety window and the upper limit of efficiency for the subsequent drying process. Without a fundamental understanding of the underlying structural characteristics established during pre-freezing, the setting of any drying parameters lacks a scientific basis and amounts to little more than a blind endeavor.
MATTON MACHINERY’s Three-Step Pre-freezing Process: Cooling, Nucleation, and Solidification
A controlled pre-freezing process can be divided into three logically interconnected stages.
Stage 1: Cooling
The solution temperature decreases from its initial value to a point below the freezing point, continuing until the first ice nucleus appears. During this stage, the solution exists in a supercooled liquid state; the ultimate depth of supercooling achieved serves as the “driving force” reserved for the subsequent phase transition. The cooling rate directly influences the distribution of supercooling: slower cooling rates tend to yield a more uniform degree of supercooling throughout the sample.
Stage 2: Nucleation and Phase Transition
Beginning with the emergence of the first ice nucleus, the system undergoes a sequence of primary nucleation followed by rapid secondary nucleation—an exothermic process. The supercooled solution absorbs the latent heat released during crystallization, causing its temperature to rise rapidly back toward the freezing point. The greater the degree of supercooling, the more latent heat of crystallization the solution can absorb and “dissipate”; consequently, a higher proportion of the water freezes instantaneously, resulting in a larger number of ice crystals and a faster rate of crystal growth. This is precisely why deep supercooling leads to the formation of a fine, dense ice crystal structure.
Stage 3: Solidification and Concentration
Ice crystals continue to grow while the concentration of the unfrozen solution steadily increases—a process known as freeze concentration. For crystalline formulations, the solute eventually reaches saturation and crystallizes, forming a eutectic mixture; the system achieves complete solidification once its temperature falls below the eutectic temperature. For amorphous formulations, the solute does not crystallize; instead, the solution’s viscosity increases sharply until it transitions into a glassy solid state. In this case, the system completes its solidification process once its temperature falls below the glass transition temperature (Tg’) of the maximally freeze-concentrated solution.
Many food factories and freeze-dryer manufacturers use the eutectic point temperature as the freezing temperature for their freeze-drying processes; consequently, they get it wrong right from the very first step.
Active Control: Cutting-Edge Technology for Achieving Optimal Pre-freezing
Since the stochastic nature of supercooling lies at the root of the problem, actively controlling ice nucleation becomes the key to achieving process consistency. Currently, various techniques aim to induce uniform nucleation at a preset, higher temperature through external intervention, thereby yielding an ideal ice crystal structure characterized by large, uniform crystals.
1. Annealing
According to a report by Searles, annealing amorphous sucrose or hydroxyethyl starch can result in a 3.5-fold increase in the primary drying rate. The challenge lies in the need to precisely optimize the annealing temperature and duration, a process that also extends the overall cycle time. Determining the exact annealing temperature and duration entails significant costs in terms of both human resources and equipment.
2. Ice Fog
TechnologyWhen the product cools to the target nucleation temperature, an “ice fog”—generated from cryogenic nitrogen and water vapor—is injected into the freeze-drying chamber to serve as seeding nuclei. Studies by Rambhatla et al. indicate that utilizing this technique to raise the nucleation temperature from -11°C to -1°C can reduce primary drying time by nearly 30%. Reports on industrial-scale applications (such as the VERSECQ® technology) demonstrate that this method effectively improves vial-to-vial uniformity and reduces primary drying time by approximately 19%.
3. Vacuum-Induced Surface Freezing
By rapidly reducing the chamber pressure, evaporative cooling is induced at the solution’s surface; this triggers freezing in the surface layer, thereby initiating nucleation throughout the entire bulk. Kramer et al. reported that this method facilitates the formation of large, “chimney-like” ice crystals, shortening the primary drying time by approximately 20%. However, this technique operates within a narrow process window and imposes stringent requirements on the precise control of shelf temperature and vacuum levels; specific parameters must be empirically determined for each distinct product. This is the technology currently employed by the TELSTAR China.
