Mixing scale up 1.pdf

SCALE –UP OF DRY BLENDING AND MIXING PROCESS BITS Pilani Pilani Campus Definition - Mixing Process that tends to result in a randomization of dissimilar particles in a system BITS Pil...

SCALE –UP OF DRY BLENDING AND MIXING PROCESS BITS Pilani Pilani Campus Definition - Mixing Process that tends to result in a randomization of dissimilar particles in a system BITS Pilani, Pilani Campus Principle mechanisms of mixing Diffusion---random action of individual particles in the mix Convection---transfer of adjacent particles groups in the mix Shear---configuration change through slip planes BITS Pilani, Pilani Campus BATCH MIXERS AND MECHANISMS Three common tumbler designs: (a) double cone, (b) V, and (c) tote or bin blenders. BITS Pilani, Pilani Campus  BATCH MIXERS AND MECHANISMS Rapid convective flow seen in particle-dynamic simulation of identical but colored spheres in a V-blender. Top: Front view reveals that unlike in some designs, convection in this blender drives grains axially, alternately outward toward the tumbler arms and inward toward its center. This axial flow strongly influences mixing. Bottom: Side view indicates that transport is dominated by a spiraling flow, seen also in drums and other blenders BITS Pilani, Pilani Campus BATCH MIXERS AND MECHANISMS Dispersive mixing is slow across the symmetry plane of a blender, here a bin design. After 10 revolutions, a front view reveals clear evidence of the initial left–right distribution of identical but colored spheres in this particle-dynamic simulation. BITS Pilani, Pilani Campus Solid-Solid Mixing---The unit operation The unit operation of solid-solid mixing can be separated into four principle steps: The bed of solid particles expands. Three dimensional shear forces become active in the power bed. The powder bed is mixed long enough to permit true randomization of particles. Randomization (no segregation), of the particles is maintained after mixing has stopped. BITS Pilani, Pilani Campus Solid-Solid Mixing---The unit operation Mixing forces and bed expansion. BITS Pilani, Pilani Campus Solid-Solid Mixing---The unit operation Mixing is an energy-consuming process which produces a random distribution of particles. It is dependent on the probability that an event happens in a given time. The law of mixing M = A[1-ekt] where M= Degree of mixing T= time A = initial resistance of the dry powder or granules K= rate at which fine powder components Plot of first order decay for mixing. disperse throughout entire system BITS Pilani, Pilani Campus  Perfect mixtures Perfect mixture: That state in which any sample removed from the mixture will have exactly the same composition as all other samples taken from the mix Illustrations of perfect and randomized binary mixtures. (a) Perfect mix and (b) randomized mix BITS Pilani, Pilani Campus  Alternatives to the Perfect Mixture ❑ There is randomization or random mixing which is that state in which the probability of finding a particle of a given component is the same at all points in the mixture and is in the same ratio of components in the entire mixture. ❑ Ordered mixing is described as the use of mechanical, adhesional, or coating forces or methods to prepare ordered units in the mix such that the ordered unit will be the smallest possible sample of the mix, and will be of near identical composition to all other ordered units in the mix, e.g., a dry granule made by wet granulation or a coated particle such as we see in fluid bed granulating. BITS Pilani, Pilani Campus Alternatives to the Perfect Mixture ORDERED MIXING--- MECHANICALLY. Dividing and recombining the powder bed any number of times until the desired subdivision unit is obtained. The smaller the units, the more uniform BITS Pilani, Pilani Campus Alternatives to the Perfect Mixture ORDERED MIXING---ADHESION Adhesional forces of particles may create ordered units of near identical composition depending on the process. Partial solubilization or the use of a binding agent during wet granulating approximates the same effect BITS Pilani, Pilani Campus Alternatives to the Perfect Mixture ORDERED MIXING---COATING Particles in an assemblage may also be coated with other ingredients to give an ordered mix either as individual or coated particle agglomerates BITS Pilani, Pilani Campus Segregation by Pouring BITS Pilani, Pilani Campus Segregation by flow from a hopper or blender BITS Pilani, Pilani Campus Sampling and Statistics of Mixing People do not often have mixing problems, but actually have sampling problems. Random samples are used to generate the data necessary to estimate the true mean and standard deviation of the mixture. 