Granulation with Dow Cellulosic Polymers II. High Shear Granulation

Granulation with Dow Cellulosic Polymers 

II. High Shear Granulation 

Introduction 

A wide variety of materials have been 

used as binders for pharmaceutical 

preparations, including cellulosic 

polymers. Over the last several years, 

newer granulation techniques have 

come to the fore. This, the second in 

a series of Technical Data Sheets 

from Dow, is designed to meet the 

needs of the pharmaceutical formulator 

in addressing these techniques 

and the value that METHOCEL* and 

HPC (hydroxypropyl cellulose) 

products have as pharmaceutical 

binders. The data presented here 

examines the use of a high shear 

mixer-granulator, in which the binders 

were applied both from aqueous 

solutions and by adding the binders 

to the dry mix followed by granulation 

with water. 

Throughout this series of granulation 

studies, it is deemed important to 

provide data not only on the 

performance of several different 

METHOCEL and HPC polymers, but 

to provide side-by-side comparison 

data for other commonly-used 

binders. Therefore, information is 

supplied for granulations prepared 

with two viscosity grades of povidone 

(polyvinylpyrrolidone, or PVP), acacia, 

and a pregelatinized starch. 

Granulation was performed in a Fuji 

vertical high shear mixer-granulator, 

model VG-25P; this model has a 25 L 

bowl. The charge in each granulation 

was 4 kg. Granulations were 

prepared at 3% and 6% binder by 

weight, with the exception of the 

pregelatinized starch, which was 

used only at 10%. The particle size 

distributions were performed on an 

ATM Sonic Sifter. The apparent and 

tap densities were determined by 

standard techniques. The 

*Trademark of The Dow Chemical Company 

† Previously referred to as hydroxypropyl methylcellulose or HPMC. 

formulations were compressed using 

0.5 inch diameter standard concave 

tooling on an instrumented tablet 

press at a series of total compression 

forces. Weight and thickness variation 

as well as tablet hardness and 

friability were also measured by 

standard techniques. Finally, tablet 

dissolution was recorded using a 

system that consisted of a Perkin- 

Elmer 3B spectrophotometer, with a 

programmable cell changer 

containing six flow-through cells and 

interfaced with a personal computer, 

a multi-channel peristaltic pump, and 

a Distek dissolution apparatus. 

Applicable USP dissolution testing 

conditions were followed, with the 

exception that the release of the drug 

was followed via continuous flow from 

the dissolution apparatus through the 

spectrophotometer. 

Three model systems were studied. 

Acetaminophen was used as an 

example of a high-dose, low-solubility 

drug; Vitamin C was examined as an 

example of a high-dose, highsolubility 

drug; and methazolamide 

was investigated as an example of a 

low-dose, low-solubility drug. While a 

detailed discussion is contained 

within the section dealing with each 

model system, a few general comments 

can be made here. Cellulosic 

polymers such as METHOCEL 

methylcellulose (MC) and 

hypromellose † and HPC have excellent 

binding properties. In some 

cases, they both outperform and are 

more cost effective than other commonly 

used binders. They are familiar 

materials, being used in many other 

pharmaceutical applications, and 

are well known for their excellent 

safety and toxicological properties. 

Finally, METHOCEL and HPC 

products have broad international 

regulatory approval. 

Model Formulation 1: 

Acetaminophen 

Composition and Preparation 

The acetaminophen formulations 

used in this study are given by the 

following: 

50% Acetaminophen 

44.5% Lactose 

3.0% Binder 

2.0% Disintegrant 

0.5% Lubricant 

75% Acetaminophen 

16.5% Lactose 

6.0% Binder 



The acetaminophen used was 

Mallinckrodt USP powder, having a 

purity of 99.9%. The lactose was 

Fast-Flo Lactose 316 from Foremost 

Whey Products. Ac-Di-Sol crosslinked 

sodium carboxymethylcellulose from 

FMC was the chosen disintegrant. 

Finally, the magnesium stearate 

used as a lubricant was supplied 

by Whittaker. 

A wide variety of binders were 

evaluated. As stated above, this was 

done not only to determine how each 

of the seven METHOCEL and HPC 

binders performed in the formulation, 

but also to compare their 

performance with other commonlyused 

granulation binders. The 

METHOCEL materials chosen were 

among the lowest viscosity products 

available in each substitution type. 

This was done in order to maximize 

the solids loading while still giving a 

solution of manageable viscosity in 

those experiments where the binder 

was placed in solution. It is 

recognized that polymer molecular 

weight (which is directly related to 

1

solution viscosity) can be an 

important variable in the effectiveness 

of a binder. The effect of molecular 

weight will be addressed in 

subsequent studies. 

