Open-loop or closed-loop control of a freeze-dryer on the basis of sensing a characteristic temperature rise with respect to at least one sample vial of a plurality of vials subjected to a freeze-drying process

Abstract

The invention relates to a method for controlling a freeze-dryer (LYO1) which is designed to subject a total set of vessels (V1 to V9), referred to as vials and having contents to be dried, to a freeze-drying process and which comprises at least one temperature measuring system designed to sense an instantaneous temperature in at least one sample vial from the total set of vials (V1 to V9) and at least one process flow controller (PFC1) designed to control the temperature of at least one placement surface of the vials in the freeze-dryer at least in a primary drying phase according to a specified temperature profile over time, the method being characterized by a) sensing, in the sample vial, a characteristic temperature rise reflecting progress of the primary drying, b) carrying out the further freeze-drying process in a defined manner according to the sensed characteristic temperature rise or a plurality of characteristic temperature rises sensed for a plurality of separate sample vials. According to a first aspect of the invention, an individual placement surface temperature is estimated for the (relevant) sample vial from the sensed characteristic temperature rise(s). According to a second aspect, an end point of the primary drying phase is identified or predicted from the characteristic temperature rises of a plurality of sample vials. The invention also relates to a corresponding controller for a freeze-dryer and a corresponding freeze-dryer.

Description

[0001] Control or regulation of a freeze dryer based on a detection of a characteristic temperature rise in relation to at least one sample vial of a plurality of samples subjected to a freeze-drying process

[0002] Vials

[0003] Description

[0004] The invention relates to a control system for a freeze dryer which is designed to subject a total number of vessels referred to as vials (such as small bottles) with contents to be dried to a freeze-drying process, comprising at least one temperature measuring system which is designed to record a current temperature in at least one sample vial from the total number of vials, and at least one process sequence control which is set up to control the temperature of at least one storage area of ​​the vials in the freeze dryer at least in a primary drying phase according to a predetermined temperature profile over time, as well as a corresponding method for controlling a freeze dryer, and a freeze dryer having such a control system.

[0005] According to a first aspect of the invention, inter alia, a control or - preferably - regulation of the temperature profile is to be achieved on the basis of a determination, in particular calculation, of the individual shelf temperature, in particular by a non-linear adaptation.

[0006] The invention is therefore based on the object of controlling or regulating the process of freeze-drying in a freeze dryer in such a way that the drying time is as short as possible, while at the same time ensuring the most reliable drying possible within the framework of GMP - Good Manufacturing Practices - for pharmaceutical products with a defined probability. In particular, the integrity of the remaining amorphous and particularly readily soluble structure after removal of the solvent, typically water, is to be ensured, while still accelerating the process. According to a second aspect of the invention, among other things, a control or - preferably - regulation of the temperature profile is to be achieved, with a particularly statistics-based detection of the end point of the primary drying phase.

[0007] The invention is therefore based on the object of controlling or regulating the process of freeze-drying in a freeze dryer in such a way that the drying time is as short as possible, but on the other hand a largely safe drying within the framework of GMP - Good Manufacturing Practices - for pharmaceutical products is ensured with a defined probability.

[0008] The basic process of freeze-drying is explained, for example, in the description of US Pat. No. 8,793,895 B2. Illustrations and explanations of a practical design of a freeze-dryer with automatic filling used in practice can be found, for example, in WO 2005 / 121671 A.

[0009] The freeze-drying process, also known as lyophilization, for pharmaceutical products essentially proceeds as follows: The freeze dryer is filled with small glass containers called vials, in large quantities per shelf. Several shelves are stacked one on top of the other. The vials are already prepared with a stopper that is only partially inserted, allowing gas to escape through side slits. The vials are filled with the pharmaceutical product as a substrate, usually in an aqueous solution.

[0010] After closing the process chamber, the temperature of the shelves is drastically reduced by the coolant lines laid in a serpentine pattern, typically to temperatures well below zero (decline). This shock-freezes the substrate. At the same time, the air in the process chamber is pumped out, creating a vacuum.

[0011] Based on the phase transition diagrams known from thermodynamics, sublimation now occurs, i.e., a direct transition from the shock-frozen solid to the gaseous state. Special design features of the freeze dryer ensure that the escaping vapor, usually water vapor, settles in a second chamber at a condenser. The freeze-drying process is now in the primary drying phase.

[0012] The major advantage of this process is the gentle removal of water from the substrate without the need for higher temperatures, which could damage the substrate. At the same time, with proper process management, an amorphous glass structure, the so-called sublimate cake, is formed in the substrate. This subsequently enables excellent solubility when the pharmaceutical product is converted back to a liquid state shortly before use, for example, with water added via a syringe. The removal of water during production ensures that the pharmaceutical product can be easily stored and transported without, for example, harmful decomposition processes occurring.

[0013] Once this primary drying phase is complete, the temperature is slowly raised back to room or storage temperature, allowing a further drying step, known as secondary drying, to begin. This removes most of the remaining water. At the end of secondary drying, the prepared stopper is mechanically pushed together into the vial, hermetically sealing it. Air is then admitted back into the process chamber. The freeze dryer is then mechanically unloaded, and the remaining process steps, such as visual inspection, stopper sealing, and packaging, are performed.

[0014] The primary drying phase takes up most of the drying process, which can last several days. Therefore, there is a strong economic interest in shortening the primary drying time to better utilize the expensive equipment.

[0015] To this end, various control approaches have been tested in the past, none of which have proven particularly effective. Therefore, the approach still chosen today is to scale the process from small to large plants using simulation and to incorporate appropriate time-related safety margins. For profile development using simulation, reference is made to EP 1 903291 A1. However, this approach has the disadvantage that, on the one hand, the margin may still be too small, thus not ensuring the required quality according to GMP, or, on the other hand, the margin may be too large, so that the freeze dryer is not used economically and the delivery time, for example, of individual therapeutics for the patient, is unnecessarily extended.

[0016] The main problem with control is that, on the one hand, vials exhibit manufacturing variations in their base thickness, which influence heat transfer. On the other hand, different placement positions on different surfaces can exhibit different properties due to the underlying coolant lines and heat radiation to the walls of the process chamber. In addition, there are slight fluctuations in filling. This multitude of influencing factors is difficult to control or predict.

[0017] Nevertheless, simple control has been attempted in the past, compare US 6,163,979 A for controlling the process pressure or the shelf temperature based on the substrate temperature, or US 4,780,964 A only for the air pressure inside the chamber.

[0018] Another approach involves using heat flow sensors in a scaled-down sample process to determine the drying state based on the heat flow between the plate and the vial (see US 10,605,527 B2). The disadvantage here is that radiant heat is not detected, yet complex scaling is required.

[0019] Another approach is to measure the gas in the process chamber (see WO 2018 / 194925 A). So-called Pirani vacuum gauges are often used for this purpose. However, this only allows the average drying rate to be measured, which is not in line with GMP, since vials with individually extended drying times are not measured if only a few vials are involved. There are also approaches to control entirely via pressure (see US 4,780,964 A). The approach of a hard specification instead of control is described in US 8,839,528 B2, but it also does not solve the actual problem of fluctuations between vials.

[0020] Fundamentally, there is a desire to be able to draw conclusions about the drying process from temperature measurements of individual vials. Individual measurements using thermocouples have shown that a characteristic temperature jump occurs at the end of the primary drying process. This is because the substrate is no longer cooled by evaporative cooling, meaning that the largely completed sublimation process essentially no longer removes any sublimation heat from the substrate to be dried in the vial. However, measuring with thermocouples is very complex and only possible in a prototype process, as the cabling itself violates hygiene requirements, disrupts automatic loading, and dissipates heat, thus massively impacting the process being measured.

[0021] Wireless temperature measurement is therefore desirable, but the energy source must not contain any harmful chemical substances and must not heat the substrate. Reference can be made to WO 2016 / 123062 A1 as relevant prior art. It discloses an arrangement for monitoring an aseptic manufacturing process, which comprises product condition sensors capable of taking measurements of the product condition, for example, temperature or humidity, at short intervals. The measurements are performed using closely spaced sensors arranged in a linear array on a single probe. This probe allows measurements to be taken at multiple levels within the product. The data from the sensors is transmitted to a data acquisition point via wireless digital short-range communication.The sensors can measure temperature and humidity at a single location. If the sensors are used, for example, in the freeze-drying of pharmaceuticals, the position of a sublimation front can be calculated for each vial, and this data can be used to control the freeze-drying process. The published application fails to provide any concrete technical solutions for how temperature can actually be measured wirelessly with sufficient accuracy.

[0022] In fact, precise wireless temperature measurement can now be performed without a dedicated energy source in the sensor using the invention described in DE 197 02 768 C1, which ostensibly relates to a completely different technical field. The energy from the radio field is stored as resonance energy in so-called temperature measuring quartz crystals of wireless measuring transducers (sensors). Due to their crystal cut, these crystals exhibit a well-defined relationship between temperature and resonance frequency and, unlike thermocouples, enable direct absolute temperature measurement without a reference cold junction. Since there is no resistive element that could produce power losses, the impact on the substrate is minimal. The sensors can be built very small, as only a rectifier diode, a few matching elements, and the antenna are required.In this regard, reference can also be made to patent documents EP 0 901 417 B1 and US 6,378,360 B1 of the same patent family, which disclose several alternative embodiments.

[0023] The excitation is carried out by means of an amplitude-modulated microwave signal of very low power, the interrogation is carried out by omitting the modulation while the carrier is still emitted, whereupon the resonating measuring quartz in turn causes a back modulation via the diode, which can be measured.

[0024] An expanded version of such wireless sensors combined with a special integrated circuit to provide a large number of measurement channels can be found in DE 102017 103 974 B3 and the corresponding publications EP 3 586 273 B1 and US 10,958,490 B2. According to this solution, the microwave signal is modulated in such a way that it transmits an identification code, to which a sensor responds to check the temperature and then wirelessly transmits a temperature measurement. The system is commercially available under the name "Tempris" and is considered the industry standard in many areas of freeze-dryer measurement. However, an average number of measurement channels can also be achieved using older approaches by selecting different resonance frequencies, via which multiple sensors can be addressed individually.

[0025] However, such systems can only measure the temperature prevailing in the respective vial, but currently do not allow for dynamic control of the shelf temperature during the freeze-drying process, as there is no way to determine the heat flow from the shelves into the vial. Thus, there is a risk that the amorphous sublimate cake will collapse due to residual moisture if the adjustment is made too early. Even sufficiently precise control of the freeze-drying process based on such temperature measurements is currently not possible, as fluctuations due to the nature, filling, and positioning of the vials continue to interfere.

[0026] Against this background, the invention is based on the object of enabling a more precise control or regulation of the freeze-drying process which better takes into account the actual course of the freeze-drying process.

[0027] To achieve this object, the invention provides a control system for a freeze dryer, which is designed to subject a total number of vessels, referred to as vials, with contents to be dried to a freeze-drying process, and comprises at least one temperature measuring system, which is designed to detect a current temperature in at least one sample vial from the total number of vials, and at least one process flow controller, which is designed to control the temperature of at least one storage area of ​​the vials in the freeze dryer at least in a primary drying phase according to a predetermined temperature profile over time.According to the invention, the process flow control is further configured to a) detect a characteristic temperature increase in the sample vial during the primary drying phase, in particular by means of the temperature measuring system, which reflects a progress of the primary drying, and b) carry out the further freeze-drying process in a defined manner depending on the detected characteristic temperature increase or a plurality of characteristic temperature increases detected for several separate sample vials.

[0028] Furthermore, the invention provides a method for controlling a freeze dryer which is designed to subject a total number of vessels, referred to as vials, with contents to be dried to a freeze-drying process and comprises at least one temperature measuring system which is designed to detect a current temperature in at least one sample vial from the total number of vials, and at least one process flow controller which is designed to control the temperature of at least one storage area of ​​the vials in the freeze dryer at least in a primary drying phase according to a predetermined temperature profile over time.According to the invention, the method is characterized by the step a) detecting a characteristic temperature increase in the sample vial reflecting a progress of the primary drying and the step b) carrying out the further freeze-drying process in a defined manner depending on the detected characteristic temperature increase or a plurality of characteristic temperature increases detected for several separate sample vials.

[0029] The proposed detection of at least one characteristic temperature rise in the sample vial allows conclusions to be drawn about the actual course of primary drying and thus an adaptation of the freeze-drying process to this actual course of primary drying by carrying out the further freeze-drying process in a defined manner depending on the characteristic temperature rise recorded for the sample vial or on the characteristic temperature rises recorded for several sample vials. This allows freeze-dried products, such as pharmaceutical products, to be produced that meet very high quality standards, and it also enables efficient freeze-drying.

[0030] The invention and preferred embodiments of the control system for a freeze dryer and of the method for controlling a freeze dryer, which provide additional advantages, are defined in the independent claims with the dependent claims of the appended set of claims which refer back thereto.

[0031] Further advantageous embodiments of the invention with regard to different aspects emerge from the items A1 to A20, B1 to B20 and C1 to C89 given at the end of the description.

[0032] Patent protection is sought for subject matter A1 to A20 and subject matter B1 to B20, regardless of the subject matter of the invention and further development defined in the appended claims. The applicant reserves the right to formulate corresponding claims.

[0033] Some preferred further development proposals in the context of the control system according to the invention for a freeze dryer will be discussed below in the context of:

[0034] The process flow controller can advantageously be configured to determine, based on the detected characteristic temperature increase or the plurality of characteristic temperature increases detected for several separate sample vials, whether and in what way the freeze-drying process should be influenced, or to propose to an operator of the freeze-dryer, preferably automatically via a user interface, a preferably predetermined or predeterminable influence on the freeze-drying process for approval. The process flow controller can then influence the freeze-drying process accordingly upon the operator's approval.

[0035] The process control system can further be configured to control at least one temperature adjustment device of the freeze dryer depending on the detected characteristic temperature rise or the plurality of characteristic temperature rises detected for several separate sample vials. Regarding the characteristic temperature rise, it is understood that this is a temperature jump or includes a temperature jump that occurs as a result of completed primary drying, when the largely completed sublimation essentially no longer extracts any sublimation heat from the contents of the sample vial to be dried, possibly referred to as the substrate.

[0036] The characteristic temperature rise can be the temperature jump with a temporally subsequent rising temperature profile or can include the temperature jump with a temporally subsequent rising temperature profile.

[0037] According to a particularly preferred implementation, the process control system can be further configured to detect and evaluate the characteristic temperature rise or a temperature jump characterizing this with a substantially negative-exponential temperature profile in the sample vial from the temperature jump toward a temperature limit, in order to estimate the temperature limit from a nonlinearity of this temperature profile. The process control system can then carry out the further freeze-drying process depending on the temperature limit. The detection can be carried out, in particular, by means of the temperature measurement system.The temperature limit value can be interpreted as the shelf temperature at the location of the sample vial, so that the course of the freeze-drying process can be deduced and the further course of the freeze-drying process can be optimized by influencing at least one process parameter.

[0038] In a further development, it is proposed that the process flow control is further configured to model, on the basis of a plurality of assumed different provisional temperature limit values, the essentially negative-exponential temperature profile in the sample vial from the recorded temperature jump as a linearized logarithmic temperature profile, by the process flow control adapting temperature-time measured value pairs representing the essentially negative-exponential temperature profile in the sample vial from the recorded temperature jump to a model straight line equation describing the respective linearized logarithmic temperature profile, and that the process flow control is further configured to compare the model straight line equations obtained for the various provisional temperature limit values ​​with each other and / or with the temperature-time measured value pairs,to estimate the most suitable provisional temperature limit as the temperature limit.,

[0039] The process flow control can be set up to iteratively and / or stochastically select different provisional temperature limit values ​​in order to model the essentially negative-exponential temperature curve in the sample vial from the recorded temperature jump as a linearized logarithmic temperature curve.

[0040] It is further proposed that the process flow control is configured to select different provisional temperature limit values ​​within at least partially sequentially performed modeling processes and / or within at least partially parallelly performed modeling processes in order to model the essentially negative-exponential temperature profile in the sample vial as of the detected temperature jump as a linearized logarithmic temperature profile.

[0041] The process flow control can further be configured to subject the model straight line equations obtained for the various provisional temperature limit values ​​to a quality assessment by determining a summary quality measure in each case, which evaluates a linear temperature-time curve corresponding to the respective model straight line equation using the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value on the basis of this quality assessment.