1. The estimation of the true arithmetic mean and the true standard deviation of a mixture 2. An accuracy and precision assessment of these estimates A Gaussian distribution yielding a normal or bell type curve. From the assays of the samples BITS Pilani, Pilani Campus Sampling and Statistics of Mixing The mean assay value of a group of random samples taken from a mixture is a measure of the central tendency of the batch population (active ingredient content). The arithmetic mean value is given by yi = the value of a given sample n = number of samples BITS Pilani, Pilani Campus Sampling and Statistics of Mixing an additional statistic is calculated to describe or characterize the spread or dispersion of individual samples about the mean. This is the standard deviation S for n number of random samples, which is given by Based on the numbers of samples used to obtain the y estimate of the true mean, so the standard deviation S is an estimate of the true population standard deviation s of the total mixture. BITS Pilani, Pilani Campus Five- and 10-Min Mix-Time Assay Data BITS Pilani, Pilani Campus Sampling and Statistics of Mixing BITS Pilani, Pilani Campus Sampling and Statistics of Mixing BITS Pilani, Pilani Campus Sampling and Statistics of Mixing It is important to note at this point that differences in the assay mean from the assay theory may not only indicate homogeneity problems, but may also be the result of poor or inadequate sampling, improper handling of the powder for assay in the lab, or may point up an assay method problem. BITS Pilani, Pilani Campus Sampling and Statistics of Mixing BITS Pilani, Pilani Campus Sampling and Statistics of Mixing BITS Pilani, Pilani Campus Sampling and Statistics of Mixing The solution to the problem lies within certain statistical guidelines including: 1. The proper number of samples required should be no less than 20, preferably 30, and more ideally 100. 2. Random sampling is the method of choice for the statistic. 3. The sampling size for determining content uniformity should approximate the unit dose size of the final product. This is an absolute necessity to show process control at the mixing stage of the product manufacture. 4. Comparison between the mean value of the sample analysis and the target value gives the first estimate of the degree of mixing. 5. 5. If the mean value of the sample analysis is on or near the target value, then calculation of the standard deviation (and/or variance) will give an indication of the uniformity of the samples. The acceptance criteria for uniformity is the USP test for tablet uniformity. BITS Pilani, Pilani Campus Sampling and Statistics of Mixing BITS Pilani, Pilani Campus Sampling and Statistics of Mixing BITS Pilani, Pilani Campus Material Properties: Basic Concepts of Dry Blending Several different types of particles encountered in tablet granulation dry blending. BITS Pilani, Pilani Campus Material Properties: Basic Concepts of Dry Blending Table: Effect of Particle Size on Powder Flow PARTICLE SIZE TYPE OF FLOW REASON Flow is usually good if Mass of individual particles shape is not interfering is relatively large Flow properties may be a Mass of individual particles problem with many pure is small and increased substances and mixtures surface area amplifies effects of surface forces Flow becomes a problem Cohesive forces or free with most substances surface energy forces are large as well as static electrical forces relative to particle size BITS Pilani, Pilani Campus Material Properties: Basic Concepts of Dry Blending Effects of particle size distribution on the bulk density of a powder DENSELY PACKED POWDERS USUALLY HAVE FLOW DIFFICULTIES!!!! BITS Pilani, Pilani Campus Material Properties: Basic Concepts of Dry Blending General particle shapes and their effect on power flow Particle shape affects powder inter-particle friction, and consequently the flow properties of the powder!!!! BITS Pilani, Pilani Campus Mixing Equipment A. Batch Type High-speed Rotation of the Rotation of the Stationary shell granulations Air mixer stationary entire mixer shell entire mixer shell or or body with a (stationary body with shell or body using or body with no body with a rotating rotating mixing rotating mixing blade moving air as agitator or mixing high shear blade & high-speed agitator agitator blade blade) Barrel V-shaped Ribbon Fluid bed processor Barrel granulator Sigma cube Double cone blade Bowl Fluid bed formulator V-shaped Planetary drier Slant double cone Conical Double cone formulator screw Slant double cone B. Continuous type Zig Zag BITS Pilani, Pilani Campus Schematic of rotating shell blenders The first general class of mixers are those that create particle movement by