The METHOCEL polymers used 

included: METHOCEL A15P LV 

(15 cps methylcellulose USP); 

METHOCEL E5P LV and E15P LV 

(5 and 15 cps hypromellose, USP 

substitution type 2910); METHOCEL 

K3P LV (3 cps hypromellose, USP 

substitution type 2208); and 

METHOCEL F4P LV (hypromellose 

USP substitution type 2906). The 

METHOCEL polymer is in many ways 

similar to the material identified as 

METHOCEL F4P LV in Part I, Fluid 

Bed Granulation. By the USP 

monographs, MC and hypromellose 

viscosities are reported as those 

obtained from 2% aqueous solutions, 

measured in a Ubbelohde (capillary) 

viscometer at 20°C. 

The HPC polymers utilized were the 

HPC-EF and HPC-LF grades. HPC- 

EF has a viscosity of 250-750 cps, 

measured as the viscosity of a 10% 

aqueous solution at 20°C with a 

Brookfield viscometer, using the #2 

spindle at 30 rpm. The viscosity of 

HPC-LF is 75-150 cps, measured as 

a 5% aqueous solution at 20°C, again 

using the Brookfield viscometer, #2 

spindle, and 30 rpm. 

2 High Shear Granulation: Acetaminophen Model 

The povidones used in this study 

were USP grades of the K29-32 and 

K90 types. 

Acacia (gum arabic, spray dried 

SDGA/1) was obtained from AEP 

Colloids, Inc. The spray dried form is 

white to very pale yellow and is in 

compacted fragments or whole 

spheres. One gram dissolves in 2 ml. 

of water, forming a solution which 

flows readily and is acidic to litmus. 

A pregelatinized starch having a coldwater 

soluble fraction of 10 to 20% 

was included in the study. This type of 

starch is a combination of 

approximately 5% amylose, 15% 

amylopectin, and 80% unmodified 

corn starch, with at least 90% of the 

particle weight less than U.S. sieve 

size 100. 

In all cases except the pregelatinized 

starch and the acacia, the binder was 

added both as an aqueous solution 

and in the dry state (in which the 

granulating liquid was simply water). 

The powders were charged into the 

granulator and premixed for 60 

seconds at high chopper speed and 

200 rpm main impeller setting to 

delump and homogeneously mix the 

contents. The binder solution or 

simply water was added to the 

granulation while the Fugi was 

operated at high chopper speed and 

200 main impeller speed until a 

proper wet mass was achieved. The 

resultant wet mass was passed 

through a CoMil (Quadro, Inc.) 

equipped with a square hole screen 

(2A-3750037/63), impeller (2A-1606- 

086L) operating at a speed of 2665 

rpm. The milled material was placed 

on trays and oven dried at 110°F to a 

final moisture endpoint of 1-2%. The 

moisture analysis was performed on 

a Mettler Infrared Drying Unit LP16-M 

(Mettler, Inc.). Upon drying, the 

granulation was dry sized via the 

same CoMil operating at the same 

speed setting and equipped with a 

round hole grater-type screen 

(2A 079G031/23120) and impeller 

(2A-1601-173). After dry sizing, the 

granulation was placed in a V-blender, 

where the lactose and disintegrant 

were added and mixed for 10 

minutes. Finally, the lubricant was 

added, and following an additional 2 

minutes of mixing, the final mixture 

was compressed at 1000, 2000, and 

3000 lbs. total compression force on 

an instrumented rotary tablet press.

Properties of the Granulation 

The complete granulations (including 

the disintegrant and lubricant) were 

characterized by a number of simple 

techniques. The apparent and tap 

densities of the granulations were 

determined using simple 

weight/volume measurements and a 

Vanderkamp tap density tester (500 

taps). From the apparent and tap 

densities, one can calculate the socalled 

compressibility index (I) given 

by the equation 

va 

I = [ 1 – x 100, 

vt 

] 

where va is the apparent density and 

vt is the tapped density. I can be used 

as a rough indication of the flow 

Table 1 

BINDERS IN SOLUTION 

Apparent Tap 

Density Density I(%) 