[0042] The process flow control can advantageously be configured to change a current shelf temperature depending on the temperature limit value by means of the temperature setting device of the freeze dryer, which is preferably assigned to at least one shelf of the freeze dryer, wherein this change is preferably based on an interpretation of the temperature limit value as an approximate shelf temperature.

[0043] The process flow control can advantageously be designed to determine a quality measure for the estimated temperature limit value which is included in the further process control.

[0044] The process flow control can be set up to evaluate the estimated temperature limit value by calculating the difference between this and the measured temperature over time, logarithmizing this difference and then performing a linear regression against the time course using the logarithmized values ​​and using this to determine a quality measure for their linearity.

[0045] The process flow control can be specifically configured to determine the estimated temperature limit value used in the further process control by determining a quality measure for iteratively changed estimated temperature limit values ​​or from several assumed estimated temperature limit values, and from this an estimated temperature limit value with an improved quality measure is used as the basis for the further process control.

[0046] It is envisaged that the process flow control is set up to determine the estimated temperature limit by evaluating this quality measure based on at least one assumed estimated temperature limit and then iteratively changing the assumed estimated temperature limit so that the quality measure is improved, preferably using established numerical iteration methods such as a gradient search or an evolutionary algorithm in the case of further parameters to be optimized.

[0047] The process flow control can be configured to determine the estimated temperature limit by evaluating each of a plurality of statically assumed estimated temperature limit values ​​and selecting from this set the one with the best evaluation according to a quality measure in order to use it as a basis for further process control or as a starting value for a subsequent iterative numerical determination of the estimated temperature limit on which further process control is to be based.

[0048] The temperature measurement system can be designed to record a respective instantaneous temperature in several sample vials from the total quantity of vials. The process control system can advantageously be configured to a) record, during the primary drying phase, characteristic temperature increases reflecting progress in the primary drying in the form of temperature jumps in the sample vials, in particular by means of the temperature measurement system, and b) carry out the further freeze-drying process in a defined manner depending on the recorded temperature jumps. This execution of the further freeze-drying process can advantageously depend on whether the end point of the primary drying phase can be identified or predicted from the temperature jumps.

[0049] In a further development, it is proposed that the process flow control is set up to determine, on the basis of the detected temperature jumps, whether the execution of the freeze-drying process is to be influenced or whether a preferably predetermined or predeterminable influence on the execution of the freeze-drying process is to be proposed to an operator of the freeze-dryer for release, in particular automatically by means of a user interface.

[0050] The process flow control can be configured to determine, on the basis of at least one of i) a plurality of previously recorded temperature jumps and ii) a recorded number of previously recorded temperature jumps and preferably on the basis of a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or a predetermined time (tEND), and, depending on this determination, to influence the execution of the freeze-drying process or to propose an influence on the execution of the freeze-drying process to an operator of the freeze-dryer for release.

[0051] Advantageously, the process flow control can be set up to assign a respective measurement time to the temperature jumps recorded for each of the sample vials and, after recording a specified rate or number of temperature jumps at the measurement times recorded in this way, to calculate at least one inferential statistical hypothesis test in order to determine, on the basis of the / a predetermined maximum probability of error, whether an end to the primary drying can be assumed in the sense of a hypothesis at the current time or the predetermined time, and, depending on this determination, to influence the execution of the freeze-drying process or to propose an influence on the execution of the freeze-drying process to an operator of the freeze-dryer for approval.

[0052] In further development, it is proposed that the process flow control is set up to calculate the mean and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution over the measurement times and to feed it into the hypothesis test.

[0053] Advantageously, the process flow control can be set up to assign the sample vials to different groups within the framework of the statistical evaluation depending on the positioning of the sample vials in the freeze dryer on the shelves or between the shelves, if desired using a factor analysis for statistical dimension reduction.

[0054] In a further development, it is proposed that the process flow control be configured to allocate the mean, variance, standard deviation, or at least one other statistical distribution measure adapted to the distribution among the groups. The process flow control is further configured to statistically weight the groups depending on the positioning of the measured sample vials contained within them on the storage area and the number of other, unmeasured vials of the total quantity of vials with a technically comparable positioning. Allocating the vials among the groups may mean or include a separate calculation of the relevant statistical distribution measure for the groups.

[0055] Advantageously, the process flow control can further be configured, in the case of a negative hypothesis being determined according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time, to repeat the determination as time progresses to determine whether an end of the primary drying can be assumed in the sense of a hypothesis at a later, then current time or the predetermined time.

[0056] Furthermore, the process flow control can be further configured, in the event of a positive hypothesis being determined according to which an end of the primary drying can be assumed at such a current time or the predetermined time, to shorten or terminate the primary drying from this time onwards and to proceed to subsequent process steps of the freeze-drying process, or to suggest such shortening or termination of the freeze-drying to the operator for a transition to the subsequent process steps for manual release.

[0057] The process flow control can further be set up to first monitor the attainment of a necessary freezing temperature of the vials based on the sample vials, if desired also by means of a hypothesis test on the hypothesis of the attainment of this freezing temperature, and further set up to further lower the shelf temperature in the event of an assumed non-attainment of the freezing temperature.

[0058] Corresponding further development proposals for the method according to the invention for controlling a freeze dryer arise from the appended method claims and the subject matters C46 to C87.

[0059] The invention further provides a control system for a freeze dryer, which is designed to carry out a control and / or regulation of at least one freeze-drying process according to the method according to the invention.

[0060] The invention also provides a freeze dryer having a control system according to the invention. This control system can advantageously be used to control the temperature of the vial shelves. In this regard, it is contemplated that the temperature profile can be individually influenced by the control system for each shelf or for groups of shelves.

[0061] The invention is explained in more detail below using several embodiments with reference to the associated figures.

[0062] Fig. 1 shows a schematic diagram of a freeze dryer with its control system.

[0063] Fig. 2 shows temperature profiles over time occurring in the freeze-drying process with characteristic temperature increases that include a temperature jump followed by a range of increasing temperature.

[0064] Fig. 3 shows schematically the temperature profiles occurring in the freeze-drying process over time with characteristic temperature increases in the form of temperature jumps together with an associated probability distribution.

[0065] Fig. 4 shows schematically in part 4 a) a wireless multi-channel temperature measuring system of a freeze dryer and in part 4 b) a measuring vial equipped with a wireless temperature sensor.

[0066] Fig. 5 shows a schematic diagram of a freeze-drying system including the process control and associated components.

[0067] Fig. 6 schematically shows an arrangement of measuring vials and test or product vials, which formed the basis for two runs with identical vial arrangements, using the respective measuring vials and test or product vials for validation measurements and validation evaluations. Fig. 7 is a diagram showing the residual moisture values ​​determined for the measuring vials and test or product vials during the validation measurements and validation evaluations.

[0068] Fig. 8 is a diagram summarizing the residual moisture values ​​for the measurement vials on the one hand and the test or product vials on the other hand.

[0069] Fig. 9 is a graph showing temperature values ​​and other measured values ​​over time for the test or product vials in a first test freeze-drying cycle (first run).

[0070] Fig. 10 is a graph showing temperature values ​​and other measured values ​​over time for the test or product vials in a second test freeze-drying cycle (second run).

[0071] Fig. 11 shows in a temperature-time diagram three typical temperature measurement curves for exemplary measuring vials, each showing a temperature jump and a subsequent negative-exponential course.

[0072] Fig. 12 shows temperature measurement curves of measuring vials corresponding to Fig. 9, including estimated baselines of the first run, which correspond to estimated temperature limits that are to be interpreted as estimated shelf temperatures.

[0073] Fig. 13 shows temperature measurement curves of measuring vials corresponding to Fig. 10, including estimated baselines of the second run, which correspond to estimated temperature limits that are to be interpreted as estimated shelf temperatures.

[0074] Fig. 14 is a graph summarizing the estimated shelf temperatures for the measurement vials on the one hand and the test or product vials on the other hand for the two runs.

[0075] Fig. 15 shows two diagrams, of which the diagram in Fig. 15 a) summarizes the temperature difference between the last measured temperature value and the estimated shelf temperature for the two runs, and the diagram in Fig. 15 b) summarizes the estimated shelf temperatures for the two runs. Fig. 16 shows two diagrams, of which the diagram in Fig. 16 a) shows the residual moisture values ​​obtained for the measuring vials versus the shelf temperature estimated from the respective measuring vial, and the diagram in Fig. 16 b) shows the residual moisture values ​​obtained for the test or product vials versus the shelf temperature estimated from the immediately adjacent measuring vial.

[0076] The function of the control is explained below with regard to a first aspect of the invention using a first exemplary embodiment, wherein the structure is shown in Fig. 1 and the evaluation is shown in Fig. 2, which preferably includes a statistical evaluation.

[0077] The control system, described here only as an example without loss of generality, uses the temporal progression of the measured values ​​determined in individual vials using a temperature measurement system to derive not only the current temperature of the vial, but also the estimated shelf temperature for this vial. From this, the heat flow into the vial is determined. This determines the energy still required for the remaining sublimation, which can then be appropriately controlled. Furthermore, the residual moisture content of the sublimate can be determined from the remaining heat flow, for example, in a table, since sublimation requires a constant supply of heat to enable the phase transition.

[0078] For this purpose, a freeze dryer LYO1 is equipped with a multi-channel wireless temperature measurement system for vials, which preferably uses temperature measuring quartz, consisting of a radio interrogation device RF1, antennas ANT1 in the process chamber and sensors TS1 to TS3 in individually measured vials V2, V5, V7 (without individual identifiers from left to right) of the entire vial set V1 to V9. The vials equipped with sensors and whose temperature has been recorded, referred to below as measuring vials, represent a very small sample. These vials can also be aptly referred to as sample vials. The other product vials for delivery to customers do not contain a sensor. The radio interrogation device is electronically connected to the process control of the freeze dryer PC1, e.g. via a computer network, which can influence the temperature of the shelves and the negative pressure in the process chamber.The entire freeze-drying system shown schematically in Figure 1, comprising the components LYO1, RF1, ANT1 and PC1, is designated by 10 and can also be referred to briefly as the “freeze-dryer” according to the invention of the embodiment.

[0079] The sensors are preferably distributed across individual measuring vials in groups in such a way that typical influencing factors such as the distance to the edge of the storage area are represented on at least one storage area, and the measuring vials are also placed at least at one defined location on as many storage areas as possible.

[0080] This results in a sample that can capture both influences between the different storage areas and influences between different positions. It is by no means necessary to measure every position on every storage area.

[0081] The measuring vials are typically automatically inserted between the product vials during loading, e.g. by a robot and a switch, and are also automatically removed again during removal, since a product with sensors may not be delivered to the recipient.

[0082] To capture the state of the sublimation process in the vials, one would traditionally also require a very precise measurement of the shelf temperature beneath the vial to determine the heat flow. There are actually prototype systems that implement this in a very complex way using measurement pads beneath the vials, but surprisingly, this isn't even necessary with the present invention:

[0083] Fig. 2 shows three typical temperature curves (T1 to T3) within vials. The time course (approximately days) is plotted on the abscissa, and the temperature course (degrees Celsius) on the ordinate. Towards the end of the primary drying stage, there is a temperature jump of approximately 10 Kelvin (see the T / t diagram in Fig. 2). After that, the temperature approaches a plateau. This is followed by secondary drying, not shown in Fig. 2. However, even with a flat curve, there is still a temperature difference of interest between the individual plateaus (Tb1 to TB3), which roughly corresponds to the shelf temperature. Therefore, one cannot simply use the maximum of the temperature measurements from the temperature jump onwards, as this would render the measurement meaningless. Waiting until the plateau is reached is also not a solution, as control should take place beforehand.

[0084] If one considers the temperature curve starting approximately in the middle of the temperature jump, one observes a negatively exponential nonlinear curve. This is the consequence of the heat conduction differential equation, whose derivatives are proportional to the linear components of the equation in this case and can therefore essentially be solved with a function that is its own derivative, namely the exponential function.

[0085] According to the invention, this non-linearity is now used and surprisingly provides the desired shelf temperature: The temperature values ​​T(t) measured over time are calculated in the statistics module SC1 from approximately the middle of the temperature jump - not before - in a preferred embodiment using the mathematical expression

[0086] Tiin(t) = In - ( T(t) - Tb ase ) is linearized, whereby the quality of the linearization depends on the correct choice of the temperature limit Tbase. This is highest when the correct temperature limit Tbase is used, which then roughly corresponds to the shelf temperature.

[0087] Then Tiin(t) represents a straight line, whose slope and offset are initially unknown. However, these can be calculated for different Tbase values, preferably using linear regression of Tiin(t) at time t.

[0088] The quality of the linear regression can then be easily verified by determining the expected linearized temperature value for the measured time points t with temperature measurements T(t) according to the straight line calculated by linear regression and evaluating its difference to the actual linearized temperature value Tiin(t), for example, using the least squares method. The result of the normalized deviations, subtracted from one, is then the r 2 -value of the linear regression and thus the proportion of the measured values ​​explained by the straight line. For further details, please refer to the relevant mathematical literature, e.g., HEDDERICH Jürgen and SACHS Lothar, Angewandte Statistik, Berlin: Springer Nature, 2020, pages 135 to 144.

[0089] The calculation is then repeated with different temperature limit values ​​Tbase, either consecutively for a sequence of values ​​or by means of a common iteration method, e.g. a classic gradient method or an evolutionary algorithm in the case of further parameters to be optimized, until a qualitatively satisfactory quality measure, here the r 2 -value is reached. The associated temperature limit value Tbase then corresponds approximately to the estimated shelf temperature (Fig. 2 horizontal lines Tb1 to Tb3).

[0090] Now that both the shelf temperature and the temperature in the vial are known, the heat flow - heat inflow Q over time - into the vial can be estimated using the following formula:

[0091] Here, k is a material and design-dependent constant, which can be determined, for example, in a prototype run if, for example, a visualization of the heat flow is desired.

[0092] The heat flow is required, on the one hand, to heat the vial including the sublimate to the storage surface temperature, and, on the other hand, it provides the enthalpy for the sublimation itself. If sublimation essentially stops, the heat flow for this process drops to zero. Conversely, due to the limitations of the phase diagram, no arbitrary sublimation and thus vial cooling can occur after the start; rather, the throughput of the heat flow and the sublimation process itself are limited.

[0093] If the heat flow now falls below a predefined level, it can be assumed that sublimation and the stability of the sublimate cake in the vial have largely ended.

[0094] The remaining dehumidification can now advantageously be accelerated by raising the shelf temperature again, thus initiating a smooth transition to secondary drying.

[0095] As a result, the equilibrium in the phase diagram is massively shifted to the higher temperature range and thus to the gaseous state during primary drying, without the risk of a liquid intermediate state being formed by exceeding the glass transition temperature or the desired amorphous sublimate cake collapsing.

[0096] In a particularly preferred embodiment of the invention, the undershoots of the determined temperature difference and thus of the heat flow are now evaluated by means of a statistical hypothesis test in order to ensure particularly safe process control.

[0097] According to GMP, at least all measured sensors must be below the specified limit before any modification of the primary drying temperature is permitted. However, in very large plants, a very high limit can be used to prevent potential defects in individual sensors from affecting the overall process.

[0098] Advantageously, in a preferred embodiment, after all measurement times of the undershoot have been available, in addition to the mean value of the measurement times, their variance or, as the root of the variance, the sample standard deviation of the standard normal distribution is determined and the Student's t-probability distribution (PD) - which provides the correct result with a small number of sensors and unknown variance of the population - or, in a simplified manner, a standard normal distribution is adapted to this.

[0099] An inferential statistical hypothesis test is then calculated at regular intervals as time progresses. The preferred embodiment uses a Student's t-test based on the Student's t-distribution. The result of the remaining one-sided integral is checked against a predetermined probability of error for the occurrence of the alternative hypothesis "primary drying may be modified". As soon as this result is confirmed, the primary drying is modified, e.g. by increasing the shelf temperature, via the process sequence controller PFC1, also preferably a software module within the hardware process controller PC1. This transition can either be fully automatic or, for safety reasons, confirmed semi-automatically by an operator, if necessary even remotely, e.g. after video control.

[0100] Then, in a particularly preferred embodiment of the invention, the described evaluation can be started again for further temperature increases. Alternatively, in a further preferred embodiment, continuous control of the shelf temperature within a predetermined range is also possible, since from this point onwards the sublimate cake has the necessary stability for classic control. A classic PID controller can be used as the controller. Since the temperature increase or control of the shelf temperature will also change the temperature limit value Tbase, this can be initially adjusted or adjusted numerically until it is correctly redetermined, for example by adding the difference between the new and the old shelf temperature. Pre-filtering of the difference according to the inertia of the thermal system of the freeze dryer, for example using an FIR, HR or adaptive filter, is particularly suitable.