rotation of the entire mixer shell or body. BITS Pilani, Pilani Campus V-shaped, double cone, and slant double-cone blenders advantages and limitations ADVANTAGES Minimal attrition when blending fragile granules Large capacity equipment available Easy to load and unload Easy to clean Minimal maintenance DISADVANTAGES High head space needed for installation (particularly with V-shaped mixers) Segregation problems with mixtures having wide particle size distribution and large differences in particle densities Tumbling-type blenders not suitable for fine particulate systems because there may not be enough shear to reduce particle agglomeration Serial dilution required for the addition of low dose active ingredients if powders are free flowing BITS Pilani, Pilani Campus V-shaped, double cone, and slant double-cone blenders mixing efficiency Effect of Powder Fill on Blending Time of Double-Cone Blenders Volume percent of Approximate blend time blender filled with (minutes) in production-size powder charge blenders 50 10 65 14 70 18 75 24 80 40 These blenders are operated by adding material to be blended to a volume of approximately 50 to 60% of the blender's total volume.!!!! BITS Pilani, Pilani Campus V-shaped, double cone, and slant double-cone blenders mixing efficiency Higher blending speeds provide more shear, more dusting may be prevalent causing segregation of fines There is also a critical speed which, if approached, will diminish blending efficiency of the mixer considerably. BITS Pilani, Pilani Campus Schematic of rotating shell blenders with agitator mixers The agitator blade gives added versatility to the tumbling blenders by virtue of the high shear attainable!!! BITS Pilani, Pilani Campus Rotating shell blenders with agitator mixers V-shaped blender with agitator mixing assembly BITS Pilani, Pilani Campus Rotating shell blenders with agitator mixers advantages and limitations ADVANTAGES ❑ Good versatility in that both wet and dry mixing can be accomplished in the blender. ❑ A wide range of shearing force may be obtained with the agitator bar design permitting the intimate mixing of very fine as well as coarse powder compositions. ❑ Serial dilution more than likely may not be needed when incorporating low dose active ingredients into the mixture. DISADVANTAGES. ❑ Possible attrition of large more friable particles or granules in a mixture as a result of the high-speed agitator mixer. ❑ Scale-up can prove to be a problem in that direct scale-up based on geometry, size, and peripheral velocity in many cases does not work. ❑ Cleaning may be a problem depending on design, since the agitator assembly must be removed and packings changed for a product changeover. ❑ Potential packing (seal) problems (packings are used to prevent leakage through the shaft entrance into the mixing chamber and to prevent the blender contents from contaminating the bearings). BITS Pilani, Pilani Campus Stationary shell or body with a rotating mixing blade Ribbon mixer BITS Pilani, Pilani Campus  Mixing Problems There are no hard, fast rules or equations to direct one to the use of a particular size and type of mixer during scale-up. Results are usually empirical, dependent heavily on the experience of the people responsible for the project Tablet and/or granulation production department already has specific sizes and certain type of mixers on hand. Selecting and purchasing a new blender suited for a specific new product may not be possible BITS Pilani, Pilani Campus  Mixing Problems PROBLEM A. Uniformly disperse high potency low dose active ingredient in a diluent for direct compression SUGGESTED APPROACHES 1. Mill or pass active ingredient and equal amount of diluent through a small screen. Rinse screen with more diluent. Place in mixer with 1/2 of remaining diluent and mix. Add remaining diluent and additives (starch, microcrystalline cellulose, etc.) and mix to a final mixture. 2. Place active ingredient, all of diluent and additives in mixer with high speed for mixing bar or chopper. Mix to a final blend. SUGGESTED EQUIPMENT 1. Cutting or hammer mill with small screen or moisturized sifter. Use tumbling mixer: V- shaped or double cone. May use sigma, ribbon, or conical screw type mixers. Should create minimum dust. 2. V-shaped or double-cone blender with high-speed agitator bar, horizontal or vertical high-speed granulator BITS Pilani, Pilani Campus Mixing Problems PROBLEM B.If uniform dispersion of high potency low dose active ingredient in a diluent for direct compression cannot be achieved. SUGGESTED APPROACHES: 1.Dissolve active ingredient in a granulating solution solvent and wet granulate in preblended diluent and additives to uniformly