3% A15P LV 0.62 0.72 15 

6% A15P LV 0.59 0.70 16 

3% E5P LV 0.62 0.77 20 

6% E5P LV 0.61 0.75 19 

3% E15P LV 0.60 0.76 20 

6% E15P LV 0.62 0.71 12 

3% F4P LV 0.60 0.78 23 

6% F4P LV 0.58 0.78 24 

3% K3P LV 0.59 0.75 21 

6% K3P LV 0.57 0.74 24 

3% HPC-EF 0.59 0.73 19 

6% HPC-EF 0.62 0.76 18 

3% HPC-LF 0.63 0.81 22 

6% HPC-LF 0.62 0.76 18 

3% PVP (K29-32) 0.58 0.75 22 

6% PVP (K29-32) 0.57 0.74 22 

3% PVP (K90) 0.53 0.71 25 

6% PVP(K90) 0.51 0.70 26 

6% Acacia 0.57 0.74 23 

10% Pregel. starch 0.57 0.68 16 

properties of a material, where I 

values of less than 15% are indicative 

of good flow characteristics, while 

values above 25% are indicative of 

poor flow properties. The data for the 

acetaminophen granulations are 

presented in Table 1, in the case 

where the binder was applied from a 

solution, and in Table 2, where the 

binder was added in a dry state to 

the mixer-granulator followed by 

granulation with water. The densities 

are in units of g/cm 3 . 

In this acetaminophen model 

system, because of composition 

differences (3% binder/50% drug and 

6% binder/75% drug), comparisons 

are made only between similar 

binder levels. 

Table 2 

BINDERS DRY 

For the "solution" method, there were 

only small differences in the apparent 

densities between the binders except 

for the PVP (K90) binder, which 

exhibited a little lower density at both 

the 3% and 6% levels. The values for 

I varied from 15-25% at the 3% level, 

and from 12-26% at the 6% level, the 

highest values in both cases 

occurring with the PVP (K90). With 

the "dry" binder addition method, the 

densities at both binder 

concentrations tended to be very 

similar to those obtained by the 

solution method. The lowest densities 

and highest I values resulted from the 

use of PVP (K90) as a binder in this 

case as well. 

Apparent Tap 


3% A15P LV 0.64 0.77 17 

6% A15P LV 0.57 0.69 16 

3% E5P LV 0.56 0.74 24 

6% E5P LV 0.60 0.76 21 

3% E15P LV 0.65 0.76 14 

6% E15P LV 0.59 0.73 19 

3% F4P LV 0.61 0.78 22 

6% F4P LV 0.57 0.75 24 

6% K3P LV 0.59 0.75 22 

3% HPC-EF 0.61 0.74 18 

6% HPC-EF 0.56 0.73 24 

3% HPC-LF 0.58 0.70 17 

6% HPC-LF 0.57 0.72 20 

3% PVP (K29-32) 0.58 0.75 23 

6% PVP (K29-32) 0.54 0.70 22 

3% PVP (K90) 0.53 0.71 25 

6% PVP(K90) 0.51 0.70 27 

High Shear Granulation: Acetaminophen Model 

3

The particle size distribution (PSD) of 

the granulations were also 

determined. The amount of material 

retained on U.S. standard sieves of 

20, 40, 60, 80, 100, and 140 mesh 

was measured, using an ATM Sonic 

Sifter. A few typical particles size 

distributions are given by Figures 1-9 

and present data for both binder 

addition methods. 

Figure 1 compares granules with 3% 

and 6% METHOCEL A15P LV. While 

the differences are minimal, 6% 

binder "dry" had less material 

retained on the 40 mesh screen and 

slightly more on the remaining sieves. 

Bar graphs for METHOCEL E5P LV 

(Figure 2) and E15P LV (Figure 3) 

demonstrate the impact of molecular 

weight on PSD. The METHOCEL 

E15P LV (15 cps) generated a larger 

overall particle size than did the 

METHOCEL E5P LV (5 cps) binder. 

The particle size distributions for 

METHOCEL K3P LV and F4P LV 

were characterized by having slightly 

less material on 40 mesh and slightly 

more material on 60 mesh than the 

distribution given in Figure 3. 

4 High Shear Granulation: Acetaminophen Model


5

HPC-EF and -LF are depicted in 

Figures 4 and 5, respectively. The 

HPC-EF and HPC-LF showed fairly 

similar particle size distributions. PVP 

(K29-32) (Figure 6) and PVP (K90) 

(Figure 7) also exhibited somewhat 

similar PSD profiles. The K29-32 

grade shows a slightly greater 

amount of material retained on the 40 

mesh. Figure 8 is a comparison of the 

cellulosic polymer, METHOCEL 

A15P LV, and pregelatinized starch. 

The METHOCEL product produced 

substantially more 20 and 40 mesh 

granules than the pregelatinized 

starch. Finally, Figure 9 shows the 

PSD of acacia and METHOCEL 

A15P LV at 6%. The acacia produced 

a PSD that was comparable to the 

cellulosic binders. This distribution 

was very different from the extremely 

fine granulation obtained when acacia 

was used in fluid bed granulation. 