[0101] The division of the functions into blocks in Fig. 1 is by no means to be regarded as static or even mandatory; rather, thanks to modern processor technology, it is even conceivable to combine the functions of the process control and the radio interrogation device on one processor, e.g. by means of a real-time operating system or through virtualization.

[0102] As an alternative to the regular repetition of calculations such as the hypothesis tests mentioned above, if sufficient computing power is available in the control system, the time of modification of the primary drying can also be calculated in advance, for example by iterating hypothesis test parameters.

[0103] Since the preferred Student's t-test for the subsequent hypothesis test is considered robust against the normal distribution assumption, it can be used in most cases; only for freeze dryers with difficult characteristics would a non-parametric test be preferred.

[0104] In principle, according to a further preferred embodiment, it is also conceivable to individually adjust the temperature of individual setting plates assigned to specific groups by means of suitable actuators.

[0105] Furthermore, the profile of the secondary drying can be adapted to the measured deviations from the standard profile of the primary drying, e.g. by changing the slope according to specified criteria for products with very high residual moisture.

[0106] For process monitoring, it is also obvious to use a function or table, which can be determined based on test series, to visualize the degree of residual moisture of the measured vials directly in a software for the user or to document it, for example, in a database in accordance with GMP.

[0107] The technical advance of this invention consists in a simple and, thanks to the control strategy used, safe and robust control of the completion of primary drying within the framework of GMP. This ensures, on the one hand, product quality and, on the other hand, the utilization of expensive freeze dryers can be significantly improved through appropriately controlled and thus accelerated primary drying.

[0108] The embodiment presented with reference to Figures 1 and 2 relates to a freeze dryer and a control system for such a system. Instead of a conventional closed-loop control system, which is hardly feasible in practice due to a multitude of influencing factors, it uses an iterative approach to determine the shelf temperature based on recording a temperature profile. For this purpose, a temperature measurement system for individual sample vials with multiple channels is used. In each channel and the vial measured via this channel, after a characteristic temperature increase is detected, a linearization of the temperature increase over time is carried out using an iteratively determined assumed shelf temperature value. The quality of the linearization is then determined using linear regression, and the assumed shelf temperature value with the best linearization quality is used to calculate the heat flow.If the heat flow falls below a certain threshold, the shelf temperature is dynamically increased. The technical advance here lies in a simple and GMP-compliant control system during the final phase of primary drying until its accelerated completion. Additionally, the same technology can also be used to monitor the freezing temperature to protect the pharmaceutical products loaded into the freeze dryer from destruction.

[0109] The function of the control is explained below with regard to a second aspect of the invention using a second exemplary embodiment, the structure being shown in Fig. 1 and the statistical evaluation being shown in Fig. 3.

[0110] The control system described here as an example without loss of generality does not even attempt to influence the process parameters during the primary drying process. Rather, it is accepted that the process is hardly controllable mechanically. Instead, a statistical approach is used. This approach is oriented towards the goal of complete primary drying with an upper probability limit for incomplete drying. For this purpose, a freeze dryer LYO1 is equipped with a multi-channel wireless temperature measurement system for vials, which preferably uses temperature measuring quartz crystals, consisting of a radio interrogator RF1, antennas ANT1 in the process chamber, and sensors TS1 to TS3 in individually measured vials V2, V5, V7 (without individual identifiers from left to right) of the entire vial set V1 to V9. The vials equipped with sensors and whose temperatures are recorded, referred to below as measuring vials, represent a very small sample.These vials can also be aptly referred to as sample vials. The other product vials for delivery to customers do not contain a sensor. The radio interrogation device is electronically connected to the process control of the freeze dryer PC1, e.g., via a computer network. This can influence the temperature of the shelves and the negative pressure in the process chamber. The entire freeze-drying system shown schematically in Figure 1, comprising the components LYO1, RF1, ANT1, and PC1, is designated 10 and can also be referred to briefly as the inventive "freeze dryer" of the exemplary embodiment.

[0111] In a preferred embodiment, the sensors are distributed among individual measuring vials in groups in such a way that typical influencing factors such as the distance to the edge of the storage area are represented on at least one storage area, and the measuring vials are placed at least at one defined location on as many storage areas as possible.

[0112] This results in a sample that can capture both influences between the different storage areas and influences between different positions. It is by no means necessary to measure every position on every storage area.

[0113] The measuring vials are typically automatically inserted between the product vials during loading, e.g., by a robot and a diverter, and automatically removed again during removal, since a product with sensors cannot be delivered to the recipient. Thus, the sensors can detect the primary drying in the measuring vials, but not that in the product vials. Surprisingly, this isn't even necessary, and this is what makes this approach so unique:

[0114] If the criterion for the end of primary drying is assumed to be a temperature jump of at least, say, 5 degrees Kelvin within a short period of time (see the T / t diagram in Fig. 3), then the controller can assign a time (t1 to t8) at which this temperature jump occurred to each measuring vial in a statistics module SC1—preferably implemented as a software module. For this purpose, the controller has a clock that measures, for example, the seconds since the start of the process and assigns and stores this time to the sensor when the temperature jump occurs. Fig. 2 shows corresponding temperature jumps at times T1, T2, and T3.

[0115] According to GMP, at least all measured sensors should exhibit such a temperature jump before the end of primary drying is considered. However, in very large plants, a very high rate can be used to prevent potential defects in individual sensors from affecting the overall process.

[0116] Advantageously, in a preferred embodiment, after all measurement times have been obtained, in addition to the mean value of the measurement times, their variance or, as the root of the variance, the sample standard deviation of the standard normal distribution is determined and the Student's t-probability distribution (PD) - which provides the correct result with a small number of sensors and unknown variance of the population - or, in a simplified manner, a standard normal distribution is adapted to this.

[0117] An inferential statistical hypothesis test is then calculated at regular intervals as time progresses. According to the preferred embodiment, this is a Student's t-test based on the Student's t-distribution, and the result of the remaining one-sided integral (p, Fig. 3 black field) is checked against a predetermined probability of error for the occurrence of the alternative hypothesis "primary drying completed". As soon as this result is confirmed, primary drying is terminated (time tEND in Fig. 3) and the transition to the secondary drying profile is initiated by the process sequence controller PFC1 - also preferably a software module within the hardware process controller PC1. This transition can either occur fully automatically or, for safety reasons, be confirmed semi-automatically by an operator, if necessary even remotely, e.g. after video control.

[0118] The division of the functions into blocks in Fig. 1 is by no means to be regarded as static or even mandatory; rather, thanks to modern processor technology, it is even conceivable to combine the functions of the process control and the radio interrogation device on one processor, e.g. by means of a real-time operating system or through virtualization.

[0119] As an alternative to regularly repeating the hypothesis test, if sufficient computing power is available in the control system, the time at which the primary drying ends can also be calculated in advance, for example by iterating hypothesis test parameters. For example, using a Newton approximation, the time at which this p-value is reached can be iteratively determined in advance from the distribution data for a specific p-value in a one-sided test. Simple statistical distributions can, if necessary, be directly inverted mathematically using an inverse function calculated or approximated by series expansion - e.g., a Taylor series. Of course, the difference between the density function of the probability distribution used and the actual distribution function as its integral (black area at p in Fig. 3) must be noted; the latter must be used to determine the remaining error probability.

[0120] Since the preferred Student's t-test is considered robust against the normal distribution assumption, it can be used in most cases; only for freeze dryers with difficult characteristics would a non-parametric test be preferred.

[0121] It is also conceivable to use an analysis of variance, e.g., an ANOVA, to make statements about the sensors distributed across individual groups of the sample. In a particularly preferred embodiment, the worst-case drying time according to GMP can be included by dividing the variance between the groups at different positions, between the groups at different storage areas, and within the measuring vials, while simultaneously calculating the mean within the groups.

[0122] For this purpose, the worst case group is formed synthetically, for example, by additionally testing the worst mean value between the shelves with the variance from the positions and that from the measuring vials themselves, but since the worst mean value was already used for the variance between the shelves, the variance between the shelves is omitted from this test.

[0123] Alternatively, dimensionality reduction can be performed using factor analysis by deriving a suitable basis from eigenvalues ​​of the covariance matrix, thus synthetically forming the worst-case group. It goes without saying that the number of product vials in the population in each group must be taken into account in the respective hypothesis tests. This can differ significantly from the number of measurement vials, for example, if the shelf positions are divided into individual rings with a defined distance from the shelf center.

[0124] Furthermore, in a preferred embodiment, the already existing temperature measurement system can be used to carry out further inferential statistical tests, for example on the shelf temperature that is sufficient for the process, and in the case of insufficient cooling, to lower the temperature profile by a few Kelvin if necessary so that the existing load of the freeze dryer with products, which can have a considerable financial value, is saved, even if problems occur in the process.

[0125] In principle, according to a further preferred embodiment, it is also conceivable to individually adjust the temperature of individual shelves assigned to specific groups using suitable actuators. Furthermore, the secondary drying profile can also be adapted to the measured deviations from the standard primary drying profile, e.g., by adjusting the slope according to predefined criteria for products with high residual moisture.

[0126] The detection of the point in time at which the temperature rise occurs after the completion of primary drying in a measuring vial can be achieved, for example, by evaluating the gradient of the change (dT / dt) within a certain period of time using a sliding window, with subsequent detection of the completion by transitioning to a less steep phase. Corresponding gradient evaluations are easily implemented in the control software. Failures of measuring vials can also be easily handled by changing the number of measuring vials in the sample or their groups according to the relevant statistical calculation rules.

[0127] A combination with other sensors such as the Pirani vacuum gauge or the automatic calibration of the mean value of all measuring vials or the exact end point of primary drying using such a Pirani vacuum gauge is also conceivable. For this purpose, the Pirani mean value is statistically correlated with the temperature curve in the control software.

[0128] The technical advance of this invention consists in a simple and, thanks to the statistics used within the framework of GMP, safe and robust control of the completion of the primary drying, thus ensuring product quality on the one hand and significantly improving the utilization of the expensive freeze dryers on the other.

[0129] The exemplary embodiment presented with reference to Figures 1 and 3 relates to a freeze dryer and a control system for such a system. Instead of a conventional closed-loop control system, which is hardly feasible in practice due to a multitude of influencing factors, it uses a statistical approach to detect the end point of primary drying and dynamically adapt the temperature profile. For this purpose, a temperature measurement system for individual sample vials with multiple channels is used. A timestamp for the end of primary drying is assigned to each channel and the vial measured via it by detecting a characteristic temperature increase. Using a statistical hypothesis test, the point in time at which primary drying is completed in all vials of the entire load with a specified error probability is then determined. Furthermore, by dividing the sample vials into groups, accuracy can be significantly increased.The technical advance here lies in the straightforward and, thanks to the use of statistics, GMP-compliant control of the completion of primary drying. In addition, the same technology can also be used to monitor the freezing temperature to protect the pharmaceutical products loaded into the freeze dryer from destruction.

[0130] A third embodiment combines the solution features of the first and second embodiments in one embodiment of a freeze-dryer control system according to the invention and is jointly represented by Figures 1, 2, and 3. The advantages of both embodiments are achieved in combination.

[0131] Figure 4 schematically shows a fourth exemplary embodiment, namely in sub-figure 4 a) a wireless and real-time operating multi-channel temperature measuring system 20 of a freeze dryer, which has a temperature query unit 22 with a connected antenna 24 and a display 26 serving to show recorded temperature profiles, and in sub-figure 4 b) one measuring vial V of several such measuring vials, each equipped with a wireless temperature sensor, whose temperature sensor is designated TS and has a sensor antenna A that extends through a closure plug VS. In the state shown, the measuring vial V is not closed by the closure plug VS in order to enable the freeze-drying of the product contained in the vial. This product is not shown.

[0132] H denotes a flexible spacer that holds the temperature sensor TS in a radially centered position within the interior of the vial and can be formed, for example, by three or four radially extending spacer legs. Four spacer legs, each offset by 90 degrees from each other, are preferred. The spacer legs are so flexible that the temperature sensor TS can be easily inserted into and removed from the interior of the vial, for example, by pulling on the antenna A.

[0133] The height position of the temperature sensor TS within the interior can be defined by the antenna A, which is held force-fit in the sealing plug VS. The temperature sensor TS can thus be held at a greater distance from the bottom of the measuring vial than shown in Fig. 4, in order to measure the product temperature unaffected by the temperature of the vial bottom. A lower end of the temperature sensor TS, remote from the antenna A, can be designed as a temperature-sensitive temperature measuring tip.

[0134] The temperature sensors of the measuring vials assigned to the temperature query unit 22 can be individually addressed by the temperature query unit 22 so that the addressed temperature sensor transmits a temperature measurement value indicating the currently detected temperature to the temperature query unit.

[0135] The temperature query unit 22 and the temperature sensor TS can operate according to the technical principles of the publications DE 197 02 768 C1, EP 0 901 417 B1 and US 6,378,360 or according to the technical principles of the publications DE 10 2017 103 974 B3, EP 3 586 273 B1 and US 10,958,490 B2.

[0136] The antenna 24 can be integrated into the pressure vessel of the freeze dryer, for example according to the teaching of EP 4 138 212 A1 .

[0137] Specifically, the temperature sensors and temperature measurement system can be components offered by Tempris GmbH (www.tempris.com, 83607 Holzkirchen, Germany). Since they do not use cables that conduct heat and cause mess and sterility problems, nor active electronics that generate additional heat, these components are ideal for directly measuring the temperature in the substrate of individual vials. The sensor, available in various sizes for different vial types, is very small and therefore has a low heat capacity. This solution from Tempris GmbH is based on special crystals that enable direct conversion of temperature into a resonant frequency, as well as a special RF protocol for measuring the resonant frequency using low-power microwave radiation.Measurement can be performed throughout the entire freeze-drying process using a large number of sensors—at least up to 30 sensors according to current technology—in individual vials serving as measuring vials. The sensors can be automatically inserted or removed in a sterile GMP production environment using a robot and require no cabling.

[0138] The embodiments discussed above with reference to Figures 1 to 3 may comprise such a wireless multi-channel temperature measuring system as a component of the freeze dryer.

[0139] A fifth embodiment is shown in Fig. 5. A freeze-drying system 100, also referred to as a "freeze dryer" for short, is shown schematically with typical components, without claiming to be exhaustive. A control system 102 is provided, which is connected to a freeze-drying device 104, which comprises a drying chamber 106 designed as a pressure chamber and a condenser chamber 108, which are connected to one another by a line 112 provided with a valve 110.

[0140] The drying chamber 106 comprises a plurality of temperature-controlled shelves 104 arranged to receive containers referred to as vials, in particular ampoules or bottles, which contain a product to be dried.

[0141] The condenser chamber 108 includes a condenser device 114, such as plates or coils, connected to a cooling device 116. The outer surfaces of the condenser device 114 are maintained at a very low temperature (e.g., -50°C) to condense the water vapor generated during the sublimation (drying phase) of the product.

[0142] The condenser chamber 108 is connected via a line provided with a valve 118

[0143] 120 is connected to a vacuum pump 122 which serves to suck out air and to create a high vacuum value - ie a very low absolute pressure - within the condenser chamber 108 and the drying chamber 106.

[0144] The control system 102 includes a control unit 126 which serves to control the operation of the freeze-drying device 100 during the freeze-drying process, i.e., to control the temperature-controlled shelves 114, the vacuum pump 122, the cooling device 116 and the valves 110 and 118.

[0145] The control system 102 further comprises a pressure sensor device 124 and, if desired, further sensors arranged within the drying chamber 106 to detect the internal pressure prevailing therein and, if necessary, other conditions and parameters during the freeze-drying process, as well as a multi-channel temperature query unit 22 integrated into the control unit 102 with an associated antenna 24, which is arranged within the drying chamber 106. The temperature query unit 22, the antenna 24, and wireless temperature sensors inserted into measuring vials located on the shelves 114 (for example, corresponding to measuring vial V in Fig. 4) form a multi-channel temperature measuring system, which can functionally correspond to the multi-channel temperature measuring system 20 in Fig. 4.

[0146] It is also expedient to provide temperature sensors arranged directly on the shelves 104, a temperature sensor associated with the condenser device 114 and other sensors commonly used in conventional freeze dryers.

[0147] The control system 102 also includes a computing unit 128, for example a computer, connected to the control unit 126 and having a user interface for inputting operating parameters and data of the freeze-drying process, as well as storage means for storing the parameters and data and the signals associated with the pressure values.