disperse SUGGESTED EQUIPMENT: 1.Sigma blade, ribbon, planetary, conical screw, or highspeed mixer. The mixer must have high enough shear to distribute granulating-active ingredient solution in preblended diluent, and additives. May also use a fluid bed granulator for this. Fluid bed drying is recommended to minimize active ingredient migration during drying. A mill may also be necessary to bring to final granule size. BITS Pilani, Pilani Campus Mixing Problems PROBLEM C.Uniformly disperse small quantities of dye lakes throughout diluent with low dose high- potency active ingredient for direct compression. SUGGESTED APPROACHES 1.Mill or pass active ingredient, dye lake and equal amount of diluent through a small screen. Rinse screen with more diluent. Place in mixer with 1/2 of remaining diluent and mix. Add remaining diluent and additives (starch, microcrystalline cellulose, etc.) and mix to a final mixture. 2.Use of high-speed mixer bar or chopper with all diluent and additives. SUGGESTED EQUIPMENT 1.Cutting or hammer mill with small screen or moisturized sifter. Use tumbling mixer: V- shaped or double cone. May use sigma, ribbon, or conical screw type mixers. Should create minimum dust. 2.V-shaped or double-cone blender with high-speed agitator bar, horizontal or vertical high- speed granulator BITS Pilani, Pilani Campus Mixing Problems PROBLEM D. Uniform dispersion of small quantities of dye lake throughout high-dose, large volume product for direct compression. SUGGESTED APPROACHES 1. Mill or pass dye lake and small quantity of additives through a small screen. Rinse screen with more additive. Place in mixer with 1/2 of remainder of ingredients and mix. Add remainder of ingredients and mix to final blend. SUGGESTED EQUIPMENT 1. Cutting or hammer mill with small screen or moisturized sifter. Use tumbling mixer: V-shaped or double cone. May use sigma, ribbon, or conical screw type mixers. Should create minimum dust. BITS Pilani, Pilani Campus Mixing Problems PROBLEM E. Uniform dispersion of dyes in high- or low-dose products SUGGESTED APPROACHES 1.Dissolve active ingredient in a granulating solution solvent and wet granulate in preblended diluent and additives to uniformly disperse SUGGESTED EQUIPMENT: 1.Sigma blade, ribbon, planetary, conical screw, or highspeed mixer. The mixer must have high enough shear to distribute granulating-active ingredient solution in preblended diluent, and additives. May also use a fluid bed granulator for this. Fluid bed drying is recommended to minimize active ingredient migration during drying. A mill may also be necessary to bring to final granule size. BITS Pilani, Pilani Campus Mixing Problems PROBLEM F. Poor flowing cohesive powders in general SUGGESTED APPROACHES 1. Use high-shear equipment. 2. Fluidized bed. SUGGESTED EQUIPMENT 1. Sigma, ribbon, and planetary mixers. V-shaped and double-cone blenders with agitator bars or high speed granulators. 2. Fluid bed granulators. BITS Pilani, Pilani Campus Mixing Problems PROBLEM G.Over mixing of lubricants yielding granulation or formula dry mixture with poor lubricating properties. SUGGESTED APPROACHES 1.Cut blending time if it does not interfere with homogenicity of final blend. 2.Use lower shear mixer if it does not interfere with homogenicity of final blend. 3.Leave lubricant out of final blend until last 5-10 min of mixing. SUGGESTED EQUIPMENT 1.High shear mixing equipment, e.g., ribbon, sigma, high-speed granulators and tumbling mixers with agitator blades. 2. Tumbling mixers. 3. Both of the above. BITS Pilani, Pilani Campus Mixing Problems Effect of active ingredient concentration on homogeneity of mixture. (Adapted from J. A. Hersey, Industrial Awareness Seminar, Powder and Bulk Solids Conference/Exhibit, Philadelphia, May 1979.) BITS Pilani, Pilani Campus Scale up considerations ✓ Rotation rate ✓ Filling level ✓ Operation time ✓ Variation in blender geometry Department of Pharmaceutics 51 BITS Pilani, Pilani Campus General mixing considerations ✓ Defining Mixedness ✓ Mixing Issues in Tumbling Blenders ✓ Process Parameters Department of Pharmaceutics 52 BITS Pilani, Pilani Campus Scale up approaches Froude number Fr = µ2R/g, Where: µ is the rotation rate, R is the vessel radius, g is the acceleration from gravity Scale up of 5 lt tumbling blender to 25 lt tumbling blender, Initial operated at 15 rpm for 15 min. Geometric similarity is the essential criteria Department of Pharmaceutics 53 BITS Pilani, Pilani Campus Scale up approaches Tip Speed Approach Tip speed approach is mainly used in the scale up of rapid mixer granulator (dry and wet mixing) and blender (dry mixing). A tip speed is defined as the tangential velocity of an impeller at a