How do I measure and document mixing quality reproducibly for pharma validations?
A reproducible mixing-quality assessment for pharma validations rests on a pre-approved, risk-based sampling and analysis concept. What matters is not a low RSD value alone, but the suitability of the sampling, an analytical method validated for the powder mixture, control of critical process parameters, and GMP-compliant, traceable documentation.
Objective and acceptance
Mixing quality, or blend uniformity, describes the even distribution of the active ingredient and, where applicable, other critical components in a powder or granule mixture. It is a critical quality attribute, because inadequate homogeneity can lead to uneven active-ingredient content in tablets, capsules, sachets or other individual dosage units.
The coefficient of variation, or relative standard deviation, is calculated from the content values of the individual samples:
RSD=CV=sxˉ×100%
Here, s is the standard deviation and x̄ is the mean of the measured active-ingredient concentrations. A value of RSD≤5% is a commonly used, demanding target for blend uniformity, but it must not be treated across the board as a universal regulatory release limit. The acceptance criteria must be product-specific, scientifically justified, risk-based, and defined in the validation protocol before the study begins.
As a reference approach widely used in the field, blend samples are commonly assessed against an RSD below 5 percent together with individual values within 90.0 to 110.0 percent of the mean. For individual dosage units, an RSD below 6 percent and supplementary limits for individual content are frequently discussed. These values must always be checked against the approved product specification, the validated control concept and the applicable regulatory expectations.
Planning sampling
The greatest source of error is often not the mixer but the sampling. A sampling plan must therefore demonstrate that it can detect possible poorly mixed zones, segregation and differences over the course of discharge. Sampling should be stratified both spatially and temporally.
When examining the final mixture, samples are taken from deliberately selected positions. These include, for example, upper, middle and lower zones, areas near the wall, central areas, the discharge, and any areas that are constructionally or process-technically critical with an elevated risk of dead space or segregation. For convective mixers, a PQRI proposal recommends sampling that is as volumetrically uniform as possible, including corners and the discharge; at least 20 positions are named as a starting point. For other mixer geometries, at least ten locations are described as a starting point.
A robust minimum approach is at least ten spatially separated sample positions. As a rule, the validation concept should provide for replicates from each position, for example three individual samples. This makes it possible to distinguish between variance within a position, variance between positions, and analytical variance. The PQRI approach describes at least three blend samples from at least ten different positions.
The sample mass should correspond to the later individual dose or, on a scientifically justified basis, represent several individual doses. Composite samples that are too large can smooth out local concentration differences and thereby conceal segregation that is actually present. The sampling technique must itself be qualified for the sampling task: depending on geometry, penetration depth, handling and particle properties, a sample thief can draw samples selectively. Recovery, precision, representativeness and possible sampling bias must therefore be demonstrated for the sampling device used.
The assessment must not end at the mixer. If the blend is subsequently tableted, encapsulated or filled, process validation should additionally include systematic in-process dosage-unit sampling across the entire compression or filling run. These samples are drawn stratified over time or production quantity, from the beginning through the middle to the end of the batch, and around relevant events such as hopper changes or stoppages. The ISPE approach, for example, calls for covering the entire manufacturing run with 30 individual dosage units, of which ten can initially be tested distributed across the batch; if needed, the remaining 20 follow according to the chosen acceptance plan.
Validating the analytics
Offline assessment requires a selective, sufficiently precise, matrix-appropriate method. HPLC or UPLC are common reference methods for quantitative API determination. Depending on the active ingredient and formulation, UV/Vis methods, titration or other suitable methods may also be possible. What matters is not the method as such, but the documented proof that it is suitable for the intended application, namely the active-ingredient content in the specific blend matrix at the intended sample quantity.
The current ICH Q2(R2) guideline requires a documented validation strategy, a validation plan, evaluation against acceptance criteria or parameter ranges, and a validation report. Together with ICH Q14, it forms a lifecycle approach to the development, validation, change and ongoing monitoring of analytical procedures.