[0148] The computing unit 128 executes a control program that implements the control method according to the invention. The computing unit preferably has a display 130 and a keyboard 132. Like the display 26 in Fig. 4, the display 130 can be used to display recorded temperature profiles, as well as to display evaluations, process parameters, current states of the freeze-drying system, positions of measuring vials, and, if desired, also positions of product vials without temperature sensors on the shelves 114 in the drying chamber 106, and whatever else may be useful to the operator of the freeze-drying system.

[0149] Elements X1 and X2 of control unit 102 symbolize that it may have further elements implemented by hardware or software, which, for example, may implement the functions of elements PFC1 and SC1 of Fig. 1, optionally in conjunction with software functionalities of the control program executed by computing unit 128. Alternatively, the functions of elements PFC1 and SC1 of Fig. 1 may be implemented entirely by software functionalities of the control program executed by computing unit 128.

[0150] Elements X1 and X2 can also represent basic hardware and software elements of the control unit 102, such as a microprocessor with RAM memory, as well as ROM and / or EPROM and / or flash memory with firmware and application software stored therein that can be executed by the microprocessor. Of course, the computing unit 128 also has such basic hardware and software elements.

[0151] By means of these functions corresponding to the elements PFC1 and SC1 of Fig. 1, the functions and advantages of the first or second embodiment, or both of these embodiments, can be achieved for the freeze-drying system of Fig. 5. Reference is made to the above explanations regarding these embodiments to avoid repetition.

[0152] In the sixth exemplary embodiment, examples of freeze-drying processes are presented below, along with subsequent evaluations and their results, to illustrate the fundamentals of the proposed inventions and developments and their validity. The freeze-drying processes discussed here are not freeze-drying processes covered by the invention that lead to a usable freeze-drying product, such as a pharmaceutical product, but rather serve solely to demonstrate the validity of the proposed inventions and developments using an arbitrarily selected test substance, in this case trehalose, a disaccharide (double sugar).

[0153] A) Preparations

[0154] The validation process procedures presented here were prepared according to the following specifications. a) Materials used

[0155] Table 1 b) Equipment c) Preparation of the formulation

[0156] For each freeze-drying cycle, trehalose was weighed and transferred into one 1000 mL and two 200 mL volumetric flasks. The trehalose was dissolved in water for injection (WFI) by stirring the substance on a magnetic stirrer until completely dissolved.

[0157] For each cycle, two trays with removable bottoms were prepared and filled with hexagonal vials (bottles or ampoules). 2.5 mL of the formulation was pipetted into each vial. d) Preparation of measuring vials and product vials

[0158] Two shelves in the drying chamber of the freeze dryer were each equipped with ten sensors in a respective 6R vial, in the configuration schematically shown in Fig. 6. The measurement vials equipped with sensors and additionally containing the formulation, which can also be aptly referred to as sample vials, are spatially distributed among a much larger number of product vials containing only the formulation, as shown in Fig. 6.

[0159] The upper shelf is shown on the left, and the lower shelf is shown on the right, as indicated by the labels "Top" and "Middle." The shelf side adjacent to the door is marked "Door." The formulation of the test product filled into the measuring vials and the product vials consists of a 5 percent trehalose solution, with a fill volume of 2.5 ml. The vials were loaded using the aforementioned tablets, which served as frames (approximately 30 mm high). The bottoms of the tablets were then removed so that all vials had direct contact with the respective shelf of the freeze dryer.

[0160] In the illustration in Fig. 6, the measuring vials are each shown with a randomly selected unique identification number, which represents the individual addressability of the temperature sensors by the freeze dryer's multi-channel temperature query unit for querying the currently measured temperature. B) Validation execution

[0161] A total of two test freeze-drying cycles were performed with the following standard parameters for trehalose:

[0162] *) Different termination criteria were applied for the two test cycles (runs 1 and 2):

[0163] 1 ) End the cycle when the Pirani pressure drops to 1.3 times the capacity pressure (50% of the capacity-Pirani difference).

[0164] 2) Termination of the cycle after the end of primary drying; indicated by the absence of a rise in the product temperature.

[0165] In both test cycles, no complete freeze-drying was carried out, but the respective freeze-drying process was terminated prematurely, before the end of the primary drying (Fig. 11 , run 1 ) or after the end of the primary drying (Fig. 12, run 2), i.e. without the secondary drying normally required to provide a freeze-dried product.

[0166] C) Initial evaluation using a Karl Fischer titration (KFT)

[0167] After the freeze-drying process was completed and the vials were sealed, they were immediately transferred to a preconditioned glove box filled with dry air (humidity < 0.1%). The vials were opened, and the product was homogenized with a spatula for 60 seconds. Approximately 50 mg was transferred to Karl Fischer vials and sealed.

[0168] The titration was performed using an 831 KF Coulometer. Water was removed from the samples by heating the 832 KF Thermoprep Oven System (Deutsche METROHM GmbH & Co. KG, Filderstadt, Germany) to 100 °C. Measurements were performed while purging the system with dry nitrogen at 60 ml / min.

[0169] For the evaluation, product vials immediately adjacent to the measuring vials were selected according to the scheme shown in Fig. 6. The selected product vials are shown in dark color.

[0170] Fig. 7 shows the residual moisture content determined in this way ([%], vertical axis of the diagram) for the vials subjected to Karl Fischer titration, identified by the array positions (horizontal axis of the diagram). Specific vial positions, "edge" (i.e., at the edge of the shelf) and "corner" (i.e., at a corner of the shelf), are marked in the diagram.

[0171] The question arises whether there is a difference (%) in humidity (bias) between the measuring vials with the Tempris sensors (also referred to as Tempris vials) and the neighboring product vials.

[0172] Fig. 8 now compares the residual moisture results for the measuring vials and the adjacent product vials, as obtained from the two cycles, using Bartlett's (Levene) test for equivalent variance.

[0173] The Kolmogorov-Smirnov test signals an alternative hypothesis (p<0.01), i.e. no normal distribution.

[0174] However, for larger n, the t-test can be assumed to be robust against violations of normality (see Eid Gollwitzer Schmitt (2017), Statistics and Research Methods, p. 371). Therefore, the pairs of differences from the two runs (cycles) can be combined.

[0175] It turns out that all groups have similar deviations, so they can be combined in a single t-test.

[0176] Using the t-test, the following assessment of possible moisture bias is obtained:

[0177] The result of the t-test for the measurement vials versus the adjacent product vials is t(19) = -0.26 (two-sided); d=0.04; p=0.80; 1 -ß>0.80 (post-hoc).

[0178] The descriptive statistics of the difference between measuring vials and neighboring product vials show

[0179] M = 0.0045; SD = 0.11

[0180] Run 1 : M = -0.056

[0181] Run 2: M = 0.047

[0182] This yields the following positive result: There is no relevant humidity difference (distortion, bias) between the measuring vials and neighboring product vials. This applies even in the worst-case scenario of a positional shift between the measuring vials and the neighboring product vials. The measuring vials are therefore excellent representatives of the product vials.

[0183] The temperatures measured by Tempris temperature sensors in the measuring vials during both runs are shown in Fig. 9 (Run 1) and Fig. 10 (Run 2), along with other process parameters and measured values, including the condenser temperature, the shelf temperature measured directly on the shelf, a target shelf temperature, the pressure in the drying chamber, and the Pirani pressure. Run 1 was terminated once a Pirani pressure of 78 mTorr was reached.

[0184] During run 2, an extended initial freezing time occurred, caused by access restrictions to process parameters in the freeze dryer used. However, the data representing this extended initial freezing time was removed from the data set underlying Fig. 10, so that the temperature curves in Fig. 10 essentially correspond to the temperature curves that would have occurred without such an extended initial freezing time, corresponding to the situation underlying Fig. 9. However, the corresponding Fig. 13 shows this extended initial freezing time, which is completely irrelevant for the validation of the invention and further development proposals.

[0185] All of these figures, as well as Fig. 12, which corresponds to Fig. 9, serve only as a general illustration, so the different temperature curves do not need to be differentiated and assigned to a respective sensor. However, such differentiation and assignment are, of course, readily possible in practice, for example, when a color display is present on a screen, and this differentiation and assignment naturally results from the data underlying the diagrams, which can be further processed.

[0186] From such temperature curves, the individual shelf temperature at the position of the respective measuring vial can now be determined, based on the simplified heat transfer equation:

[0187] The negative exponential temperature curve after the completion of sublimation is: dm In addition, reference is made to the explanations above for the first embodiment, in which the heat flow m with Q and the material and design-dependent constant k mentioned there corresponds to the inverse of the material and design-dependent constant ki used here.

[0188] The constant ki (here also referred to as k1) is the product of the specific heat of sublimation of water Hsub (energy requirement per unit mass to be sublimated) and the heat of sublimation referred to in the relevant literature as k vThe value referred to as the value between the temperature difference and the area-neutral power per unit of temperature difference, which must be multiplied by the contact area A of an individual vial on the support plate and the temperature difference. This depends, among other things, on design details, particularly of the vial. The result of multiplying by the temperature difference is the actual heat output. If the heat of sublimation is divided by this heat output, the relationship between the sublimation rate in the form of a mass unit per unit of time and the temperature difference is obtained.

[0189] The constant k2 describes, similar to a radioactive decay process, the dependence of the sublimation rate on the remaining mass after the transition from the saturation region, which is determined by the maximum usable sublimation capacity, to the region limited by the remaining mass. This constant is also material- and process-dependent.

[0190] Both constants result indirectly from the method according to the invention, in particular through the application of the statistical method of linear regression, and do not need to be explicitly introduced into it.

[0191] By integrating the temperature difference during the entire run, knowing the original ice mass, the k vvalue can be determined or, conversely, the residual moisture can be estimated. It is important to know the individual support surface temperature under the measured vial in order to be able to determine the temperature difference to the substrate within the vial. This does not exactly correspond to the temperature set by the support surface heating, e.g. using silicone oil, since vials that are located in front of the measured vial on the support surface's heating coil cool it down. The great advantage of the method according to the invention is that the time series can be used to determine the individual temperature of the support surface even if the temperature of the via has not yet fully converged to it.

[0192] A proposed algorithm for iteratively determining shelf temperature includes the following steps:

[0193] 1 . Find the transition to the exponential section after the temperature increase.

[0194] 2. Linearize with Tim = In ( - T + Tbase).

[0195] 3. Try to fit the linear output using linear regression.

[0196] 4. Calculate the goodness of fit using the r-value of linear regression.

[0197] 5. Iteratively try different Tbase values ​​and use the best fitting Tbase value.

[0198] This algorithm is completely independent of the formulation, the size and fill level of the measuring vial, etc.

[0199] Fig. 11 shows three typical temperature diagrams (temperature [°C] versus time [minutes]) of three different measuring vials. "BL" stands for baseline, which corresponds to the shelf temperature, "NExpT" stands for "negative exponential trend," which begins at the point identified by the arrow, and "Delta" represents the difference between the actual instantaneous vial temperature at the latest recorded time in the diagram and the shelf temperature determined or estimated from the negative exponential temperature curve.

[0200] Figures 12 and 13 show the temperature diagram obtained using the Tempris temperature sensors in the measuring vials for the two runs (Fig. 14: Run 1; Fig. 13: Run 2; each temperature diagram (temperature [°C] versus time [minutes]) along with the baselines determined by applying the algorithm and thus the resulting shelf temperature. This shelf temperature determined in this way enables more reliable control or regulation of a freeze-drying process, better than temperature values ​​obtained using a temperature sensor arranged on the shelf itself.

[0201] For the sake of completeness, it should be noted that, due to an access restriction of the freeze dryer used, the system remained in the initial freezing state for significantly longer than in Run 1, as can be seen from a comparison of Figures 12 and 13. However, this is not a relevant difference between the two runs for the subsequent freeze-drying process for validation.

[0202] E) Further evaluations a of the data obtained for the two runs

[0203] Fig. 14 now compares the obtained shelf temperatures using Bartlett's (Levene) test for equivalent variance. It shows that all groups have similar deviations, so they can be combined in a single t-test.

[0204] The Kolmogorov-Smirnov test accepts a null hypothesis, i.e. a normal distribution (p=0.063). The t-test for the alternative hypothesis that there is a difference between the two

[0205] Runs at the estimated shelf temperature yield t(18) = 1 .48 (two-sided); d=0.04; p=0.16; 1 -ß>0.49 (post-hoc)

[0206] The descriptive statistics of the difference between the two runs show

[0207] M = 0.29, SD = 0.87

[0208] The null hypothesis is therefore valid. This means that there is no difference in the measured shelf temperature between the two runs.

[0209] This evaluation shows that the algorithm for the time series of shelf temperature and the exponential parameter fitting performed to determine Tbase (also referred to as “exponential fitting”) provides valid results.

[0210] Fig. 15 compares in Fig. 15 a) the temperature difference between the last measured temperature, i.e. the temperatures inside the measuring vials at the time of premature termination of the respective freeze-drying cycle, and the obtained shelf temperature for the two runs, and in Fig. 15 b) the obtained shelf temperatures for the two runs.

[0211] The temperature difference between the last measured temperature and the obtained control temperature is proportional to the last heat transfer that occurred, i.e., the heat flow that occurred at the time of termination. c) Consideration of the residual moisture results

[0212] Fig. 16 compares in Fig. 16 a) the residual moisture values ​​obtained by means of the Karl Fischer titration (KFT) (vertical axis of the diagram) for the measuring vials in relation to the respective temperature difference (horizontal axis of the diagram) between the last measured temperature and the obtained control temperature, and in Fig. 16 b) the residual moisture values ​​obtained by means of the Karl Fischer titration (KFT) (vertical axis of the diagram) for the product vials immediately adjacent to the measuring vials in relation to the respective temperature difference (horizontal axis of the diagram) between the last measured temperature of the immediately adjacent measuring vial and the obtained control temperature.

[0213] Once again it is clear that the measuring vials are excellent representatives of the product vials.

[0214] The data underlying the two diagrams allow for further evaluations that take into account the positions of the measuring vials on the shelves and can be helpful for the optimization of fully completed freeze-drying processes, including secondary drying, for the provision of freeze-dried products. and views

[0215] Based on the principles and proposals presented and explained above and in the overall disclosure herein, the skilled person is able to establish a model of the freeze-drying process (lyo-process) into which the temperatures of a large number of measuring vials and the shelf temperatures determined for the individual measuring vials, and thus the heat flows (in particular by adjusting the kv constants), are included as additional real-time input, and into which the positions of these measuring vials are included. With such a model, already established models of the freeze-drying process, which are well known to the skilled person, can then be supplemented and calibrated in order to make them robust against changes in the parameters involved, thus obtaining robust model results.

[0216] This enables robust simulation and thus robust control of the freeze-drying process carried out by a freeze dryer. The core of the approach presented here and the resulting application and further development possibilities is the shelf temperature determination enabled by the present invention, which is completely independent of the vial types used and their size, the FD type, the formulation, etc. The shelf temperature determination is also independent of the freeze dryer used, since its relevant properties can be easily taken into account by adapting at least one constant used in the proposed algorithm.

[0217] However, the shelf temperature determination enabled by the invention can also be advantageously used without such modeling of the freeze-drying process to optimize the parameters of a conventional freeze-drying process and shorten the required process time until completion of the primary drying phase, for example, by adjusting the shelf temperature or shelf temperatures during ongoing freeze-drying operation depending on the temperatures measured by the measuring vials. Such optimizations, to ensure the required quality while simultaneously avoiding unnecessarily long process times until completion of the primary drying phase, can also be carried out heuristically based on test freeze-drying cycles.In particular, the control or regulation logic suitable for shortening the process times for adjusting the shelf temperature or temperatures during ongoing freeze-drying operation can be heuristically designed and optimized depending on the temperatures measured by means of the measuring vials.

[0218] In particular, further analysis and modeling of the residual moisture as a function of heat flow is also considered, in particular using statistical tests, for example on the basis of data such as those shown in Fig. 16.

[0219] The invention and further development proposals explained above using the second exemplary embodiment (see in particular Fig. 3) can also advantageously contribute to avoiding unnecessarily long process times for the primary drying phase. The proposed statistical live analysis enables the determination of a moisture content below a defined limit during the freeze-drying run. The live statistics then make it possible to determine the probability that all vials, i.e., the measuring vials and the product vials, are below this limit. This allows the freeze-drying cycle to be shortened and generally improved in the event of problems. Better and more complete measurements enable both a faster freeze-drying cycle and better quality of the resulting freeze-dried products.