4. High-Pressure Shift Freezing
According to a report by Konstantinidis et al., this method can reduce the primary drying time for a 5% mannitol solution by 40%. Research by Bursac et al. demonstrated that for a 5% sucrose solution, primary drying time was reduced by 27%, accompanied by a significant improvement in the uniformity of the dried cake.
5. Ultrasound-Assisted Freezing
At the target temperature, ultrasonic waves are applied to the sample; the resulting cavitation effect is utilized to trigger nucleation. Studies by Nakagawa et al. have demonstrated that raising the nucleation temperature from -8°C to -2°C transforms the morphology of ice crystals from fine and disordered to large and dendritic, thereby reducing the primary drying time by as much as 60%.
The choice of technique depends on the characteristics of the product formulation (specifically, its sensitivity to mechanical stress and temperature), process objectives (whether uniformity or efficiency takes precedence), as well as existing equipment capabilities and modification costs. No single technique is universally applicable; rather, a scientifically sound selection is predicated upon a profound understanding of both the underlying mechanisms and the specific requirements of the product. Nevertheless, the core principle remains consistent: to transform the pre-freezing stage from an uncontrollable, stochastic event into a scientifically defined, monitorable, and reproducible process step.
MATTON’s Philosophy: Integrating Science Throughout Lyophilization Process Development
At MATTON, we are committed to a scientific approach to lyophilization, ensuring that every data point is substantiated by evidence. We reject experimentation based on vague empirical intuition, advocating instead for a development pathway driven by data and mechanistic understanding.
When approaching a new lyophilized product, our development process begins with a systematic understanding. As we design the lyophilization cycle profile for our clients, our steps include:
Step 1: Characterization of Critical Temperatures and Formulation Understanding
We begin by employing a comprehensive suite of tools—such as Differential Scanning Calorimetry (DSC) and Freeze-Drying Microscopy (FDM)—to precisely determine the formulation’s eutectic temperature (T_eu), glass transition temperature (T_g’), and collapse temperature (T_C). This process involves more than merely obtaining a few temperature values; it aims to answer fundamental questions: Does this formulation form crystals or an amorphous glass during the freeze-concentration phase? What is its maximum permissible drying temperature? Which excipients are prone to crystallization, and at what stage does this occur?
Step 2: Investigation of Pre-freezing Behavior and Variable Correlation
We systematically investigate how different cooling protocols influence the actual distribution of ice nucleation temperatures. For instance, we do not simply record the “average” nucleation temperature; rather, we focus on its distribution range, as the breadth of this distribution directly reflects the inherent variability of the process. We quantify the degree of supercooling and correlate it with the subsequent morphology of the ice crystals (assessed through structural analysis of the lyophilized cake). Key operational variables we monitor include: cooling rate, shelf temperature at the time of loading, and the contact interface between the sample and its container—factors that directly impact the degree of supercooling.
Step 3: Mechanistic Process Design and Optimization
Through the design of comparative experiments, we analyze how different pre-freezing strategies—such as conventional shelf cooling versus the application of controlled nucleation techniques—impact the critical parameter of drying layer resistance (R_p). We evaluate the product using methods such as the Pressure Rise Test, while simultaneously recording the primary drying duration and the product temperature profile. Our objective, through this collaborative product development process, is to empower our clients’ teams with a robust, efficient, and reliable lyophilization process!
Building a Trust Bridge Between Machines and Processes
The development of a lyophilization process is a long-term, systematic undertaking.
Manufacturers require genuine professional expertise, a profound grasp of physical principles, and the capability to utilize specialized tools to precisely control process performance. The pre-freezing stage constitutes the initial—and indeed, the decisive—link in this systematic chain. Scientifically controlling this step serves as the foundational starting point for achieving a robust, efficient, and consistent process. MATTON MACHINERY is dedicated to transforming lyophilization process development from an experience-dependent “craft” into a science- and data-driven “engineering discipline.” If you are currently facing challenges regarding process uniformity, efficiency bottlenecks, or scale-up complexities—and if you resonate with MATTON’s core values—we look forward to engaging with you.
