A survey of in-line and online laser diffraction analysis to improve process monitoring and control capabilities. For in-line analysis, the equipment is installed close to the product's drain. Samples for analysis are manually extracted from production when needed. In terms of process relevance, this approach has some limitations associated with offline analysis, although the response time is very short. At the same time, the sample is unlikely to be reused after analysis. If the material is expensive or the supply is short, this may become a topic worth exploring. Figure 2: Real-time particle size data measured at the beginning of spray drying At 10:57 am, the particle size suddenly increased sharply and decreased again during transmission. This detected incident is related to the process operation: The powder is released from the outer wall of the measuring container. In the production process, oil-in-water microspheres having a hard surface adhere to the walls of the container to form agglomerates. In order to prevent excessive material thickening, the powder is usually released using a crowbar beater wall or air enthalpy. This intervention causes the material to pass quickly through the measurement zone, thereby reducing transmission to measure how light passes through the sample quickly. The increase in particle size is due to the nature of the released material (partly due to aggregation). Figure 3: Identifying the chaos that occurred during a process Being able to detect the confusion of planned and unplanned processes and assess their impact is valuable. For example, for this process, the frequency of any operation taken to remove thickened material will affect the amount and size of particle release, making the product uniform. In an integrated plant, process chaos can negatively impact the operation of downstream units. The PAT solution provides continuous measurement and high data acquisition speed, and details the dynamics of the process, providing information for quality-derived design (QbD). Another alternative is to mathematically simulate the results and give the information you need. Using data analysis techniques, the measured particle size distribution can be divided into two modes to provide a reference for the measurement and aggregation of the isolated microspheres. If particles above a certain particle size are defined as aggregates, they can be excluded by correcting the particle size range.
The FDA's goal of promoting PAT and quality from design (QbD) is to encourage pharmaceutical companies to focus more on production efficiency. Traditional production relies heavily on off-line testing of online materials and end products. FDA seeks to modernize production practices by promoting knowledge-based development and design, using analytical tools that continuously monitor critical production processes. For some variables, long-term or continuous measurements in the operating environment are still a serious technical burden. But for others, there are already commercially viable solutions available. Particle size is such a variable.
For particle drug systems, particle size is an important parameter. The particle size determines the nature of the drug release and also affects other properties such as flow, stability and mouthfeel of the suspension. As a result, particle size is a routine measurement item in powder-related production processes, and laser diffraction is an alternative analytical method in many applications. Laser diffraction is a fast, reliable and reproducible technology that is equally suitable for offline, inline, online and online analysis. Therefore, it can be used for early development, pilot testing and full scale production.
The benefits of using a real-time laser diffraction particle size analyzer to monitor the grinding process include increased throughput, better product quality, and reduced waste. This technique is equally relevant and beneficial for conventional operations such as spray drying. In this study, a pilot scale spray dryer for producing oily microspheres was monitored and controlled using an in-line and on-line particle size analyzer.
Production of oil-in-water microspheres
Spray drying is widely used in the pharmaceutical industry, and micro-coating for solid or oily liquids is a valuable technology. The micro-coating not only controls the release rate of the active pharmaceutical ingredient, but also prevents oxidation and degradation of the core material because the outer coating forms a protective physical barrier. Micro-coating is also a useful technique for prescriptions that require masking. Nevertheless, to achieve the desired performance, the particle size of the microsphere product needs to be strictly controlled.
In this study, the goal was to produce oil-in-water microspheres with a particle size range of 10-35 um. In general, the data obtained by off-line analysis of samples collected from the production line on a regular basis are used to control the particle size of the microspheres. In this case, offline measurements involve light microscopy associated with image analysis. Such techniques are very useful for early research and development because they provide virtual images and valuable information about particle morphology and shape, and it is also an effective way to evaluate particle size.
Despite this, light microscopy and image analysis are not ideal for process monitoring and control. Because only a small number of samples are measured, the quality of the data is highly dependent on the extraction and analysis of representative samples. Equally important, the sampling and measurement process is too slow for real-time monitoring. The corresponding decision was to investigate the ability of in-line and online laser diffraction analyzers to improve process monitoring and control.
In-line and in-line particle size determination techniques for spray drying
Choose to apply Malvern's online particle size monitoring system. The online laser diffraction particle size analyzer integrates inline, online and online options specifically designed for the process area. These systems are suitable for wet and dry fluids with particle sizes ranging from 0.1 to 2000 microns. In this survey, the same equipment was used to measure online and inline measurements at different time points.