[0220] The invention and further development proposals herein are motivated, among other things, by the following considerations:

[0221] - To date, there is no measurement of heat transfer in a larger number of individually measured vials in production environments.

[0222] - Knowledge of heat transfer allows us to gain insights into the remaining moisture.

[0223] - Knowledge of the remaining moisture in production allows for shorter freeze-drying cycles (so-called lyo cycles) and / or better quality.

[0224] - Wireless temperature measurement using Tempris temperature sensors inserted into measuring vials enables live heat transfer measurements, if necessary using time series and calibration of the k-value (a material and design-dependent constant, see the explanations above for the first embodiment) within simulations.

[0225] According to the first aspect, the invention also provides, inter alia, the following items:

[0226] A1. A control system for a freeze dryer (LYO1), comprising at least one temperature measuring system which can record the temperature in at least one random sample set from a total quantity of vials - vessels with contents to be dried - (V1 to V9) for each vial in the sample, whereby even a sample size of only one vial is permissible, and at least one process control system (PFC1) which can control the temperature of at least one storage area of ​​the vials in the primary drying phase according to a predetermined temperature profile over time, characterized in that during the primary drying phase, a temperature jump is recorded for each measured vial in the sample as a result of completed primary drying - which occurs when the largely completed sublimation essentially no longer extracts any sublimation heat from the substrate to be dried in the vial -From the temperature jump thus recorded in at least one measured vial, the essentially negative exponential curve of the temperature in the vial against a temperature limit value - which approximately corresponds to the shelf temperature - is used to estimate the temperature limit value using the non-linearity of this curve. If a difference between the measured temperature and the estimated temperature limit value falls below the threshold value - in the case of several vials, this difference is also summarized using statistical methods such as averaging or hypothesis tests and recognized as an undershoot - the process control increases the shelf temperature in order to increase the temperature difference and thus cause a greater heat flow into the vial.

[0227] A2. Control system for a freeze dryer according to item A1, characterized in that an evaluation of the estimated temperature limit value is carried out by calculating the difference between this and the measured temperature over time, logarithmizing this difference and then performing a linear regression against the time course on the logarithmized values ​​and determining a quality measure for their linearity over this time.

[0228] A3. Control system for a freeze dryer according to one of the items A1 to A2, characterized in that the estimated temperature limit is determined by evaluating at least one assumed estimated temperature limit using a quality measure and then iteratively changing the assumed estimated temperature limit so that the quality measure is improved, wherein established numerical iteration methods such as a gradient search or an evolutionary algorithm can be used for this purpose in the case of further parameters to be optimized.

[0229] A4. Control system for a freeze dryer according to one of the items A1 to A3, characterized in that the estimated temperature limit is determined by evaluating a plurality of statically assumed estimated temperature limit values, and selecting from this set the one with the best evaluation according to a quality measure, whereby a starting value for a subsequent iterative numerical determination can also be found in this way.

[0230] A5. Control system for a freeze dryer according to one of the items A1 to A4, characterized in that the mean value and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution are calculated over the measuring times of the temperature difference undershoots and these are fed to a suitable hypothesis test.

[0231] A6. Control system for a freeze dryer according to any one of items A1 to A5, characterized in that after the shelf temperature has been raised, the temperature difference to an estimated temperature limit is again calculated and if the shelf temperature falls below this limit again, the shelf temperature is raised again, wherein the estimated temperature limit can also be mathematically adjusted in accordance with the increase.

[0232] A7. Control system for a freeze dryer according to any one of items A1 to A6, characterized in that after raising the shelf temperature, the temperature difference to an estimated temperature limit is again calculated and adjusted to a desired value, whereby a conventional PID controller with a limited adjustment range can be used for this purpose, whereby the estimated temperature limit can also be mathematically adjusted according to the adjustment.

[0233] A8. Control system for a freeze dryer according to one of the items A1 to A7, characterized in that a residual moisture content in at least one vial is estimated and visualized or documented from the temperature difference(s), for which purpose both a formula-based and a tabular conversion can be used.

[0234] A9. Freeze dryer, characterized in that a control system according to one of the items A1 to A8 is used to control the temperature of the vial storage areas.

[0235] A10. Freeze dryer according to item A9, characterized in that the temperature profile can be individually influenced by the control system for each shelf or in groups of shelf.

[0236] A11. A method for controlling a freeze dryer (LYO1), which consists of at least one temperature measuring system, which can record the temperature in at least one random set from a total quantity of vials - vessels with contents to be dried - (V1 to V9) for each vial in the sample, whereby even a sample size of only one vial is permissible, and at least one process control (PFC1), which can control the temperature of at least one storage area of ​​the vials in the primary drying phase according to a predetermined temperature profile over time, characterized in that during the primary drying phase, a temperature jump is recorded for each measured vial in the sample as a result of completed primary drying - which occurs when the largely completed sublimation essentially no longer extracts any sublimation heat from the substrate to be dried in the vial -From the temperature jump thus recorded in at least one measured vial, the essentially negative exponential curve of the temperature in the vial against a temperature limit value - which approximately corresponds to the shelf temperature - is used to estimate the temperature limit value using the non-linearity of this curve. If a difference between the measured temperature and the estimated temperature limit value falls below the threshold value - in the case of several vials, this difference is also summarized using statistical methods such as averaging or hypothesis tests and recognized as an undershoot - the process control increases the shelf temperature in order to increase the temperature difference and thus cause a greater heat flow into the vial.

[0237] A12. Method according to item A11, characterized in that an evaluation of the estimated temperature limit value is carried out by calculating the difference between this and the measured temperature over time, logarithmizing this difference and then performing a linear regression against the time course on the logarithmized values ​​and determining a quality measure for their linearity over this time.

[0238] A13. Method according to one of the items A11 to A12, characterized in that the estimated temperature limit is determined by evaluating at least one assumed estimated temperature limit using a quality measure and then iteratively changing the assumed estimated temperature limit so that the quality measure is improved, wherein established numerical iteration methods such as a gradient search or an evolutionary algorithm can be used for this purpose in the case of further parameters to be optimized.

[0239] A14. Method according to one of the items A11 to A13, characterized in that the estimated temperature limit is determined by evaluating a plurality of statically assumed estimated temperature limit values, and selecting from this set the one with the best evaluation according to a quality measure, whereby a starting value for a subsequent iterative numerical determination can also be found in this way.

[0240] A15. Method according to one of the items A11 to A14, characterized in that the mean value and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution are calculated over the measuring times of the temperature difference undershoots and these are fed to a suitable hypothesis test.

[0241] A16. Method according to one of the items A11 to A15, characterized in that after the raising of the storage surface temperature, a calculation of the temperature difference to an estimated temperature limit is carried out again and if this temperature falls below the storage surface temperature again, the storage surface temperature is raised again, wherein the estimated temperature limit can also be mathematically adjusted in accordance with the increase.

[0242] A17. Method according to one of the items A11 to A16, characterized in that after raising the shelf temperature, the temperature difference to an estimated temperature limit is again calculated and adjusted to a desired value, wherein a conventional PID controller with a limited adjustment range can be used for this purpose, wherein the estimated temperature limit can also be mathematically adjusted according to the adjustment.

[0243] A18. Method according to one of the items A11 to A17, characterized in that a residual moisture content in at least one vial is estimated from the temperature difference(s) and visualized or documented, for which purpose both a formula-based and a tabular conversion can be used. A19. Method according to one of the items A11 to A18, characterized in that a control system according to one of the items A1 to A8 is used to control the temperature of the storage surfaces of the vials.

[0244] A20. Method according to one of the items A11 to A19, characterized in that the temperature profile is individually influenced by the control system for each setting plate or in groups of setting plates.

[0245] According to the second aspect, the invention also provides, inter alia, the following items:

[0246] B1. A control system for a freeze dryer (LYO1), comprising at least one temperature measuring system which can record the temperature in at least one random set of a total quantity of vials - vessels with contents to be dried - (V1 to V9) for each vial in the sample, and at least one process control system (PFC1) which can control the temperature of at least one storage area of ​​the vials in the primary drying phase according to a predetermined temperature profile over time, characterized in that during the primary drying phase, a temperature jump is recorded for each measured vial in the sample as a result of completed primary drying - which occurs when the largely completed sublimation essentially no longer extracts any sublimation heat from the substrate to be dried in the vial -at least one measurement time (t1 to t8) for this vial is assigned to the temperature jump recorded in this way for a measured vial of the sample; after recording a specified rate or number of temperature jumps at the measurement times recorded in this way, at least one inferential statistical hypothesis test is calculated which makes a statement as to whether, at the current or a specified time (tEND), the end of primary drying can be assumed for all vials with a specified probability of error; this hypothesis test is either repeated with the actual advancing time for these vials if the test has so far concluded that primary drying is not yet complete, or - by iteration or mathematical inversion - the time of the assumed end of primary drying (tEND) is numerically calculated in advance, with the statement of the hypothesis test,that at a time determined in this way, the end of the primary drying can be assumed, the primary drying is abbreviated or terminated from this time onwards and the process flow control proceeds to the subsequent process steps, whereby the abbreviation or termination can either be carried out directly by the control system or is suggested to a plant operator for manual release.

[0247] B2. Control system for a freeze dryer according to item B1, characterized in that the mean value and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution is calculated over the measuring times and fed into the hypothesis test.

[0248] B3. Control system for a freeze dryer according to item B1 or B2, characterized in that, depending on the positioning of the measured vials of the sample in the freeze dryer on the shelves or between the shelves, an assignment to different groups is carried out within the framework of the statistical evaluation, for which purpose a factor analysis for statistical dimension reduction can also be used.

[0249] B4. Control system for a freeze dryer according to item B3, characterized in that the mean value or the variance or the standard deviation or at least one other statistical distribution measure adapted to the distribution is distributed among the groups, wherein the groups are statistically weighted depending on the positioning of the measured vials contained therein on the storage area and the number of further non-measured vials with a technically comparable positioning.

[0250] B5. Control system for a freeze dryer according to any one of items B1 to B4, characterized in that the hypothesis test is a Student's t-test or a test against the standard normal distribution or a variance test - ANOVA - or a non-parametric test.

[0251] B6. Control system for a freeze dryer according to one of the items B1 to B5, characterized in that the control system first monitors the attainment of the necessary freezing temperature of the vials based on the measured vials - if necessary also by means of a hypothesis test for the hypothesis of the attainment of this freezing temperature - and, in the event of an assumed non-attainment of the freezing temperature, further lowers the shelf temperature.

[0252] B7. Control system for a freeze dryer according to one of the items B1 to B6, characterized in that a wireless temperature measuring system (RF1) is used.

[0253] B8. Control system for a freeze dryer according to one of the items B1 to B7, characterized in that the temperature profile in the secondary drying phase is adapted or selected according to the previous course of the primary drying phase.

[0254] B9. Freeze dryer, characterized in that a control according to one of the items B1 to B8 is used to control the temperature of the storage areas of the vials.

[0255] B10. Freeze dryer according to item B9, characterized in that the temperature profile for each shelf or in groups of shelf can be individually influenced by the control system by means of actuators. B11. Method for controlling a freeze dryer (LYO1), which consists of at least one temperature measuring system which can record the temperature in at least one sampled set from a total quantity of vials - vessels with contents to be dried - (V1 to V9) for each vial in the sample, and at least one process control system (PFC1) which can control the temperature of at least one storage area of ​​the vials in the primary drying phase according to a predetermined temperature profile over time, characterized in that during the primary drying phase, a temperature jump is recorded for each measured vial in the sample as a result of completed primary drying - which occurs,when the largely completed sublimation essentially no longer extracts any sublimation heat from the substrate to be dried in the vial - , the temperature jump thus recorded for a measured vial of the sample is assigned at least one measurement time (t1 to t8) for this vial, after recording a defined rate or number of temperature jumps at the measurement times thus recorded, at least one inferential statistical hypothesis test is calculated, which makes a statement as to whether at the current or a predetermined time (tEND) in all of all vials, an end of the primary drying can be assumed with a predetermined probability of error, this hypothesis test is either repeated for these with the actually advancing time if the test so far has concluded that the primary drying is not yet finished,or - by iteration or mathematical inversion - the time of the assumed end of the primary drying (tEND) is calculated numerically in advance, with the statement of the hypothesis test that at a time thus determined, the end of the primary drying can be assumed, the primary drying is abbreviated or terminated from this time onwards and the process flow control proceeds to the subsequent process steps, whereby the abbreviation or termination can either be carried out directly by the control system or is suggested to a plant operator for manual release.

[0256] B12. Method according to item B11, characterized in that the mean and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution is calculated over the measurement times and fed into the hypothesis test.

[0257] B13. Method according to item B11 or B12, characterized in that, depending on the positioning of the measured vials of the sample in the freeze dryer on the shelves or between the shelves, an assignment to different groups is carried out within the framework of the statistical evaluation, for which purpose a factor analysis for statistical dimension reduction can also be used.

[0258] B14. Method according to item B13, characterized in that the mean or the variance or the standard deviation or at least one other statistical distribution measure adapted to the distribution is distributed among the groups, wherein the groups are statistically weighted depending on the positioning of the measured vials contained therein on the storage area and the number of further non-measured vials with a technically comparable positioning.

[0259] B15. Method according to one of the items B11 to B14, characterized in that the hypothesis test is a Student's t-test or a test against the standard normal distribution or a variance test - ANOVA - or a non-parametric test.

[0260] B16. Method according to one of the items B11 to B15, characterized in that firstly the attainment of the necessary freezing temperature of the vials is monitored on the basis of the measured vials - if necessary also by means of a hypothesis test to determine the hypothesis of the attainment of this freezing temperature - and in the event of an assumed non-attainment of the freezing temperature, the storage surface temperature is additionally lowered.

[0261] B17. Method according to one of the items B11 to B16, characterized in that a wireless temperature measuring system (RF1) is used.

[0262] B18. Control system for a freeze dryer according to one of the items B11 to B17, characterized in that the temperature profile in the secondary drying phase is adapted or selected according to the previous course of the primary drying phase.

[0263] B19. Method according to one of the items B11 to B18, characterized in that a control according to one of the items B1 to B8 is used to control the temperature of the storage surfaces of the vials.

[0264] B20. Method according to one of the items B11 to B19, characterized in that the temperature profile for each setting plate or in groups of setting plates is individually influenced by the control system by means of actuators.

[0265] The invention provides, according to the first aspect, the second aspect and further Among other things, the following items are also available:

[0266] C1. A control system for a freeze dryer (LYO1) designed to subject a total number of vessels (V1 to V9) referred to as vials with contents to be dried to a freeze-drying process, comprising at least one temperature measuring system designed to detect a current temperature in at least one sample vial from the total number of vials (V1 to V9), and at least one process sequence controller (PFC1) designed to control the temperature of at least one storage area of ​​the vials in the freeze dryer, at least in a primary drying phase, according to a predetermined temperature profile over time;characterized in that the process flow controller (PFC1) is further configured to a) detect a characteristic temperature rise in the sample vial during the primary drying phase, reflecting a progress of the primary drying, and b) carry out the further freeze-drying process in a defined manner depending on the detected characteristic temperature rise or a plurality of characteristic temperature rises detected for several separate sample vials.;

[0267] C2. Control system for a freeze dryer (LYO1) according to item C1, characterized in that the process sequence control system (PFC1) is configured to determine, for the further execution of the freeze-drying process, on the basis of the detected characteristic temperature increase or the plurality of characteristic temperature increases detected for a plurality of separate sample vials, whether and in what manner the execution of the freeze-drying process is to be influenced or to propose an influence on the execution of the freeze-drying process to an operator of the freeze-dryer for approval.

[0268] C3. Control for a freeze dryer (LYO1) according to item C1 or C2, characterized in that the process flow control (PFC1) is further configured to control at least one temperature setting device of the freeze dryer depending on the detected characteristic temperature increase or the plurality of characteristic temperature increases detected for several separate sample vials.

[0269] C4. Control system for a freeze dryer (LYO1) according to one of the items C1 to C3, characterized in that the characteristic temperature rise is a temperature jump or comprises a temperature jump which occurs as a result of a completed primary drying when the largely completed sublimation essentially no longer extracts any sublimation heat from the contents of the sample vial to be dried, possibly referred to as the substrate.

[0270] C5. Control system for a freeze dryer (LYO1) according to one of the items C1 to C4, characterized in that the characteristic temperature rise is the temperature jump with a temporally subsequent rising temperature profile or comprises the temperature jump with a temporally subsequent rising temperature profile.