Typical performance characteristics are specificity or selectivity, range and linearity, accuracy or recovery, repeatability, intermediate precision, robustness and, where relevant to the question at hand, the limit of detection and limit of quantification. Separating process variance from method variance is particularly important. High scatter can arise from the mixer, from the sampling, from sample preparation, or from the measurement method. Method precision and recovery should therefore be determined separately and taken into account when interpreting the blend-uniformity results.
PAT can supplement conventional sample-based analytics. NIR spectroscopy is often suitable for inline, on-line or at-line observation of the mixture. Raman spectroscopy can also be used for a suitable measurement task, for instance where higher molecular specificity is needed. Such systems provide many measurement points over the mixing time and can help to understand the mixing kinetics, identify the point of homogeneity and define a robust endpoint. However, PAT does not automatically replace qualified sampling and testing of individual dosage units. The spectral model, the reference method, the calibration, the measurement location, the data processing, the model monitoring and the change management must be documented and qualified just as thoroughly as an offline procedure. ICH Q14 and ICH Q2(R2) describe this lifecycle-oriented framework for analytical procedures.
Validating the process
Mixing quality should be systematically investigated as early as process development. A design of experiments can show how fill level, mixing time, rotational speed, order of addition, dosing speed, particle size distribution, moisture, API concentration and, where applicable, liquid additions affect homogeneity. From this, critical quality attributes, critical process parameters and a robust operating window can be derived.
Process performance qualification deliberately reflects the planned commercial conditions. These include the actual formulation, the intended batch size, the planned fill level, the production equipment, the approved raw-material specifications, and the final dosing and discharge strategy. A worst-case consideration is advisable: for example minimum and maximum API concentration, different permitted bulk densities, critical moisture ranges, the smallest and largest validated fill level, or raw materials with an unfavourable particle size distribution.
The "three PPQ batches" rule can be a sensible starting point, but it is no substitute for a scientific justification. The number of confirmation batches, the number of samples and the statistical evaluation must be derived from product and process risk, prior knowledge, process understanding and the capability of the control concept. Blend uniformity should be assessed during both process development and process qualification; later routine manufacturing can likewise include systematic in-process dosage-unit sampling for ongoing assurance.
GMP documentation
The documentation must make it traceable, from the raw data, what was planned, what was actually carried out, what results were obtained, and why the acceptance decision is technically justified. The set of documents should include at least the following elements:
A validation master plan or an overarching quality plan with responsibilities, lifecycle strategy and reference to the applicable GMP requirements.
A pre-approved protocol for the mixing-quality study or PPQ, with objective, risk analysis, formulation, batch size, defined acceptance criteria, sampling plan, analytical methods, evaluation plan and provisions for deviations.
A detailed sampling plan with sample position, timing or quantity assignment, sample mass, labelling, the sampler used, sampling instructions, and transport and storage conditions.
Evidence for the qualification of the mixer, the control system, the sampling device, the scales, dosing units and other equipment critical to mixing quality.
Complete batch and process data: raw-material lots, weighed-in quantities, dosing sequence, fill level, mixing time, rotational speed, temperature, moisture, pressure, any interruptions, operator interventions and discharge data.
Analytical raw data including sample preparation, calibrations, system suitability, integration and evaluation parameters, and test protocols.
The statistical evaluation with individual values, mean, standard deviation, RSD, minimum, maximum, comparison between positions and, where applicable, analysis of variance, confidence intervals or tolerance intervals.
Documented deviations, investigations, root-cause analyses, CAPA measures and effectiveness checks.
A final report with a summary of results, an assessment against the predefined criteria, a conclusion on the validated operating window, and a plan for continued process verification.
Electronic records must ensure data integrity: data must be attributable to the person acting, legible, recorded contemporaneously, original, accurate, complete, consistent, enduring and available, and readily retrievable. This is often summarised under ALCOA+. For electronic recipes, process control systems, LIMS, PAT software and electronic test reports, access rights, audit trails, time stamps, data review, backup, archiving and controlled change management must additionally be provided.
