What accuracy is realistically achievable with micro-dosing systems for vitamins and enzymes?
When dosing vitamins, enzymes and other highly active micro-components, gravimetrically controlled systems are generally the most robust solution. Unlike purely volumetric dosing units, they record the mass actually dosed and can compensate better for fluctuations in product flow. Under industrial conditions, with suitable design, stable product properties and a sufficiently large target quantity, dosing deviations of less than ±1 % of the set point are realistic; values of around ±0.5 % are frequently achievable. Requirements down to approximately ±0.1 % may be attainable in specially optimised applications, but must be demonstrated specifically for the product concerned, the dosing range and the actual operating conditions.
How precisely a micro-dosing system works does not depend on the load cell or the dosing unit alone. What is decisive above all is the flowability, bulk density, moisture, grain shape and particle size distribution of the powder. Cohesion, bridging, compaction, electrostatic charging and build-up likewise influence the mass actually discharged. The smaller the weighed-in quantity or the continuous mass flow, the greater the relative effect of these influences. Dosing in the ppm range is therefore technically possible, but places high demands on the dosing unit, product guidance, calibration and validation.
With continuous loss-in-weight systems, the mass flow is regulated via the measured decrease in weight of the supply hopper. Realistic accuracy figures frequently lie in the region of approximately ±0.5 to ±1 %, but must always be related to a defined mass flow and a defined time window. Short-term fluctuations can be greater during the refilling phase or with changing flow behaviour, for example, than the deviation averaged over a longer period. With batch dosing on the gain-in-weight principle, by contrast, the target vessel is filled up to the specified set weight; this method is particularly suitable where individual recipe constituents are to be weighed in exactly.
Volumetric dosing units such as screw or rotary valve feeders are not fundamentally unsuitable. They can work economically and with sufficient repeatability with free-flowing products of constant bulk density. If bulk density, moisture, fill level, compaction or flow behaviour change, however, the mass actually dosed can shift despite an unchanged setting. For valuable, highly active or quality-critical components such as vitamins and enzymes, gravimetric feedback is therefore usually preferable.
The distinction between repeatability and accuracy is also important. High repeatability means that successive dosings scatter only slightly. It does not, however, ensure that the mean value corresponds exactly to the set weight. That requires correct calibration, allowance for in-flight material, a suitable reference weighing and regular checking of the overall system. The dosing accuracy actually achievable is thus always the result of the interplay of product, dosing technology, control, mechanical installation situation and consistent process management.
Micro-dosing and mixing quality: two separate quality characteristics
With vitamin, enzyme and trace element premixes, two quality characteristics must be considered together: the accuracy of the dosing and the homogeneity of the subsequent mixture. The dosing system determines what mass of a micro-component is introduced into the process. The mixer then ensures that this component is distributed evenly throughout the total quantity. High mixing quality cannot correct faulty dosing; conversely, even an exactly dosed micro-component does not automatically produce a homogeneous premix.
Modern gravimetric dosing systems, for example gain-in-weight or loss-in-weight systems, can achieve high accuracies with readily dosable products and sufficiently long dosing times. In industrial use, deviations of less than ±1 % of the set point are frequently realistic; values of around ±0.5 % are attainable with suitable design. Higher accuracies always presuppose an application-specific check. Purely volumetric dosing units can work economically where flow behaviour is constant and bulk density is stable, but they react more sensitively to changes in bulk density, moisture, compaction or flowability.
For vitamin and enzyme premixes, therefore, the nominal precision of a dosing unit is not the only relevant factor. Equally important are the reproducibility of the material flow, a suitable charging strategy, the avoidance of build-up and bridging, and sufficient dosing and mixing time. The permissible dosing tolerances follow from the formulation, from labelling and product safety, from the margin for overdosing and from the homogeneity requirements for the finished premix.
Mixing quality in the amixon® precision mixer
amixon® precision mixers of the VM, HM, AM, KoneSlid® and SpherHelics® series are designed for the intensive yet product-protecting homogenisation of dry, moist or suspended bulk materials. With a suitable formulation, reproducible charging and a validated mode of operation, they can also distribute very small component proportions reliably. amixon® defines precision mixing for concentrations down to 1:100,000 as a reproducibly achievable coefficient of variation below 5 %. This statement must, however, always be verified within the framework of specific product trials, a suitable sampling plan and sufficiently sensitive analytics.
An "ideal random mixture" should not be described as a blanket product property. It is rather a statistical target that depends on the particular material system, the formulation, the fill level, the mixing time, the charging and the sampling. With very small active ingredient proportions in particular, agglomerates, differences in grain size or density and subsequent segregation during discharge or transport can affect homogeneity.
A widely usable fill level range can offer great flexibility in day-to-day production. Nevertheless, whenever batch size, formulation or fill level changes, it should always be checked whether mixing time, speed, charging sequence and sampling plan remain suitable. Renewed validation is not dispensable as a matter of principle but depends on the risk and the quality management of the operator concerned.
Product protection and de-agglomeration
For even distribution, micro-components must be present in the premix in as fine and reproducible a form as possible. Cohesive powders, lumped vitamins or agglomerated enzyme preparations can limit homogeneity if they are not sufficiently broken up. At the same time, excessive energy input can impair sensitive active ingredients, coatings, granulates or encapsulated components.
In basic operation, amixon® mixers work at comparatively low circumferential speeds and thus permit product-protecting mixing. Where necessary, a cutting rotor can be used specifically for de-agglomeration. What is decisive is the application-specific design: as much shear and size-reduction energy as necessary, but as little as possible. Success should be assessed not by mixing time alone but through analytically supported homogeneity tests, particle examinations and, where appropriate, activity measurements of the active ingredient.
Continuous precision mixing with the AMK
Where micro-dosing and mixing take place continuously, dosing, control and the mixing process must be considered as an overall system. The gravimetric dosing units feed the individual components continuously in the intended mass ratio. The continuous mixer homogenises these material streams and discharges the mixture continuously. The stability of the dosing streams, the control during refilling operations, the residence time distribution and the mixing dynamics are particularly decisive for the resulting product quality.
The AMK continuous mixer can be integrated into continuous production processes. According to amixon®, its fill level is adjustable across a wide range; the residence time is independent of the rotational frequency of the mixing tools. In continuous operation, the already homogenised product inventory present in the mixer forms a mixing base. Newly entering, comparatively small material streams are mixed into it. A sufficiently generous mean residence time can attenuate short-term fluctuations in the inlet concentration over time; amixon® describes this effect as the low-pass behaviour of the continuous mixing process.
The residence time is thus an important design parameter, but no substitute for stable dosing. Depending on the process and the mixer principle, mean residence times can lie in the region of 0.5 to 3 minutes, for example. Their distribution must be designed so that the required homogeneity is achieved in steady-state operation and so that product changes, start-up procedures and dosing deviations are kept under control.
Trials with original products
Whether a specific target can be met can be assessed reliably only in trials with representative original products. In the amixon® pilot plant, the mixture can be investigated under conditions close to practice. The focus is above all on the mixing quality of the chosen mixer, the behaviour of cohesive or agglomerated micro-components, the influence of fill level or throughput, and the tendency to segregate during discharge.
The dosing accuracy itself is determined by the particular dosing system selected by the customer or the plant integrator. For a sound overall assessment, dosing unit and mixer should therefore be considered and tested together. A suitable sampling plan, stratified in space and time, together with the separation of process, sampling and analytical scatter, creates the basis for a sound assessment vis-à-vis quality assurance, customers and, where applicable, the authorities.