[0271] C6. Control for a freeze dryer (LYO1) according to one of the items C1 to C5, characterized in that the process sequence control (PFC1) is further configured to record and evaluate the characteristic temperature rise or a / the temperature jump characterizing this with a substantially negative-exponential course of the temperature in the sample vial from the temperature jump towards a temperature limit value in order to estimate the temperature limit value from a non-linearity of a / this course of the temperature.

[0272] C7. Control for a freeze dryer (LYO1) according to item C6, characterized in that the process sequence control (PFC1) is further configured to model, on the basis of a plurality of assumed different provisional temperature limit values, the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump as a linearized logarithmic temperature profile, in that the process sequence control (PFC1) adapts temperature-time measured value pairs representing the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump to a model straight line equation describing the respective linearized logarithmic temperature profile, and in that the process sequence control (PFC1) is further configured toto compare the model straight line equations obtained for the various provisional temperature limits with each other and / or with the temperature-time measurement pairs in order to estimate the best-fitting provisional temperature limit as the temperature limit.

[0273] C8. Control according to item C7, characterized in that the process sequence control (PFC1) is configured to iteratively and / or stochastically select different provisional temperature limit values ​​in order to model the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump as a linearized logarithmic temperature course.

[0274] C9. Control according to one of the items C6 to C8, characterized in that the process sequence control (PFC1) is set up to select different provisional temperature limit values ​​within at least partially sequentially carried out modeling processes and / or within at least partially parallel carried out modeling processes in order to model the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump as a linearized logarithmic temperature course.

[0275] C10. Control according to one of the items C6 to C9, characterized in that the process sequence control (PFC1) is set up to subject the model straight line equations obtained for the various provisional temperature limit values ​​to a quality assessment by determining a summary quality measure in each case, which evaluates a linear temperature-time curve corresponding to the respective model straight line equation on the basis of the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value on the basis of this quality assessment.

[0276] C11 . Control for a freeze dryer (LYO1 ) according to one of the items C6 to C10, characterized in that the process sequence control (PFC1 ) is designed to change a current shelf temperature depending on the temperature limit value by means of the / a temperature setting device of the freeze dryer, which is preferably assigned to at least one shelf of the freeze dryer, wherein this change is preferably based on an interpretation of the temperature limit value as an approximate shelf temperature.

[0277] C12. Control for a freeze dryer (LYO1) according to one of the items C6 to C11, characterized in that the process sequence control (PFC1) is set up to determine a temperature difference between the current temperature detected for the sample vial and the estimated temperature limit value and to subject this temperature difference to an undershoot test with respect to a predetermined minimum difference, wherein the process sequence control (PFC1) is further set up to raise a current shelf temperature in the event of a detected undershoot of the minimum difference in order to increase the temperature difference and thus bring about a greater heat flow into the sample vial.

[0278] C13. Control system for a freeze dryer (LYO1) according to item C12, characterized in that the process flow control system (PFC1) is configured to determine a combined temperature difference in the case of several sample vials and to subject it to the undershoot test in order to raise the current shelf temperature in the case of a detected undershoot of the minimum difference.

[0279] C14. Control system for a freeze dryer (LYO1) according to item C13, characterized in that the process flow control system (PFC1) is configured to determine the combined temperature difference from the instantaneous temperatures and estimated temperature limit values ​​recorded for the plurality of sample vials using at least one statistical method such as averaging or hypothesis testing.

[0280] C15. Control system for a freeze dryer according to one of the items C12 to C14, characterized in that the process sequence control (PFC1) is designed to calculate a mean value and a variance or standard deviation or at least one other statistical distribution measure adapted to the distribution over measuring times of temperature difference undershoots and to apply a hypothesis test to these.

[0281] C16. Control system for a freeze dryer according to one of the items C12 to C15, characterized in that the process sequence control (PFC1) is configured to determine, after the shelf temperature has been raised, a temperature difference again from an estimated temperature limit value, which may be adjusted if desired in accordance with the rise in the shelf temperature, and to subject this temperature difference to the undershoot test with respect to the predetermined minimum difference in order to raise the shelf temperature again if the minimum difference is again undershot.

[0282] C17. Control system for a freeze dryer according to one of the items C12 to C16, characterized in that the process sequence control (PFC1) is designed to provide a control, if desired PID control with a limited control range, of the temperature difference to an estimated temperature limit value after raising the shelf temperature, which comprises a renewed determination of the temperature difference to an estimated temperature limit value, if desired adapted in accordance with the control system.

[0283] C18. Control system for a freeze dryer according to one of the items C12 to C17, characterized in that the process sequence control (PFC1) is designed to estimate a residual moisture content in at least the sample vial or at least one of the sample vials from the temperature difference or the temperature differences.

[0284] C19. Control system for a freeze dryer (LYO1) according to one of the items C6 to C18, characterized in that the process sequence control system (PFC1) is designed to determine a quality measure for the estimated temperature limit value used in the further process control system.

[0285] C20. Control system for a freeze dryer (LYO1) according to one of the items C6 to C19, characterized in that the process sequence control system (PFC1) is designed to carry out an evaluation of the estimated temperature limit value by calculating the difference between this and the measured temperature over time, logarithmizing this difference and then carrying out a linear regression against the time course using the logarithmized values ​​and using this to determine a quality measure for their linearity.

[0286] C21 . Control for a freeze dryer (LYO1 ) according to one of the items C6 to C20, characterized in that the process sequence control (PFC1 ) is set up to determine the estimated temperature limit value used in the further process control by determining a quality measure for iteratively changed estimated temperature limit values ​​or from several assumed estimated temperature limit values, and from this an estimated temperature limit value with an improved quality measure is used as the basis for the further process control.

[0287] C22. Control system for a freeze dryer (LYO1) according to one of the items C6 to C21, characterized in that the process sequence control system (PFC1) is configured to determine the estimated temperature limit by evaluating this quality measure based on at least one assumed estimated temperature limit and then iteratively changing the assumed estimated temperature limit such that the quality measure improves, preferably using established numerical iteration methods such as a gradient search or an evolutionary algorithm in the case of further parameters to be optimized. C23.Control system for a freeze dryer (LYO1) according to one of the items C6 to C22, characterized in that the process sequence control system (PFC1) is designed to determine the estimated temperature limit value by evaluating each of a plurality of statically assumed estimated temperature limit values ​​and selecting from this set the one with the best evaluation according to a quality measure in order to use it as a basis for further process control or as a starting value for a subsequent iterative numerical determination of the estimated temperature limit value on which further process control is to be based.

[0288] C24. Control for a freeze dryer (LYO1) according to one of the objects C1 to C23, characterized in that the temperature measuring system is designed to detect a respective instantaneous temperature in a plurality of sample vials from the total number of vials (V1 to V9), and in that the process sequence control (PFC1) is set up to a) detect, during the primary drying phase, characteristic temperature increases reflecting progress in the primary drying in the form of temperature jumps in the sample vials, and b) carry out the further freeze-drying process in a defined manner depending on the detected temperature jumps.

[0289] C25. Control system for a freeze dryer (LYO1) according to item C24, characterized in that the process sequence control system (PFC1) is configured to determine, for the further execution of the freeze-drying process, on the basis of the detected temperature jumps, whether the execution of the freeze-drying process is to be influenced or whether an influence on the execution of the freeze-drying process is to be proposed to an operator of the freeze-dryer for approval.

[0290] C26. Control for a freeze dryer (LYO1) according to item C24 or C25, characterized in that the process sequence control (PFC1) is set up to determine, on the basis of at least one of i) a plurality of previously detected temperature jumps and ii) a detected number of previously detected temperature jumps and preferably on the basis of a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or a predetermined time (tEND), and to influence the execution of the freeze-drying process depending on this determination or to propose an influence on the execution of the freeze-drying process to an operator of the freeze-dryer for release.

[0291] C27. Control system for a freeze dryer (LYO1) according to one of the items C24 to C26, characterized in that the process sequence control system (PFC1) is set up to assign a respective measurement time (t1; t8) to the temperature jumps recorded for a respective one of the sample vials and, after recording a defined rate or number of temperature jumps at the measurement times recorded in this way, to calculate at least one inferential statistical hypothesis test in order to determine, on the basis of the / a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or the predetermined time (tEND), and, depending on this determination, to influence the execution of the freeze-drying process or to propose an influence on the execution of the freeze-drying process to an operator of the freeze-dryer for approval.

[0292] C28. Control system for a freeze dryer (LYO1) according to item C27, characterized in that the process sequence control system (PFC1) is configured to calculate the mean value and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution over the measuring times (t1; ...; t8) and to supply it to the hypothesis test.

[0293] C29. Control system for a freeze dryer (LYO1) according to item C27 or C28, characterized in that the process flow control system (PFC1) is designed to assign the sample vials to different groups as part of the statistical evaluation depending on the positioning of the sample vials in the freeze dryer on the shelves or between the shelves, if desired using a factor analysis for statistical dimension reduction.

[0294] C30. Control for a freeze dryer (LYO1) according to item C29, characterized in that the process flow control (PFC1) is designed to distribute the mean or the variance or the standard deviation or at least one other statistical distribution measure adapted to the distribution among the groups, wherein the process flow control (PFC1) is further designed to statistically weight the groups depending on the positioning of the measured sample vials contained therein on the storage area and the number of other, unmeasured vials of the total quantity of vials with a technically comparable positioning.

[0295] C31 . Control system for a freeze dryer (LYO1 ) according to one of the items C27 to C30, characterized in that the hypothesis test is a Student's t-test or a test against the standard normal distribution or a variance test - ANOVA - or a non-parametric test.

[0296] C32. Control system for a freeze dryer (LYO1) according to one of the items C26 to C31, characterized in that the process sequence control system (PFC1) is further configured, in the case of a negative hypothesis being determined according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), to repeat the determination as time progresses as to whether an end of the primary drying can be assumed in the sense of a hypothesis at a later, then current time or the predetermined time (tEND).

[0297] C33. Control system for a freeze dryer (LYO1) according to one of the items C26 to C32, characterized in that the process sequence control system (PFC1) is further configured, in the case of a negative hypothesis determined according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), to numerically calculate in advance the time (tEND) of the assumed end of the primary drying by means of an iteration or a mathematical inversion.

[0298] C34. Control system for a freeze dryer (LYO1) according to one of the items C26 to C33, characterized in that the process sequence control system (PFC1) is further configured, in the case of a positive hypothesis being determined according to which an end of the primary drying can be assumed at such a current time or the predetermined time (tEND), to shorten or terminate the primary drying from this time onwards and to proceed to subsequent process steps of the freeze-drying process, or to suggest such shortening or termination of the freeze-drying to the operator for a transition to the subsequent process steps for manual release.

[0299] C35. Control system for a freeze dryer (LYO1) according to one of the items C1 to C34, characterized in that the process sequence control system (PFC1) is configured to first monitor the attainment of a necessary freezing temperature of the vials on the basis of the sample vials, if desired also by means of a hypothesis test for the hypothesis of the attainment of this freezing temperature, and is further configured to additionally lower the shelf temperature in the event of an assumed non-attainment of the freezing temperature.

[0300] C36. Control system for a freeze dryer (LYO1) according to one of the items C1 to C35, characterized in that the temperature measuring system is or comprises a wireless temperature measuring system (RF1).

[0301] C37. Control system for a freeze dryer (LYO1) according to item C36, characterized in that the wireless temperature measuring system (RF1) is designed to supply associated wireless temperature detection transponders, which are inserted or can be inserted into the or a respective one of the sample vials, with electrical operating energy from a radio frequency signal generated by the wireless temperature measuring system (RF1) and to receive radio frequency signals from such temperature detection transponders and to decode them in order to obtain temperature measurement values ​​transmitted thereby.

[0302] C38. Control system for a freeze dryer according to item C37, characterized in that the wireless temperature measuring system (RF1) is designed to unambiguously assign temperature measurement values ​​transmitted by a plurality of temperature detection transponders and then decoded to a respective one of the temperature detection transponders.

[0303] C39. Control according to item C37 or C38 for a freeze dryer, characterized in that the wireless temperature measuring system (RF1) is designed to modulate the high-frequency signal in such a way that it transmits an identification code which is intended to identify a temperature detection transponder to be interrogated with regard to the detected temperature and which is intended to wirelessly transmit a temperature measurement value.

[0304] C40. Control system according to item C39 for a freeze dryer, characterized in that the wireless temperature measuring system (RF1) is designed to receive radio frequency signals from such temperature detection transponders and to decode them in order to obtain temperature measurement values ​​transmitted thereby, together with identification codes identifying the respective temperature detection transponder from which such temperature measurement value originates.

[0305] C41. A control system for a freeze dryer according to item C37 or C38, characterized in that the wireless temperature measuring system (RF1) is designed to generate radio-frequency signals of different frequencies in order to address, via the frequency, one of several temperature detection transponders with different resonant frequencies, which are intended to wirelessly transmit a temperature measurement value. C42. A control system for a freeze dryer (LYO1) according to one of items C37 to C41, characterized in that the wireless temperature measuring system (RF1) is designed to receive and decode a backscatter amplitude modulation signal transmitted by the respective temperature detection transponder in order to obtain at least the transmitted temperature measurement value.

[0306] C43. Control system for a freeze dryer (LYO1) according to one of the items C1 to C42, characterized in that the process sequence control system (PFC1) is configured to adapt a temperature profile to be used in a secondary drying phase following the primary drying phase in accordance with the previous course of the primary drying phase or to offer it to an operator in an adapted and selectable manner for release.

[0307] C44. Freeze dryer, characterized in that a control system according to one of the items C1 to C43 is used to control the temperature of the vial storage areas.

[0308] C45. Freeze dryer according to item C44, characterized in that the temperature profile can be individually influenced by the control system for each shelf or in groups of shelf.

[0309] C46. Method for controlling a freeze-dryer (LYO1) which is designed to subject a total number of vessels (V1 to V9) referred to as vials with contents to be dried to a freeze-drying process and comprises at least one temperature measuring system which is designed to record a current temperature in at least one sample vial from the total number of vials (V1 to V9), and at least one process sequence controller (PFC1) which is designed to control the temperature of at least one storage area of ​​the vials in the freeze-dryer at least in a primary drying phase according to a predetermined temperature profile over time;characterized by a) detecting a characteristic temperature rise in the sample vial reflecting a progress of the primary drying, b) carrying out the further freeze-drying process in a defined manner depending on the detected characteristic temperature rise or a plurality of characteristic temperature rises detected for several separate sample vials;

[0310] C47. Method for controlling a freeze dryer (LYO1) according to item C46, ​​characterized by:

[0311] Determine, on the basis of the recorded characteristic temperature rise or the plurality of characteristic temperature rises recorded for several separate sample vials, whether and in what way the execution of the freeze-drying process is to be influenced or an operator of the freeze-dryer is to be suggested to influence the execution of the freeze-drying process for release.

[0312] C48. Method for controlling a freeze dryer (LYO1) according to item C46 or C47, characterized by:

[0313] Controlling at least one temperature setting device of the freeze dryer depending on the detected characteristic temperature increase or the plurality of characteristic temperature increases detected for several separate sample vials.

[0314] C49. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C48, characterized in that the characteristic temperature rise is a temperature jump or comprises a temperature jump which occurs as a result of completed primary drying, when the largely completed sublimation essentially no longer extracts any sublimation heat from the contents of the sample vial to be dried, possibly referred to as the substrate. C50. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C49, characterized in that the characteristic temperature rise is the temperature jump with a temporally subsequent rising temperature profile or comprises the temperature jump with a temporally subsequent rising temperature profile.

[0315] C51 . Method for controlling a freeze dryer (LYO1 ) according to one of the items C46 to C50, characterized by:

[0316] Recording and evaluating the characteristic temperature rise or a / the temperature jump characterizing this with a substantially negative-exponential temperature curve in the sample vial from the temperature jump against a temperature limit value in order to estimate the temperature limit value from a non-linearity of a / this temperature curve.

[0317] C52. Method for controlling a freeze dryer (LYO1) according to item C51, characterized by: on the basis of a plurality of assumed different provisional temperature limit values, respective modeling of the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump as a linearized logarithmic temperature course.

[0318] C53. Method for controlling a freeze dryer (LYO1) according to item C52, characterized in that the respective modeling comprises an adaptation of temperature-time measured value pairs representing the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump to a model straight line equation describing the respective linearized logarithmic temperature course.

[0319] C54. Method for controlling a freeze dryer (LYO1) according to item C53, characterized in that the model straight line equations obtained for the various provisional temperature limit values ​​are compared with each other and / or with the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value.