Tablet Physical Properties 

The hardness and friability of the 0.5 

inch tablets compressed at 2000 lbs. 

total compression force are given in 

Tables 3 and 4. For both 

measurements, 20 tablets were used; 

the reported friabilities are those after 

4 minutes in a Roche-type friabilator. 

Examining the data of Table 3, it is 

clear that good, hard tablets were 

produced, with the exception of those 

compressed from the acacia 

granulations; acacia has again been 

shown to be a relatively poor binder 

for this formulation. In most cases, the 

Table 3 


Hardness Friability 

(SCU) ± sd (%) 

3% A15P LV 17.1 ± 1.7 0.30 

6% A15P LV 16.2 ± 1.6 0.30 

3% E5P LV 19.9 ± 2.0 0.27 

6% E5P LV 16.8 ± 2.2 0.39 

3% E15P LV 13.3 ± 1.3 0.30 

6% E15P LV 17.1 ± 1.7 0.50 

3% F4P LV 19.0 ± 2.3 0.28 

6% F4P LV 15.9 ± 1.6 0.38 

3% K3P LV 20.6 ± 2.2 0.31 

6% K3P LV 19.3 ± 2.6 0.40 

3% HPC-EF 27.2 ± 3.7 0.24 

6% HPC-EF 15.8 ± 1.3 0.38 

3% HPC-LF 26.0 ± 3.1 0.28 

6% HPC-LF 19.1 ± 1.5 0.34 

3% PVP (K29-32) 20.9 ± 2.3 0.37 

6% PVP (K29-32) 20.5 ± 2.8 0.38 

3% PVP (K90) 23.5 ± 2.5 0.35 

6% PVP (K90) 24.3 ± 2.9 0.33 

6% Acacia 8.5 ± 1.6 capped 

10% Pregel. Starch 13.9 ± 2.6 capped 

tablets with 3% binder were harder 

than those with 6% binder; this is 

most likely due to the differences in 

the formulations and the inherent 

incompressiblity of acetaminophen. 

The exceptions to this observation 

were the METHOCEL E15P LV and 

PVP (K90) binders. At the 3% level, 

the hardest tablets were made from 

HPC-EF and -LF granulations, from 

the povidone granulations, and from 

the METHOCEL K3P LV granulation. 

Similarly, at the 6% level, the hardest 

tablets were those resulting from 

granulations prepared with the PVPs, 

METHOCEL K3P LV, and HPC-LF. 

Table 4 

BINDERS DRY 

Overall, very good friabilities were 

achieved by the solution addition 

method. The binders giving the lowest 

friability tablets were 3% HPC-EF, 3% 

METHOCEL E5P LV, 3% HPC-LF, 

and 3% METHOCEL F4P LV. (Note 

that these all were formulations 

containing 50% acetaminophen.) The 

tablets with the highest friability were 

the 6% acacia and 10% 

pregelatinized starch, which both 

capped. Tablets containing 6% 

METHOCEL E15P LV, which had a 

friability of only 0.50%, were the most 

friable of those that did not cap. 


(SCU) ± sd (%) 