Summary
Mixing quality is demonstrated reproducibly for pharma validations when sampling, analytics, process control and documentation are validated together. A low RSD, for which a target value of at most 5 percent is often used for blend uniformity, is not sufficient on its own. The value must come from a predefined, representative and stratified sampling plan that covers the critical zones of the mixing chamber and the entire discharge or filling run.
For examination of the final mixture, at least ten spatially separated, risk-based selected positions should be considered; replicates per position help to distinguish between method, sampling and process variance. The sample mass should be oriented to the later individual dose. For tablets, capsules or filled individual doses, in-process dosage-unit sampling distributed in time across the whole process supplements the evidence from the mixer.
The active-ingredient analysis must be fit for purpose for the specific mixture and application. HPLC or UPLC often supply the reference data; PAT with NIR or Raman can secure the mixing kinetics and mixing endpoint with a high data frequency. ICH Q2(R2) and ICH Q14 require a lifecycle-oriented approach for this, with a validation strategy, defined acceptance criteria, documented data and ongoing monitoring.
Process validation, finally, defines and confirms the robust operating window. It examines critical influencing factors such as fill level, mixing time, rotational speed, order of addition, raw-material variability and discharge under realistic and justified worst-case conditions. Complete GMP documentation, data integrity in line with ALCOA+, deviation management and continued process verification ensure that homogeneity, once demonstrated, remains reproducible in the long term.
How amixon® makes mixing quality measurable, monitorable and documentable
For pharma validations, it is not enough to show a low scatter of individual analytical values. The evidence must demonstrate that the distribution of active ingredient and excipients is achieved reproducibly under defined process conditions, that the sampling actually detects possible inhomogeneities, and that all quality-relevant data are documented in a GMP-compliant manner. amixon® supports this evidence with reproducible mixing processes, suitable sampling concepts, defined PLC recipes, and mixing systems designed to be hygienic and fully dischargeable.
Mixing quality is usually assessed via the relative standard deviation, also called RSD or CoV:
RSD=CoV=sxˉ×100%
Here, s is the standard deviation of the individual values and x̄ is the mean of the measured concentrations. A low RSD value is an important indicator of homogeneity. For pharmaceutical applications, however, it is only reliable if it results from representative sampling, suitable and validated analytics, and clearly defined process control. A target value of at most 5 percent RSD can be a sensible starting point for many mixing tasks, but it does not replace product-specific, scientifically justified acceptance criteria defined in advance in the validation protocol. The requirements must be oriented to the active-ingredient dose, therapeutic window, dosage form, analytical performance, segregation risk and the approved control concept.
Representative sampling
Every mixing-quality assessment stands or falls with the sampling. A seemingly homogeneous sample can give a false impression if it comes only from a favourable zone, or if the sampler selectively captures fine, coarse, cohesive or differently dense particles. The sampling plan is therefore built up on a risk-based and stratified basis as part of validation.
When assessing the final mixture, samples are planned from different spatial zones of the mixing chamber. These include upper, middle and lower product areas, zones near the wall and central zones, the area of the discharge, and any critical areas with an elevated risk of dead space or segregation. The number and position of the samples depend on mixer geometry, batch size, formulation, API content, particle size distribution, flow properties and process risk. A robust starting point is at least ten spatially separated sample positions. For higher risks or complex mixing tasks, a significantly larger number may be required. A published PQRI approach recommends at least three individual samples from at least ten spatially separated positions for many batch mixers; for convective mixers, even more positions are often considered depending on the geometry.