[0320] C55. Method for controlling a freeze dryer (LYO1) according to one of the items C52 to C54, characterized by:

[0321] Selecting, iteratively and / or stochastically, different provisional temperature limit values ​​in order to model the essentially negative-exponential temperature curve in the sample vial from the recorded temperature jump as a linearized logarithmic temperature curve.

[0322] C56. Method for controlling a freeze dryer (LYO1) according to one of the items C52 to C55, characterized in that different provisional temperature limit values ​​are selected within at least partially sequentially carried out modeling sequences and / or within at least partially parallelly carried out modeling sequences in order to model the essentially negative-exponential course of the temperature in the sample vial from the detected temperature jump as a linearized logarithmic temperature course.

[0323] C57. Method for controlling a freeze dryer (LYO1) for one of the objects C52 to C56, characterized in that model straight line equations are obtained for the various provisional temperature limit values ​​and these are subjected to a quality assessment by determining in each case a summary quality measure which evaluates a linear temperature-time curve corresponding to the respective model straight line equation on the basis of the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value on the basis of this quality assessment.

[0324] C58. Method for controlling a freeze-dryer (LYO1) according to one of the items C51 to C57, characterized by: changing at least one instantaneous shelf temperature depending on the temperature limit value, wherein this change is preferably based on an interpretation of the temperature limit value as an approximate shelf temperature.

[0325] C59. Method for controlling a freeze dryer (LYO1) according to one of the items C51 to C58, characterized by:

[0326] Determining a temperature difference between the current temperature recorded for the sample vial and the estimated temperature limit.

[0327] C60. Method for controlling a freeze dryer (LYO1) according to item C59, characterized by:

[0328] Subjecting this temperature difference to an undershoot test with respect to a predetermined minimum difference and, in the event of a detected undershoot of the minimum difference, raising an instantaneous shelf temperature to increase the temperature difference and thus cause a greater heat flow into the sample vial.

[0329] C61 . Method for controlling a freeze dryer (LYO1 ) according to item C60, characterized by:

[0330] Determining, in the case of multiple sample vials, a combined temperature difference, and

[0331] Subjecting the summarized temperature difference to the undershoot test in order to raise the current shelf temperature in the event that the minimum difference is detected to be undershot.

[0332] C62. Method for controlling a freeze dryer (LYO1) according to item C61, characterized in that the combined temperature difference is determined from the instantaneous temperatures recorded for the plurality of sample vials and estimated temperature limit values ​​using at least one statistical method such as averaging or hypothesis testing.

[0333] C63. Method for controlling a freeze dryer (LYO1) according to one of the items C60 to C62, characterized by:

[0334] Calculating, over measurement times of temperature difference undershoots, a mean and a variance or standard deviation or at least one other statistical distribution measure adapted to the distribution, and

[0335] Applying a hypothesis test to the variance or standard deviation or at least one other statistical distribution measure fitted to the distribution.

[0336] C64. Method for controlling a freeze dryer (LYO1) according to one of the items C60 to C63, characterized by:

[0337] - after raising the shelf temperature, re-determining a temperature difference to an estimated temperature limit value, which may be adjusted if desired according to the increase in the shelf temperature, and

[0338] - Subject this temperature difference to the undershoot test with regard to the specified minimum difference in order to raise the shelf temperature again if the minimum difference is again undershot.

[0339] C65. Method for controlling a freeze dryer (LYO1) according to one of the items C60 to C64, characterized in that the temperature difference to an estimated temperature limit is controlled, if desired PID-controlled, after the shelf temperature has been raised, comprising a renewed determination of the temperature difference to an estimated temperature limit, if desired adapted according to the control.

[0340] C66. Method for controlling a freeze dryer (LYO1) according to one of the items C59 to C65, characterized by: estimating, from the temperature difference or the temperature differences, a residual moisture in at least the sample vial or at least one of the sample vials.

[0341] C67. Method for controlling a freeze dryer (LYO1) according to one of the items C51 to C66, characterized by:

[0342] - Determination of a quality measure for the estimated temperature limit value used in further process control.

[0343] C68. Method for controlling a freeze dryer (LYO1) according to one of the items C51 to C67, characterized by:

[0344] - Evaluate the estimated temperature limit by calculating the difference between this and the measured temperature over time, logarithmizing this difference and then performing a linear regression against the logarithmized values ​​over time and determining a quality measure for their linearity.

[0345] C69. Method for controlling a freeze dryer (LYO1) according to one of the items C51 to C68, characterized by:

[0346] - Determining the estimated temperature limit value used in further process control by determining a quality measure for iteratively changed estimated temperature limits or from several assumed estimated temperature limits and using this as the basis for further process control an estimated temperature limit value with an improved quality measure.

[0347] C70. Method for controlling a freeze dryer (LYO1) according to one of the items C51 to C69, characterized by:

[0348] - Determining the estimated temperature limit by evaluating this quality measure based on at least one assumed estimated temperature limit and then iteratively changing the assumed estimated temperature limit so that the quality measure improves, preferably using established numerical iteration methods such as a gradient search or an evolutionary algorithm in the case of further parameters to be optimized.

[0349] C71 . Method for controlling a freeze dryer (LYO1 ) according to one of the items C51 to C70, characterized by:

[0350] - Determination of the estimated temperature limit value, starting from several statically assumed estimated temperature limit values, each of these is evaluated and from this set the one with the best evaluation is selected according to a quality measure in order to use it as a basis for further process control or to use it as a starting value for a subsequent iterative numerical determination of the estimated temperature limit value on which further process control is to be based.

[0351] C72. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C71, characterized by:

[0352] Recording a respective current temperature in several sample vials from the total number of vials (V1 to V9).

[0353] C73. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C72, characterized by: a) during the primary drying phase, detecting characteristic temperature increases reflecting progress in the primary drying in the form of temperature jumps in the sample vials, and b) carrying out the further freeze-drying process in a defined manner depending on the detected temperature jumps.

[0354] C74. Method for controlling a freeze dryer (LYO1) according to item C73, characterized by:

[0355] Determine, based on the detected temperature jumps, whether intervention in the execution of the freeze-drying process is necessary or whether intervention in the execution of the freeze-drying process should be suggested to a freeze-dryer operator for approval.

[0356] C75. Method for controlling a freeze dryer (LYO1) according to item C73 or C74, characterized by:

[0357] Determining, on the basis of at least one of i) a plurality of previously recorded temperature jumps and ii) a recorded number of previously recorded temperature jumps and preferably on the basis of a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or a predetermined time (tEND), and depending on this determination, influencing the implementation of the freeze-drying process or suggestions from an operator of the freeze-dryer to influence the implementation of the freeze-drying process for release.

[0358] C76. Method for controlling a freeze dryer (LYO1) according to one of the items C73 to C75, characterized by:

[0359] Assigning a respective measurement time (t1; ...; t8) to the temperature jumps recorded for each of the sample vials, and, after recording a specified rate or number of temperature jumps, calculating at least one inferential statistical hypothesis test on the measurement times thus recorded in order to determine, based on the / a predetermined maximum probability of error, whether an end of the primary drying can be assumed at the current time or the predetermined time (tEND) in the sense of a hypothesis, and depending on this determination, influencing the execution of the freeze-drying process or making suggestions to an operator of the freeze-dryer about influencing the execution of the freeze-drying process for release. C77. Method for controlling a freeze-dryer (LYO1) according to item C76, characterized by:

[0360] - Calculating, over the measurement times (t1 ; ; t8), the mean and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution,

[0361] - Applying the variance or standard deviation or at least one other statistical distribution measure fitted to the distribution to the hypothesis test.

[0362] C78. Method for controlling a freeze dryer (LYO1) according to item C76 or C77, characterized in that, depending on the positioning of the sample vials in the freeze dryer on the shelves or between the shelves, these are assigned to different groups within the framework of the statistical evaluation, if desired using a factor analysis for statistical dimension reduction.

[0363] C79. Method for controlling a freeze dryer (LYO1) according to item C78, ​​characterized in that the mean or the variance or the standard deviation or the at least one other statistical distribution measure adapted to the distribution is distributed among the groups, and in that the groups are statistically weighted depending on the positioning of the measured sample vials contained therein on the storage area and the number of other, unmeasured vials of the total quantity of vials with a technically comparable positioning.

[0364] C80. Method for controlling a freeze-dryer (LYO1) according to one of the items C76 to C79, characterized in that the hypothesis test is a Student's t-test or a test against the standard normal distribution or a variance test - ANOVA - or a non-parametric test.

[0365] C81 . Method for controlling a freeze dryer (LYO1 ) according to one of the items C75 to C80, characterized in that in the case of a determined negative hypothesis according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), the determination is repeated as time progresses to determine whether an end of the primary drying can be assumed in the sense of a hypothesis at a later, then current time or the predetermined time (tEND).

[0366] C82. Method for controlling a freeze dryer (LYO1) according to one of the items C75 to C81, characterized in that in the case of a negative hypothesis determined, according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), the time (tEND) of the assumed end of the primary drying is calculated numerically in advance by an iteration or a mathematical inversion.

[0367] C83. Method for controlling a freeze dryer (LYO1) according to one of the items C75 to C82, characterized in that in the case of a positive hypothesis determined, according to which an end of the primary drying can be assumed at such a current time or the predetermined time (tEND), the primary drying is abbreviated or terminated from this time and a transition is made to subsequent process steps of the freeze-drying process, or such an abbreviation or termination of the freeze-drying is proposed to the operator for manual release for a transition to the subsequent process steps.

[0368] C84. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C83, characterized in that firstly the attainment of a necessary freezing temperature of the vials is monitored on the basis of the sample vials, if desired also by means of a hypothesis test for the hypothesis of the attainment of this freezing temperature, and that furthermore in the event of an assumed non-attainment of the freezing temperature the storage surface temperature is additionally lowered.

[0369] C85. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C84, characterized in that the instantaneous temperature in the sample vial or vials is detected wirelessly.

[0370] C86. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C85, characterized in that a temperature profile to be used in a secondary drying phase following the primary drying phase is adapted according to the previous course of the primary drying phase or is offered to an operator in an adapted and selectable manner for approval.

[0371] C87. Method for controlling a freeze dryer (LYO1) according to one of the items C46 to C86, in that a temperature control is carried out for the storage areas of the vials, wherein, if desired, the temperature profile is influenced for each storage plate or in groups of storage plates.

[0372] C88. Control system for a freeze dryer (LYO1), which is designed to carry out a control and / or regulation of at least one freeze-drying process according to the method according to one of the items C46 to C87.

[0373] C89. Freeze dryer equipped with a control system (LYO1) according to any one of items C1 to C45 or according to item C88.

[0374] A method for controlling a freeze dryer (LYO1) which is designed to subject a total number of vessels (V1 to V9) referred to as vials with contents to be dried to a freeze-drying process and comprises at least one temperature measuring system which is designed to detect a current temperature in at least one sample vial from the total number of vials (V1 to V9), as well as at least one process sequence controller (PFC1) which is designed to control the temperature of at least one storage area of ​​the vials in the freeze dryer at least in a primary drying phase according to a predetermined temperature profile over time, is characterized by a) detecting a characteristic temperature increase in the sample vial reflecting a progress of the primary drying,b) Carrying out the further freeze-drying process in a defined manner depending on the recorded characteristic temperature increase or a plurality of characteristic temperature increases recorded for several separate sample vials.

[0375] According to a first aspect of the invention, an individual shelf temperature for the (respective) sample vial is estimated from the detected characteristic temperature rise(s). According to a second aspect, an end point of the primary drying phase is detected or predicted from the characteristic temperature rises of several sample vials. The invention also provides a corresponding control system for a freeze dryer and a corresponding freeze dryer.

Claims

Claims 1. A control system for a freeze dryer (LYO1) designed to subject a total number of vessels (V1 to V9) referred to as vials with contents to be dried to a freeze-drying process, comprising at least one temperature measuring system designed to detect a current temperature in at least one sample vial from the total number of vials (V1 to V9), and at least one process sequence controller (PFC1) designed to control the temperature of at least one storage area of the vials in the freeze dryer, at least in a primary drying phase, according to a predetermined temperature profile over time;characterized in that the process flow controller (PFC1) is further configured to a) detect a characteristic temperature rise in the sample vial during the primary drying, reflecting a progress of the primary drying, and b) carry out the further freeze-drying process in a defined manner depending on the detected characteristic temperature rise or a plurality of characteristic temperature rises detected for several separate sample vials.; 2. Control for a freeze dryer (LYO1) according to claim 1, characterized in that the process sequence control (PFC1) is set up to determine for the further implementation of the freeze-drying process on the basis of the detected characteristic temperature increase or the plurality of characteristic temperature increases detected for a plurality of separate sample vials whether and in what way the implementation of the freeze-drying process is to be influenced or an operator of the freeze-dryer is to be suggested an influence on the implementation of the freeze-drying process for release.

3. Control for a freeze dryer (LYO1) according to claim 1 or 2, characterized in that the process flow control (PFC1) is further configured to control at least one temperature setting device of the freeze dryer depending on the detected characteristic temperature rise or the plurality of characteristic temperature rises detected for several separate sample vials.

4. Control for a freeze dryer (LYO1) according to one of claims 1 to 3, characterized in that the characteristic temperature rise is a temperature jump or comprises a temperature jump which occurs as a result of a completed primary drying when the largely completed sublimation essentially no longer extracts any sublimation heat from the contents of the sample vial to be dried, possibly referred to as substrate.

5. Control for a freeze dryer (LYO1) according to one of claims 1 to 4, characterized in that the characteristic temperature rise is the temperature jump with a temporally subsequent rising temperature profile or comprises the temperature jump with a temporally subsequent rising temperature profile.

6. Control for a freeze dryer (LYO1) according to one of claims 1 to 5, characterized in that the process sequence control (PFC1) is further configured to record and evaluate the characteristic temperature rise or a / the temperature jump characterizing this with a substantially negative-exponential course of the temperature in the sample vial from the temperature jump towards a temperature limit value in order to estimate the temperature limit value from a non-linearity of a / this course of the temperature.

7. Control system for a freeze dryer (LYO1) according to claim 6, characterized in that the process sequence control (PFC1) is further configured to determine, on the basis of a plurality of assumed different provisional temperature limit values, the substantially to model the negative-exponential course of the temperature in the sample vial from the recorded temperature jump as a linearized logarithmic temperature curve, in that the process sequence control (PFC1) adapts the temperature-time measured value pairs representing the essentially negative-exponential course of the temperature in the sample vial from the recorded temperature jump to a model straight line equation describing the respective linearized logarithmic temperature curve, and in that the process sequence control (PFC1) is further set up to compare the model straight line equations obtained for the various provisional temperature limit values with one another and / or with the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value.

8. Control according to claim 6 or 7, characterized in that the process sequence control (PFC1) is set up to subject the model straight line equations obtained for the various provisional temperature limit values to a quality assessment by determining a summary quality measure which evaluates a linear temperature-time curve corresponding to the respective model straight line equation on the basis of the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value on the basis of this quality assessment.

9. Control for a freeze dryer (LYO1) according to one of claims 6 to 8, characterized in that the process sequence control (PFC1) is set up to change a current shelf temperature depending on the temperature limit value by means of the / a temperature setting device of the freeze dryer, which is preferably assigned to at least one shelf of the freeze dryer, wherein this change is preferably based on an interpretation of the temperature limit value as an approximate shelf temperature.

10. Control system for a freeze dryer (LYO1) according to one of claims 6 to 9, characterized in that the process flow control (PFC1) is set up to determine a temperature difference between the current temperature recorded for the sample vial and the estimated temperature limit value and to subject this temperature difference to an undershoot test with respect to a predetermined minimum difference, wherein the process sequence control (PFC1) is further set up to raise a current shelf temperature in the event of a detected undershoot of the minimum difference in order to increase the temperature difference and thus cause a greater heat flow into the sample vial.

11. Control system for a freeze dryer according to claim 10, characterized in that the process sequence control (PFC1) is designed to provide a control, if desired PID control with a limited control range, of the temperature difference to an estimated temperature limit value after the shelf temperature has been raised, which comprises a renewed determination of the temperature difference to an estimated temperature limit value, if desired adjusted according to the control.

12. Control system for a freeze dryer according to claim 10 or 11, characterized in that the process control system (PFC1) is configured to estimate a residual moisture content in at least the sample vial or at least one of the sample vials from the temperature difference or the temperature differences.