3% A15P LV 21.5 ± 2.0 0.21 

6% A15P LV 14.3 ± 1.5 0.42 

3% E5P LV 20.6 ± 2.9 0.27 

6% E5P LV 17.1 ± 2.0 0.38 

3% E15P LV 14.3 ± 1.6 0.36 

6% E15P LV 14.6 ± 1.8 0.45 

3% F4P LV 24.8 ± 2.0 0.24 

6% F4P LV 16.6 ± 2.1 0.32 

6% K3P LV 19.9 ± 2.3 0.39 

3% HPC-EF 20.6 ± 1.6 0.24 

6% HPC-EF 22.6 ± 1.6 0.24 

3% HPC-LF 23.5 ± 1.3 0.25 

6% HPC-LF 23.3 ± 1.1 0.27 

3% PVP (K29-32) 22.1 ± 2.6 0.35 

6% PVP (K29-32) 24.0 ± 2.2 0.35 

3% PVP (K90) 28.3 ± 2.3 0.24 

6% PVP (K90) 23.4 ± 3.1 0.34 


7

The dry binder addition method (see 

Table 4) also produced tablets with 

excellent hardness overall. The 

results were somewhat mixed 

concerning which binder level gave 

the harder tablets (in 5 of the 9 

cases, the tablets at 3% binder were 

harder). At the 3% binder 

concentration, the hardest tablets 

were made with the PVP (K90), 

METHOCEL F4P LV, HPC-LF, and 

PVP (K29-32) materials. Note that 

these are essentially the same 

binders and essentially the same 

hardness values that were obtained 

by the solution addition method. This 

was a somewhat surprising result, 

since it is generally considered that 

binders are more effective when they 

are placed in solution prior to the 

granulation operation. Similarly, at the 

6% level, the best binders were PVP 

(K29-32), PVP (K90) ≈ HPC-LF, HPC- 

EF, and METHOCEL K3P LV. In 

terms of friability, the lowest losses 

were predominantly from tablets 

containing 3% binders, namely 3% 

METHOCEL A15P LV, 3% 

METHOCEL F4P LV ≈ 3% and 6% 

HPC-EF, and 3% PVP (K90). The 

highest friabilities, which were still 

less than 0.50%, resulted from the 

use of 6% METHOCEL E15P LV 

and A15P LV. 

The statistical significance of 

differences in the two binder addition 

methods was examined. Due to the 

differences in the formulations, only 

comparisons between 3% 

solution/3% dry and 6% solution/6% 

dry were made. Of the 17 sets of data 

(results of the 3% METHOCEL 

K3P LV excluded), the dry binder 

addition method was statistically 

harder in 7 cases: 3% METHOCEL 

A15P LV, E15P LV, and F4P LV, 6% 

HPC-EF and -LF, 6% PVP (K29-32), 

and 3% PVP (K90). The solution 

binder addition method produced 

8 High Shear Granulation: Acetaminophen Model 

statistically harder tablets in 4 cases: 

6% METHOCEL A15P LV and 

E15P LV, and 3% HPC-EF and -LF. 

The remaining 6 cases showed no 

statistically significant difference in 

hardness between the binder addition 

methods. 

As mentioned on page 3, the 

formulations were compressed at 

1000, 2000, and 3000 lbs. total 

compression force. For both binder 

addition methods, tablet hardness 

increased with force with very few 

exceptions, those being 6% acacia 

(hardness was essentially constant), 

10% pregelatinized starch (tablets 

compressed at 3000 lbs. were about 

5 SCU softer than those compressed 

at 2000 lbs.), and the 3% K3P LV 

(solution) and 3% PVP (K29-32) (dry) 

tablets (which had slightly softer 

tablets produced at 3000 lbs. 

compression compared to 2000 lbs. 

compression). The friabilities were 

measured at 2, 4, and 6 minutes at 

each force. At a given force, the 

percent weight loss naturally 

increased with time. At a given time 

(e.g., 4 minutes), there was the 

expected decrease in friability as the 

compression force, and thereby the 

tablet hardness increased. There was 

a proportionally smaller decrease in 

friability between 2000 lbs. and 3000 

lbs. than there was between 1000 lbs. 

and 2000 lbs. There were a few 

cases in which the friability of tablets 

compressed at 3000 lbs. was greater 

than those compressed at 2000 lbs. 

despite the fact that the one 

compressed at the higher force had a 

higher hardness. One example of this 

behavior is 3% PVP (K29-32) 

(solution), where the hardness and 

friability at 2000 lbs. compression 

force were 20.9 SCU and 0.37%, 

while the hardness and friability at 

3000 lbs. compression force were 

21.9 SCU and 0.75%, respectively. 

The weight and thickness variation of 

the tablets for each formulation and 

binder addition method were also 

measured (n = 20). The thickness 

variation in all cases was excellent, 

varying from a low of 0.12% relative 

standard deviation (RSD, equal to the 

standard deviation/mean x 100) for 

the 6% METHOCEL F4P LV (solution 

method) formulation, to 0.84% RSD 

for the 3% HPC-EF (solution) 

formula; the range of thickness 

variation for the dry addition method 

was 0.18–0.67% RSD. Similarly, the 

weight variation was very good for the 

majority of the formulations, varying 

from 0.47% RSD for the 3% 

METHOCEL F4P LV (dry) formula, to 

1.39% RSD for the 6% METHOCEL 

E15P LV (solution) case. The binders 

that performed well in one of the 

binder addition methods tended to 

perform well in the other addition 

method as well. 

Tablet Dissolution Properties 

The in vitro dissolution properties 

varied both as a function of the 

nature of the binder and the amount 

of binder in the formulation, but 

relatively little as a function of the 

binder addition method. The USP 

dissolution conditions for 

acetaminophen tablets were used 

(Type 2 apparatus at 50 rpm, 900 mL 

of pH 5.8 phosphate buffer). The 

time to reach 80% dissolved is 

designated by t 80% . A number of 

these dissolutions are presented in 

Figures 10-18.