The sample mass should be oriented to the later individual dose or, on a scientifically justified basis, represent several individual doses. Composite samples that are too large can level out local content variations and thereby conceal an inhomogeneity that is actually present. The sampler itself must be qualified. It must be demonstrated that the geometry, sampling depth, handling, product-contact surfaces and emptying of the sampler do not cause any systematic bias. Recovery, repeatability, intermediate precision and possible selectivity toward particular particle sizes or densities therefore form part of the assessment of the sampling procedure.
amixon® mixing systems can be designed with a sampling concept that supports investigation across location and time. In batch mixers, dead-space-free discharge units and the large CleverCut® accessibility enable defined sampling points within the mixing chamber and during discharge. In the pilot plant, the sampling plan is coordinated together with amixon® experts to match the original product, formulation, mixer design and the later process route. A deliberate distinction is made between process variance, sampling variance and analytical variance, so that scatter in the measured values is not prematurely attributed to the mixing process.
The investigation does not end at the mixing chamber. Discharge samples staggered over time, from the beginning, middle and end of discharge, show whether the mixture segregates at the outlet, during conveying, in an intermediate vessel or during filling. For tablets, capsules or other individual doses, in-process dosage-unit sampling stratified across the entire manufacturing run supplements the evidence from the final mixture. For this, individual doses are examined at the beginning, middle and end of the campaign, and where applicable after stoppages, hopper changes or other relevant process events. This connection between blend uniformity and even dosing of the end product is a central component of a robust control concept.
Analytics and PAT
The analytical methods must suit the respective active-ingredient/excipient matrix and be validated for the intended use. HPLC or UPLC are frequently used as reference methods for the quantitative determination of active-ingredient content. Depending on the formulation, UV/Vis methods, titration, gravimetric methods or other specific methods may also be suitable. What matters is that the method captures the active ingredient in the intended sample quantity selectively, accurately and precisely.
Typical performance characteristics include specificity, linearity or working range, accuracy or recovery, repeatability, intermediate precision, robustness and, where relevant, the limit of detection and limit of quantification. Analytical method validation must clearly show what proportion of the total scatter originates from the method and what proportion actually arises from mixing, sampling or the process. ICH Q2(R2) requires a documented validation strategy, a validation protocol, predefined acceptance criteria and a validation report for this purpose. Together with ICH Q14, this guideline forms the framework for the development, validation and ongoing monitoring of analytical procedures.
PAT methods can supplement classical offline analytics. Where the formulation is suitable, NIR spectroscopy enables inline, on-line or at-line observation of the mixing kinetics. A high number of spectra recorded in sequence over time can indicate when a stable mixed state is reached and whether the mixture changes with further mixing. Raman spectroscopy or chemical imaging methods can provide additional information on active-ingredient distribution, material state or local concentration differences for suitable tasks. A prerequisite is always a qualified model secured against reference data, with clearly regulated calibration, model maintenance, monitoring and change management. PAT supplements representative sampling; it does not automatically replace it.
Reproducible process
Reproducible mixing quality begins with a stable and fully defined process. The critical process parameters are established on the basis of development trials and risk analyses. These can include mixing time, tool rotational speed or tool circumferential speed, fill level, the order and speed of raw-material addition, dosing locations, particle size distribution, bulk density, moisture, temperature, pressure, liquid addition and discharge sequence.
amixon® mixers can be operated independently of fill level across a wide range of approximately 10 to 100 percent of the usable volume. Whether the required mixing quality is achieved across the entire intended fill-level range with a given formulation must be confirmed through product trials and as part of qualification. The SinConvex® forced restratification generates a controlled, three-dimensional circulation of the product: near the wall, the product is moved upward, while in the centre it flows downward under gravity and is fed back into the outer mixing zone. These recurring product circuits can support rapid spatial redistribution of all constituents. The rotational speed and mixing time intended for the respective application are stored in the PLC recipe and monitored.