13. Control system for a freeze dryer (LYO1) according to one of claims 6 to 12, characterized in that the process sequence control system (PFC1) is designed to determine a quality measure for the estimated temperature limit value which is included in the further process control system.

14. Control system for a freeze dryer (LYO1) according to one of claims 6 to 13, characterized in that the process control system (PFC1) is arranged to carry out an evaluation of the estimated temperature limit value by determining the difference between this and measured temperature is calculated over time, this is logarithmized and then a linear regression is carried out against the time course using the logarithmized values and a quality measure for their linearity is determined using this.

15. Control system for a freeze dryer (LYO1) according to one of claims 6 to 14, characterized in that the process sequence control system (PFC1) is designed to determine the estimated temperature limit value used in the further process control by determining a quality measure for iteratively changed estimated temperature limit values or from a plurality of assumed estimated temperature limit values, and from this an estimated temperature limit value with an improved quality measure is used as the basis for the further process control.

16. Control for a freeze dryer (LYO1) according to one of claims 6 to 15, characterized in that the process sequence control (PFC1) is set up to determine the estimated temperature limit by evaluating this quality measure based on at least one assumed estimated temperature limit and then iteratively changing the assumed estimated temperature limit such that the quality measure is improved, preferably using established numerical iteration methods such as a gradient search or an evolutionary algorithm in the case of further parameters to be optimized.

17. Control system for a freeze dryer (LYO1) according to one of claims 6 to 16, characterized in that the process sequence control (PFC1) is designed to determine the estimated temperature limit value by evaluating each of a plurality of statically assumed estimated temperature limit values and selecting from this set the one with the best evaluation according to a quality measure in order to use it as a basis for further process control or as a starting value for a subsequent iterative numerical determination of the to use the estimated temperature limit value to be used as the basis for further process control.

18. Control system for a freeze dryer (LYO1) according to one of claims 1 to 17, characterized in that the temperature measuring system is designed to record a respective instantaneous temperature in several sample vials from the total number of vials (V1 to V9), and in that the process sequence control (PFC1) is set up to a) record, during the primary drying phase, characteristic temperature increases reflecting progress in the primary drying in the form of temperature jumps in the sample vials, and b) carry out the further freeze-drying process in a defined manner depending on the recorded temperature jumps.

19. Control system for a freeze dryer (LYO1) according to claim 18, characterized in that the process sequence control system (PFC1) is configured to determine, for the further execution of the freeze-drying process, on the basis of the detected temperature jumps, whether the execution of the freeze-drying process is to be influenced or whether an operator of the freeze-dryer is to be suggested an influence on the execution of the freeze-drying process for release.

20. Control for a freeze dryer (LYO1) according to claim 18 or 19, characterized in that the process sequence control (PFC1) is set up to determine, on the basis of at least one of i) a plurality of previously detected temperature jumps and ii) a detected number of previously detected temperature jumps and preferably on the basis of a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or a predetermined time (tEND), and to influence the execution of the freeze-drying process depending on this determination or to propose an influence on the execution of the freeze-drying process to an operator of the freeze-dryer for release.

21. Control for a freeze dryer (LYO1) according to one of claims 18 to 20, characterized in that the process sequence control (PFC1) is set up to assign a respective measuring time (t1; t8) to the temperature jumps recorded for a respective one of the sample vials and, after recording a predetermined rate or number of temperature jumps at the measuring times recorded in this way, to calculate at least one inferential statistical hypothesis test in order to determine, on the basis of the / a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or the predetermined time (tEND), and, depending on this determination, to influence the execution of the freeze-drying process or to propose to an operator of the freeze-dryer an influence on the execution of the freeze-drying process for release.

22. Control for a freeze dryer (LYO1) according to claim 21, characterized in that the process sequence control (PFC1) is set up to calculate the mean value and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution over the measuring times (t1; ...; t8) and to feed it to the hypothesis test.

23. Control system for a freeze dryer (LYO1) according to claim 21 or 22, characterized in that the process sequence control system (PFC1) is designed to assign the sample vials to different groups as part of the statistical evaluation depending on the positioning of the sample vials in the freeze dryer on the shelves or between the shelves, if desired using a factor analysis for statistical dimension reduction.

24. Control system for a freeze dryer (LYO1) according to claim 23, characterized in that the process control system (PFC1) is designed to calculate the mean value or the variance or the standard deviation or to distribute at least one other statistical distribution measure adapted to the distribution to the groups, wherein the process flow control (PFC1) is further configured to statistically weight the groups depending on the positioning of the measured sample vials contained therein on the storage area and the number of other, unmeasured vials of the total quantity of vials with a technically comparable positioning.

25. Control system for a freeze dryer (LYO1) according to one of claims 20 to 24, characterized in that the process sequence control system (PFC1) is further configured, in the case of a negative hypothesis being determined according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), to repeat the determination as time progresses as to whether an end of the primary drying can be assumed in the sense of a hypothesis at a later, then current time or the predetermined time (tEND).

26. Control for a freeze dryer (LYO1) according to one of claims 20 to 25, characterized in that the process sequence control (PFC1) is further configured, in the case of a negative hypothesis determined according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), to numerically calculate in advance the time (tEND) of the assumed end of the primary drying by means of an iteration or a mathematical inversion.

27. Control system for a freeze dryer (LYO1) according to one of claims 20 to 26, characterized in that the process sequence control system (PFC1) is further configured, in the case of a positive hypothesis being determined according to which an end of the primary drying can be assumed at such a current time or the predetermined time (END), to shorten or terminate the primary drying from this time onwards and to proceed to subsequent process steps of the freeze-drying process, or to indicate to the operator such shortening or termination of the To propose freeze-drying for a transition to the subsequent process steps for manual release.

28. Control system for a freeze dryer (LYO1) according to one of claims 1 to 27, characterized in that the process sequence control system (PFC1) is set up to first monitor the attainment of a necessary freezing temperature of the vials on the basis of the sample vials, if desired also by means of a hypothesis test for the hypothesis of the attainment of this freezing temperature, and is further set up to additionally lower the shelf temperature in the event of an assumed non-attainment of the freezing temperature.

29. Control system for a freeze dryer (LYO1) according to one of claims 1 to 28, characterized in that the process sequence control system (PFC1) is designed to adapt a temperature profile to be used in a secondary drying phase following the primary drying phase in accordance with the previous course of the primary drying phase or to offer it to an operator in an adapted and selectable manner for release.

30. Method for controlling a freeze dryer (LYO1) which is designed to subject a total quantity of vessels (V1 to V9) referred to as vials with contents to be dried to a freeze-drying process and comprises at least one temperature measuring system which is designed to detect a current temperature in at least one sample vial from the total quantity of vials (V1 to V9), and at least one process sequence controller (PFC1) which is designed to control the temperature of at least one storage area of the vials in the freeze dryer at least in a primary drying phase according to a predetermined temperature profile over time; characterized by a) detecting a characteristic temperature increase in the sample vial reflecting a progress of the primary drying, b) carrying out the further freeze-drying process in a defined manner depending on the detected characteristic Temperature rise or a plurality of characteristic temperature rises recorded for several separate sample vials.

31. Method for controlling a freeze dryer (LYO1) according to claim 30, characterized by: Determine, on the basis of the recorded characteristic temperature rise or the plurality of characteristic temperature rises recorded for several separate sample vials, whether and in what way the execution of the freeze-drying process is to be influenced or an operator of the freeze-dryer is to be suggested to influence the execution of the freeze-drying process for release.

32. Method for controlling a freeze dryer (LYO1) according to claim 30 or 31, characterized by: Controlling at least one temperature setting device of the freeze dryer depending on the detected characteristic temperature rise or the plurality of characteristic temperature rises detected for several separate sample vials.

33. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 32, characterized in that the characteristic temperature rise is a temperature jump or comprises a temperature jump which occurs as a result of a completed primary drying when the largely completed sublimation essentially no longer extracts any sublimation heat from the contents of the sample vial to be dried, possibly referred to as the substrate.

34. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 33, characterized in that the characteristic temperature rise is the temperature jump with a temporally subsequent rising temperature profile or comprises the temperature jump with a temporally subsequent rising temperature profile.

35. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 34, characterized by: Recording and evaluating the characteristic temperature rise or a / the temperature jump characterizing this with a substantially negative-exponential temperature curve in the sample vial from the temperature jump against a temperature limit value in order to estimate the temperature limit value from a non-linearity of a / this temperature curve.

36. Method for controlling a freeze dryer (LYO1) according to claim 35, characterized by: on the basis of a plurality of assumed different provisional temperature limit values, respective modeling of the substantially negative-exponential course of the temperature in the sample vial from the detected temperature jump as a linearized logarithmic temperature course.

37. Method for controlling a freeze dryer (LYO1) according to claim 36, characterized in that the respective modeling comprises an adaptation of temperature-time measured value pairs representing the substantially negative-exponential course of the temperature in the sample vial from the detected temperature jump to a model straight line equation describing the respective linearized logarithmic temperature course.

38. A method for controlling a freeze dryer (LYO1) according to claim 37, characterized in that the model straight line equations obtained for the various provisional temperature limit values are compared with each other and / or with the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value.

39. Method for controlling a freeze dryer (LYO1) according to one of claims 36 to 38, characterized in that model straight line equations are obtained for the various provisional temperature limit values and these are subjected to a quality assessment by determining in each case a summary quality measure which evaluates a linear temperature-time curve corresponding to the respective model straight line equation on the basis of the temperature-time measured value pairs in order to estimate the best-fitting provisional temperature limit value as the temperature limit value on the basis of this quality assessment.

40. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 39, characterized by: Changing at least one current shelf temperature depending on the temperature limit value, wherein this change is preferably based on an interpretation of the temperature limit value as an approximate shelf temperature.

41. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 40, characterized by: Determining a temperature difference between the current temperature recorded for the sample vial and the estimated temperature limit.

42. Method for controlling a freeze dryer (LYO1) according to claim 41, characterized by: Subjecting this temperature difference to an undershoot test with respect to a predetermined minimum difference and, in the event of a detected undershoot of the minimum difference, raising an instantaneous shelf temperature to increase the temperature difference and thus cause a greater heat flow into the sample vial.

43. Method for controlling a freeze dryer (LYO1) according to claim 42, characterized in that the temperature difference to a estimated temperature limit after raising the shelf temperature, if desired PID-controlled, comprising a renewed determination of the temperature difference to an estimated temperature limit, if desired adjusted according to the control.

44. Method for controlling a freeze dryer (LYO1) according to one of claims 41 to 43, characterized by: - Estimating, from the temperature difference or temperature differences, a residual moisture content in at least the sample vial or at least one of the sample vials.

45. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 44, characterized by: - Determination of a quality measure for the estimated temperature limit value used in further process control.

46. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 45, characterized by: - Evaluate the estimated temperature limit by calculating the difference between this and the measured temperature over time, logarithmizing this difference and then performing a linear regression against the logarithmized values over time and determining a quality measure for their linearity.

47. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 46, characterized by: - Determination of the estimated temperature limit value used in further process control by determining a quality measure for iteratively changed estimated temperature limits or from several assumed estimated temperature limits and from this an estimated temperature limit value with a improved quality measure is used as the basis for further process control.

48. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 47, characterized by: - Determining the estimated temperature limit by evaluating this quality measure based on at least one assumed estimated temperature limit and then iteratively changing the assumed estimated temperature limit so that the quality measure improves, preferably using established numerical iteration methods such as a gradient search or an evolutionary algorithm in the case of further parameters to be optimized.

49. Method for controlling a freeze dryer (LYO1) according to one of claims 35 to 48, characterized by: - Determination of the estimated temperature limit value, starting from several statically assumed estimated temperature limit values, each of these is evaluated and from this set the one with the best evaluation is selected according to a quality measure in order to use it as a basis for further process control or to use it as a starting value for a subsequent iterative numerical determination of the estimated temperature limit value on which further process control is to be based.

50. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 49, characterized by: Recording a respective current temperature in several sample vials from the total number of vials (V1 to V9). 51 . Method for controlling a freeze dryer (LYO1 ) according to one of claims 30 to 50, characterized by: a) during the primary drying phase, detecting characteristic values reflecting progress in the primary drying Temperature increases in the form of temperature jumps in the sample vials, and b) carrying out the further freeze-drying process in a defined manner depending on the recorded temperature jumps.

52. Method for controlling a freeze dryer (LYO1) according to claim 51, characterized by: Determine, based on the detected temperature jumps, whether intervention in the execution of the freeze-drying process is necessary or whether intervention in the execution of the freeze-drying process should be suggested to a freeze-dryer operator for approval.

53. Method for controlling a freeze dryer (LYO1) according to claim 51 or 52, characterized by: Determining, on the basis of at least one of i) a plurality of previously recorded temperature jumps and ii) a recorded number of previously recorded temperature jumps and preferably on the basis of a predetermined maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or a predetermined time (tEND), and depending on this determination, influencing the implementation of the freeze-drying process or suggestions from an operator of the freeze-dryer to influence the implementation of the freeze-drying process for release.

54. Method for controlling a freeze dryer (LYO1) according to one of claims 51 to 53, characterized by: Assigning a respective measurement time (t1; ...; t8) to the temperature jumps recorded for each of the sample vials, and, after recording a fixed rate or number of temperature jumps, calculating at least one inferential statistical hypothesis test on the measurement times recorded in this way in order to To determine, on the basis of a given maximum probability of error, whether an end of the primary drying can be assumed in the sense of a hypothesis at the current time or the given time (tEND), and depending on this determination, to influence the execution of the freeze-drying process or to make suggestions to an operator of the freeze-dryer to influence the execution of the freeze-drying process for release.

55. Method for controlling a freeze dryer (LYO1) according to claim 54, characterized by: - Calculating, over the measurement times (t1 ; ... ; t8), the mean and the variance or standard deviation or at least one other statistical distribution measure adapted to the distribution, - Applying the variance or standard deviation or at least one other statistical distribution measure fitted to the distribution to the hypothesis test.

56. Method for controlling a freeze dryer (LYO1) according to claim 54 or 55, characterized in that depending on the positioning of the sample vials in the freeze dryer on the shelves or between the shelves, these are assigned to different groups within the framework of the statistical evaluation, if desired using a factor analysis for statistical dimension reduction.

57. Method for controlling a freeze dryer (LYO1) according to claim 56, characterized in that the mean value or the variance or the standard deviation or the at least one other statistical distribution measure adapted to the distribution is distributed among the groups, and in that the groups are statistically weighted depending on the positioning of the measured sample vials contained in them on the storage area and the number of other, unmeasured vials of the total quantity of vials with a technically comparable positioning.

58. Method for controlling a freeze dryer (LYO1) according to one of claims 53 to 57, characterized in that in the case of a negative hypothesis determined, according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), the determination is repeated as time progresses as to whether an end of the primary drying can be assumed in the sense of a hypothesis at a later, then current time or the predetermined time (tEND).

59. Method for controlling a freeze dryer (LYO1) according to one of claims 53 to 58, characterized in that in the case of a negative hypothesis determined, according to which an end of the primary drying cannot yet be assumed at the current time or the predetermined time (tEND), the time (tEND) of the assumed end of the primary drying is calculated numerically in advance by an iteration or a mathematical inversion.

60. Method for controlling a freeze dryer (LYO1) according to one of claims 53 to 59, characterized in that in the case of a positive hypothesis being determined, according to which an end of the primary drying can be assumed at such a current time or the predetermined time (tEND), the primary drying is abbreviated or terminated from this time onwards and the system proceeds to subsequent process steps of the freeze-drying process, or such an abbreviation or termination of the freeze-drying is proposed to the operator for manual release in order to proceed to the subsequent process steps.

61. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 60, characterized in that firstly the attainment of a necessary freezing temperature of the vials is monitored using the sample vials, if desired also by means of a hypothesis test for the hypothesis of the attainment of this freezing temperature, and that further If the freezing temperature is assumed not to be reached, the shelf temperature is further reduced.

62. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 61, characterized in that a temperature profile to be used in a secondary drying phase following the primary drying phase is adapted according to the previous course of the primary drying phase or is offered to an operator in an adapted and selectable manner for release.

63. Method for controlling a freeze dryer (LYO1) according to one of claims 30 to 62, characterized in that a temperature control is carried out for the storage areas of the vials, wherein, if desired, the temperature profile is influenced for each storage plate or in groups of storage plates.

64. Control system for a freeze dryer (LYO1), which is designed to carry out a control and / or regulation of at least one freeze-drying process according to the method according to one of claims 30 to 63.

65. Freeze dryer equipped with a control (LYO1) according to one of claims 1 to 29 or according to claim 64.

Images