When installed on the output line of the spray dryer (Figure 1), the sample preparation is not required and the analyzer can measure all outlet flows, at which point the status is "in process". For some large-volume units, this may not be an option. In this case, online analysis (a portion of the flow is converted for measurement) is more suitable. The online system is fully closed, which reduces the risk of exposure to process materials. The analysis is completely automated, so the operator's input is small and reproducible. 
In both layouts, air cleaning prevents particles from adhering to the surface of the lens, maintaining measurement accuracy. Using an in-line system, an additional air supply (introduced via the venturi, the venturi is used to take the sample into the analyzer) provides a dispersion function prior to measurement, and careful design of the flow rate ensures that any aggregated particles break without causing microspheres Wear and tear.
Monitoring process
The real-time recorded data at the beginning of the spray dryer is shown in Figure 2. Due to lag time, it is almost impossible to effectively track transient operations, startup, shutdown cycles, or changes in product standards using off-line analysis. For this process, the initial phase is expected to take about 10 minutes, so in order to ensure the quality of the product, the output of the initial stage is discarded. The online results show that the actual time required to reach steady state is much faster than the above inference, only a few minutes. Here, the online device provides the data needed to make a precise decision to determine when to start sample collection and then cut off the waste to maximize production. 
Because the analyzer can track the production process in real time, the operator can detect the problem in time, observe the effects of the planned change, or monitor and correct any long time drift. Performing an inspection over a longer period of time can highlight the ability of the device identification process to be confusing (see Figure 3). 
Although the system clearly allows the operator to confirm steady state operation, or vice versa, the comparison of online and inline data reveals a significant difference between the two. On-line analysis showed a double particle size distribution peak, while the in-line results were narrower particle size distribution peaks with almost no secondary or larger particle size patterns. The average particle size in the line is much smaller than that of the line. These results are directly related to the state of the sample being measured. The on-line analyzer measures the sample without the need for sample preparation and is able to correctly detect the presence of aggregated particles in the effluent fluid. The oily nature of the microspheres and the smaller particle size make them more susceptible to aggregation. Therefore, the measurement of the particle size of the separated particles becomes more challenging. When this tendency to aggregate is more pronounced, the higher the oil load, the more pronounced the bimodal distribution.
The results in the line indicate that the gas stream used for sample dispersion can successfully break up most of the aggregated particles. This largely eliminates the second peak in the double peak. Therefore, the results in the line can more accurately reflect the particle size of the discrete microspheres. Nevertheless, data from the online can more accurately reflect the true nature of the export fluid. If the target is a dispersed microsphere, the product from the spray dryer will contain a large amount of aggregate material that requires further processing.
Real-time analysis of aggregated particles
These results increase the debate about whether particle size measurements need to be implemented to achieve good process control. Here an online analyzer is used to measure the material being produced and to show if the operation is in steady state. However, it cannot set the particle size of the discrete microspheres in the initial configuration. Is this important?
The answer depends on how the airflow at the outlet is used and how the parameters are related to product performance. If the online analyzer is able to determine the appropriate set point, it will be able to sensitively detect variables associated with this value. In this case, the relationship between the particle size of the discrete particles and the on-line results will depend on the tendency of the material to aggregate. As noted earlier, the degree of aggregation depends on the amount of oil carried but other factors also have an effect. The correlation between discrete particle size and on-line measurement is therefore not linearly related. For the exploration of pilot scale, it is necessary to provide a corresponding correction scheme.
One option is to integrate a decentralized system on the production line just before the measurement area (Figure 4). The correct setting of the airflow will bring the in-line and online data closer together. Still, it is clear that this will have an impact on the overall process flow. Alternative methods include periodic dispersion during the measurement process, or the sample loop with purge gas - this is effective online rather than online. The best solution is process dependencies, which are influenced by the scale of operations and the basic needs of the application. 
in conclusion
This experimental study confirmed the feasibility of laser diffraction particle size analysis for continuous monitoring of PAT operations in the spray drying process. This particular application is a requirement for the production of oily microspheres because the product itself tends to aggregate. There are different options for how the sample dispersion is incorporated into the in-line, on-line, and on-line configurations to ensure accurate measurement of the particle size of the discrete particles.
With continuous analysis, the operator can control the process in a highly responsive manner, as the problem and the results of any operation are obvious. Therefore, this reduces the start time from ten minutes to two minutes because the time point reached by the steady state can be accurately measured instead of being estimated. This reduces waste and increases production.
More generally, the use of appropriate PAT equipment provides detailed insight into the process, provides the basis for product optimization, and drives plant and operational practices. The current increasing use of PAT in the pilot scale allows developers to quickly and efficiently obtain the information required by QbD to help drive challenging pharmaceutical production transformations.