9

The METHOCEL products are 

compared in Figures 10-13. Figure 10 

(solution binder addition method, 3% 

level) shows that the polymers all 

have t 80% of about 20 minutes with 

the exception of METHOCEL 

E15P LV, which had a t 80% of 

approximately 7 minutes longer, but 

which still passed the USP criteria. 

The dry binder addition method 

(Figure 11) shows more 

differentiation. METHOCEL E5P LV 

clearly was the fastest with a 

dissolution time of 15 minutes; 

METHOCEL A15P LV and F4P LV 

were next, with about the same time 

as was observed with the solution 

addition method. METHOCEL 

E15P LV was just 2-3 minutes slower 

than the solution method, with a t 80% 

right at the USP limit. (The 

dissolution results for the 

METHOCEL K3P LV product were 

anomolous and are not included 

pending repetition of the granulation.) 

Increasing the binder content to 6% 

resulted in the typical lengthening of 

the time for dissolution. In Figure 12, 

the curve for METHOCEL A15P LV is 

virtually unchanged from the 3% 

solution case. However, the remaining 

METHOCEL products are shifted to 

longer times, with METHOCEL 

F4P LV, E5P LV, and E15P LV all at 

t 80% ≈ 35 minutes, and with 

METHOCEL K3P LV at nearly 45 

minutes. In the dry binder addition 

case (Figure 13), slightly different 

profiles were obtained. METHOCEL 

A15P LV, E5P LV, and F4P LV form 

one group with t 80% ≈ 25 minutes, 

while METHOCEL E15P LV and 

K3P LV both have t 80% greater than 

the USP limit, but with different 

curve shapes. 


The remaining figures in this section 

are comparisons of METHOCEL 

polymers with the HPC polymers, 

with the PVP products, and with 

acacia and pregelatinized starch. 

Figure 14 illustrates that using the dry 

addition method at 3% results in 

virtually no difference between the 

two viscosity grades of HPC, both 

having 80% drug dissolution in 25 

minutes. If the same binders at the 

same level utilized the "solution" 

method, the HPC-LF gave the same 

release profile, while the HPC-EF 

gave a slightly longer (t 80% of 31 

minutes) profile. Turning to the 6% 

level, Figure 15 shows the results of 

the solution method, with the HPC-EF 

and HPC-LF curves indistinguishable 

and with the time for 80% release at 

the USP limit. Using the dry addition 

method was not different in terms of 

the release profile for HPC-LF, but 

once again the EF grade had a 

slightly longer release with t 80% ≈ 

32 minutes. 

Figures 16 and 17 compare two 

METHOCEL products with 

polyvinylpyrrolidone. The dry method 

at 3% (Fig. 16) shows that 

METHOCEL E5P LV and PVP 

(K29-32) are equivalent in dissolution 

behavior, with the K90 grade just 

slightly longer but well within the 

specified time period. The solution 

addition method gave results 

essentially the same. Turning to the 

6% level (Fig. 17), METHOCEL A15P 

LV and PVP (K29-32) are practically 

equivalent, with the K90 polymer 

again slightly longer, now just within 

the specification. Using the dry 

method at the 6% binder level 

produced dissolution curves where 

the METHOCEL A15P LV was just 

slightly longer than the PVP (K90) 

and which was at the 30 minute limit. 

Lastly, a comparison of acacia, a 

pregelatinized starch, and 

METHOCEL A15P LV is illustrated in 

Figure 18. Here it appears that the 

derivatized starch is also acting as a 

disintegrant, giving the fastest drug 

release of all the binders evaluated. 

Acacia, despite giving quite soft and 

very friable tablets, nevertheless gave 

a somewhat unexpectedly long time 

for drug dissolution. 

Conclusions: Acetaminophen 

Model Formulation 

Granulations with acceptable 

densities and particle size 

distributions were obtained when 2 

acetaminophen formulations having 

3% and 6% binder levels were 

prepared in a high shear mixergranulator. 

There were only minor 

differences in these properties 

between granulations prepared by 

adding the binder as an aqueous 

solution or by adding the binder in a 

dry state to the active followed by 

granulation with water. The cellulosic 

polymers (METHOCEL and HPC 

products) tended to give granulations 

with a higher proportion of granules 

of 20-60 mesh, whereas the PVP 

binders produced somewhat more 

material in the 80-140 mesh region. 

With the given milling conditions, 

about 10% fines (


2: Vitamin C 


In this section of the study, the 

following formulations were used: 

75% Vitamin C 

19.5% Lactose 

3.0% Binder 



75% Vitamin C 

16.5% Lactose 

6.0% Binder 



Note that these formulations are 

identical with those used in the earlier 

fluid bed study. The formulation at 6% 

binder is identical with the 

acetaminophen model presented in 

the previous section except for the 

active ingredient; this will facilitate 

comparisons on the effects of the 

drug on the relative performance of 

the various binders. Here, Roche fine 

powder was utilized; all of the 

remaining components were the 

same as described on page 1. 