A DoE approach is advisable for process development and validation. Systematic trials make it possible to assess the effects of fill level, mixing time, rotational speed, API concentration, raw-material variability, moisture, order of addition and discharge on mixing quality. This produces a robust process window that describes not just a single ideal point, but a range within which the desired homogeneity is reliably achieved. The investigation should also include realistic worst-case conditions, such as the minimum and maximum intended fill level, critical raw-material lots, unfavourable but permitted moisture ranges, or particularly low active-ingredient concentrations.
The documented mixing recipe contains setpoints and tolerances for all quality-relevant parameters. These include raw-material identity and weighed-in quantity, batch size, fill level, order of addition, dosing times, mixing time, rotational speed, temperature, pressure, liquid quantities where applicable, spray duration, post-mixing time and discharge sequence. Deviations from the setpoint time, setpoint rotational frequency or other defined limits are detected, alarmed and documented. The most effective process monitoring is often to consistently adhere to the once-validated combination of parameters and to assess every relevant deviation in a controlled manner.
GMP documentation
For a pharmaceutical validation, every study must be planned, approved, carried out and evaluated in a traceable manner in advance. A complete set of documents comprises a validation master plan or an overarching quality plan, an approved protocol for the mixing-quality study or process performance qualification, a detailed sampling plan, qualified analytical methods, complete batch and process data, analytical raw data, a statistical evaluation, and a final validation report.
The protocol defines the objective, responsibilities, formulation, batch size, fill level, acceptance criteria, number of confirmation batches, number of samples, sample positions, sample mass, sample labelling, analytical methods, statistical evaluation, deviation management and decision rules. The raw data must include individual values, mean, standard deviation, RSD, minimum, maximum and, where applicable, comparisons between positions, analyses of variance, confidence intervals or tolerance intervals. Deviations must be assessed, investigated and addressed with suitable CAPA measures.
Data capture must meet the principles of ALCOA+: data must be attributable to the person acting, legible, recorded contemporaneously, original, accurate, complete, consistent, enduring and available, and readily retrievable. For electronic systems such as the PLC, process control system, LIMS, PAT software or ERP connection, role-based access rights, audit trails, time stamps, verifiable data backup, archiving and controlled change management are required. This makes it possible to trace which raw materials were used in which batch, which mixing parameters applied, which process values were actually present, who made interventions, and how the batch was assessed or released.
amixon® systems can be equipped with barcode or RFID capture and a connection to the operator's ERP system. This allows raw-material lot, formulation, mixing program, process parameters, operator interventions, cleaning status and batch release to be linked together in a traceable way. In GMP environments, amixon® can support the preparation and compilation of documentation for Design Qualification, Installation Qualification and Operational Qualification. Process performance qualification and the product-related validation decision remain, however, with the pharmaceutical manufacturer within its own quality system.
Hygienic design and CPV
Reproducible mixing quality also depends on hygiene, residual discharge and cleanability. Product residues from a previous batch, active-ingredient build-up, moisture or cleaning residues, and hard-to-reach areas can affect the quality of a subsequent batch. Constructional features are therefore relevant for pharmaceutical applications, such as product-contact surfaces welded free of joints and ground smooth, mixing tools supported only at the top with no lower shaft passage in the product area, large CleverCut® inspection doors with a permanently dead-space-free OmgaSeal® seal, dead-space-free discharge units, and integrated wash lances.
Cleaning, whether carried out dry or wet, must be validated for the respective worst-case product changeover. This includes defining critical contact surfaces, establishing suitable sampling or rinse points, analytical detection limits, and acceptance criteria for cleaning residues. Depending on the project, hygiene concepts can be aligned with EHEDG guidelines, FDA hygiene guidelines and the 3-A Sanitary Standards.
Validation does not end after successful process performance qualification. As part of continued process verification, mixing parameters, raw-material data, analytical results, RSD values, discharge data, deviations and trends are monitored on an ongoing basis. This makes it possible to identify early whether raw-material quality, equipment condition, dosing performance, cleaning effectiveness or mixing behaviour is changing. Ongoing statistical process monitoring ensures that the process window demonstrated during qualification is maintained in the long term.