point on its tip. The tip speed is a function of the RPM and diameter of the impeller. The following formula is used to calculate the tip speed of an impeller. TS = pi*D* RPM / 60 Where TS is the tip speed (distance/second) D is the diameter of the impeller (m or ft) RPM is the rotations per minute BITS Pilani, Pilani Campus Scale up approaches Example 1: During the formulation of a solid oral dosage form, a blending process was optimized at a pilot plant scale up using a 1200 L V-blender. The 1200 L blender was operated for 12 rpm for 25 min to obtain the desired blend uniformity. Calculate the operating parameters, if the blending process was scaled up from 1200 L to 6000 L V- blender. Tip radius of 1200 L V Blender is 1000 mm and for 6000 L V-blender, it is 1600 mm. BITS Pilani, Pilani Campus Scale up approaches Calculate the tip speed for 1200 L V-Blender at 12 rpm (https://www.simsite.com/tip-speed-calculator): Tip speed= 0.628 m/s Now, calculate the tip speed for the 6000 L V-blender, assuming rpm as 12: Tip speed= 1.005 m/s Since the tip speed for the 1200 L V-blender and 6000 L V-blender does not match, it is not recommended to operate the 6000 L V- blender at this rpm. Now, change the rpm of the 6000 L V-blender in such a way to match the tip speed approx. 0.628 m/s: At 8 rpm, the tip speed for a 6000 L V-blender is calculated to be 0.670 m/s, while at 7 rpm, it is 0.586 m/s, and at 9 rpm, it is approx. 0.754 m/s. Therefore, we can finalise the tip speed of 0.670 m/s at 8 rpm, which is closer to tip speed of a 1200 L V-blender. Minutes are kept the same as that operated for 1200 L V-blender, i.e., 25 minutes. BITS Pilani, Pilani Campus Scale up approaches Limitations of the Tip Speed approach Tip speed approach does not let the user to alter the time for which the machine is operated. The time of mixing could have a significant impact to the degree of randomization achieved. Moreover, it is important to consider blend fill level and their particle size distribution for appropriate mixing of the blend, which needs to be taken into consideration by the tip speed approach. BITS Pilani, Pilani Campus Scale up approaches Number of revolution approach This approach take into consideration of time and rpm and calculates total number of revolution. In case of scale up, there may be cases where the same rpm is not possible to execute. In such case, time is increased so as to match total number of revolution. Total number of revolution= Time X rpm BITS Pilani, Pilani Campus Scale up approaches Example 1: During the formulation of a solid oral dosage form, a blending process was optimized at a pilot plant scale up using a 200 L Bin blender. The 200 L blender was operated for 15 rpm for 20 min to obtain the desired blend uniformity. Calculate the operating parameters, if the blending process was scaled up from 200 L to 1500 L Bin blender. Assume we have to operate 1500 L at 12 rpm as the machine is qualified in 8-12 rpm range. Total number of revolution in 200 L Bin Blender = 20 X 15= 300 Since, we have to operate 1500L Bin Blender at 12 rpm, we shall calculate time required to match number of revolutions= 300/12=25 minutes. Therefore, 1500 L Bin Blender should be operated at 12 rpm for 25 minutes so as to match the number of revolutions of 200 L Bin Blender. BITS Pilani, Pilani Campus Scale up approaches Limitations of the number of revolution approach Although matching of number of revolution is most simple and easy way to scale up and scale down and used very widely in pharmaceuticals system, this method has certain limitations. Number of revolution approach do not take into consideration of the tip speed and it may eventually lead to a long time, in some cases, which may be 35 or 40 minutes of mixing. Such a long time of mixing may result in demixing. BITS Pilani, Pilani Campus Recommendations and conclusions ✓ Make sure that changes in scale have not changed the dominant mixing mechanism in the blender (i.e., convective to dispersive). ✓ The number of revolutions is a key parameter, but rotation rates are largely unimportant. ✓ When performing scale-up tests, take enough samples to give an “accurate” description of the mixture state in the vessel. Department of Pharmaceutics 61 BITS Pilani, Pilani Campus  Recommendations and conclusions ✓ One simple way to increase mixing rate is to decrease the fill level— while this may be undesirable from a throughput point of view, decreased fill level also reduces that probability that dead zones will form. ✓ Addition of asymmetry into the vessel, either by design or the addition of baffles, can have a tremendous impact on mixing rate. Department of Pharmaceutics 62 BITS Pilani, Pilani Campus