The same binders were evaluated as 

described on pages 1 and 2. The 

processing of the Vitamin C 

formulations was different from that 

used in the acetaminophen 

formulations. In this case, the active 

and the lactose (and the binder in 

those experiments where the binder 

was not placed in solution) were 

placed in the Fuji and mixed for 60 

seconds at 200 rpm, chopper speed 

high. The granulating liquid was then 

added by means of the attached 

funnel, and mixing was continued at 

200 rpm, chopper speed high until a 

proper wet mass was achieved. The 

resultant wet mass was discharged 

12 High Shear Granulation: Vitamin C Model 

and wet milled in a CoMil (model 

197S) using a square hole screen 

(2A-3750037/63) and an impeller 

(2A-1607-086L) at 2665 rpm. This 

combination of screen and impeller 

worked very well for wet milling for all 

binders with the exception of PVP 

(K90), which had very poor wet 

milling properties. After tray drying at 

110°F overnight to a moisture content 

of < 1%, the granulations were then 

dry milled with the CoMil using a 

round hole grater-type screen 

(2A-079G031/23120) and impeller 

(2A-1601-173), again at 2665 rpm. 

The disintegrant and lubricant were 

added to the Vitamin C granulations 

in a twin shelled blender and mixed 

for 2 minutes. The complete mixtures 

then were compressed at 1000, 

2000, and 3000 lbs. total 

compression force. 


The apparent and tap densities and 

the compressibility indices I of the 

Vitamin C granulations were 

determined using the procedures 

described on page 3, and are given 

in Tables 5 and 6. As in Tables 1 and 

2, the apparent and tapped densities 

are in g/cm 3 . 

In this model system using the Fuji 

VG-25P, the densities increased as 

the binder level was increased from 

3% to 6%. For the granulations 

produced with the METHOCEL and 

the HPC products, the increase was 

small. However, the PVP binders 

(especially the K29-32 grade) showed 

a slightly greater increase in density 

with an increase in binder 

concentration. Once again, this 

behavior can be understood by 

examining the particle size 

distributions that are presented in the 

figures that follow. Comparing the 

results where the binder was 

prehydrated with those where it was 

not shows that in the large majority of 

cases, higher densities were obtained 

by the "dry" binder addition method; 

there was no trend in the I values. 

Finally, note that for the comparable 

6% binder levels, the compressibility 

indices for Vitamin C granulations are 

smaller (i.e., indicative of better flow) 

than those obtained for the 

acetaminophen granulations. It was 

observed during tablet compression 

that the Vitamin C granulations had 

very good flow behavior. 

Particle size distributions (PSD), 

determined as described on page 4, 

are given in Figures 19-27. The 

figures show results for both of the 

binder addition methods. Figure 19 

compares granule sizes with 3% and 

6% METHOCEL A15P LV. Increasing 

the amount of binder gives small 

increases in the quantity of material 

≤ 60 mesh, but a decrease in the 

amount of 40 mesh material. The 6% 

binder level (solution) gave the finest 

overall PSD. Granulation with 

METHOCEL E5P LV (Figure 20) 

shows a PSD that is typical of 

virtually all of the other binders 

tested. Increasing the amount of 

binder gave an increase in granules 

on 20 mesh, little change at 40 mesh 

and < 140 mesh, and small 

decreases in all other fractions. The 

amount of material 80-140 mesh is 

quite small. For the most part, the 

binder addition method had relatively 

little effect on the PSD at both the 3% 

and 6% levels. This general pattern 

was true for METHOCEL E15P LV, 

METHOCEL F4P LV, METHOCEL 

K3P LV (Figure 21), HPC-EF, HPC- 

LF (Figure 22), and PVP (K29-32). 

In Figure 23, it is seen that the PVP 

(K90) granulations of Vitamin C 

followed a somewhat different pattern 

than the other binders just described. 

With this binder, there was less 

material on the 40 mesh screen, 

somewhat more material on 80 mesh,

and substantially more material


The hardness and friability of the 0.5 

inch tablets compressed at 2000 lbs. 

total compression force are given in 

Table 7 for the wet binder addition 

method, and in Table 8 for the dry 

addition method. For all 

measurements, 20 tablets were used; 

the reported friabilities are those after 

4 minutes in a Roche-type friabilator. 

The hardness and friability data can 

be analyzed in a number of ways. For 

the most part, tablets of satisfactory 

hardness and friability were obtained 

with each of the binders. From an 

inspection of Table 7, the top 5 

binders in terms of hardness are 6% 

HPC-LF, 6% METHOCEL 

E15P LV/6% PVP (K90), 3% PVP 

(K90), and 6% HPC-EF. Similarly, the 

Table 7 



(SCU) ± sd (%) 

3% A15P LV 11.9 ± 0.7 0.42 

6% A15P LV 9.8 ± 0.9 0.45 

3% E5P LV 9.5 ± 0.9 0.63 

6% E5P LV 10.2 ± 1.0 0.53 

3% E15P LV 9.2 ± 1.0 0.54 

6% E15P LV 13.0 ± 1.6 0.41 

3% F4P LV 10.4 ± 0.7 0.58 

6% F4P LV 11.2 ± 0.9 0.45 

3% K3P LV 10.2 ± 1.0 0.66 

6% K3P LV 10.7 ± 1.2 0.48 

3% HPC-EF 9.8 ± 0.9 0.71 

6% HPC-EF 12.1 ± 1.3 0.45 

3% HPC-LF 10.2 ± 1.0 0.71 

6% HPC-LF 13.4 ± 1.2 0.41 

3% PVP (K29-32) 8.0 ± 1.5 1.06 

6% PVP (K29-32) 11.5 ± 1.9 0.91 

3% PVP (K90) 12.6 ± 1.6 0.56 

6% PVP (K90) 13.0 ± 1.9 0.53 

6% Acacia 6.0 ± 1.2 2.50 

10% Pregel. Starch 11.1 ± 1.5 0.59 


best binders in terms of friability are 

6% METHOCEL E15P LV/6% HPC- 

LF, 3% METHOCEL A15P LV, and 

6% METHOCEL A15P LV, 6% 

METHOCEL F4P LV/6% HPC-EF. A 

similar inspection of Table 8 (dry 

binder addition method) shows that 

the top 5 binders in terms of 

hardness are 6% PVP (K90), 6% 

HPC-EF, 6% HPC-LF, and 6% 

METHOCEL E5P LV/3% PVP (K90). 

The least friable tablets in the "dry" 

case are 6% A15P LV, 6% HPC-LF, 

6% HPC-EF, 3% METHOCEL E15P 

LV, and 6% METHOCEL E15P LV. 

Comparing the two sets of rankings 

for both hardness and friability, it is 

clear that the HPC products and PVP 

(K90) consistently gave the hardest 

tablets, while the HPC products, 

METHOCEL E15P LV, and 

Table 8 

BINDERS DRY 

METHOCEL A15P LV were the 

binders that most consistently gave 

the lowest friabilities. If Tables 7 and 8 

are examined for the 3 binders that 

gave the lowest values for hardness 

and the highest values for friability, it 

can be seen that PVP (K29-32) falls 

into both categories for both binder 

addition methods. The worst binder in 

those granulations prepared from 

binder solutions was acacia, with the 

3% METHOCEL E15P LV tablets also 

being relatively soft. For the dry 

binder addition method, the 6% 

METHOCEL E15P LV tablets, along 

with the PVP (K29-32) tablets, 

among the softest. It is once again 

observed that there is not a simple 

inverse relationship between 

hardness and friability. 


(SCU) ± sd (%) 

3% A15P LV 10.3 ± 1.0 0.53 

6% A15P LV 11.1 ± 0.7 0.41 

3% E5P LV 10.1 ± 0.9 0.57 

6% E5P LV 11.7 ± 1.3 0.54 

3% E15P LV 9.5 ± 0.7 0.48 

6% E15P LV 8.6 ± 0.7 0.49 

3% F4P LV 11.0 ± 1.3 0.55 

6% F4P LV 10.5 ± 1.0 0.56 

3% K3P LV 9.6 ± 1.0 0.68 

6% K3P LV 11.1 ± 1.3 0.54 

3% HPC-EF 9.7 ± 1.0 0.67 

6% HPC-EF 13.1 ± 1.4 0.45 

3% HPC-LF 10.7 ± 1.0 0.72 

6% HPC-LF 12.4 ± 1.2 0.44 

3% PVP (K29-32) 7.2 ± 0.9 0.85 

6% PVP (K29-32) 8.6 ± 0.9 0.79 

3% PVP (K90) 11.7 ± 1.3 0.61 

6% PVP (K90) 13.8 ± 1.9 0.49

It was observed previously that the 

hardest and least friable tablets were 

those which used 6% binder (with the 

exception of the 3% PVP (K90) 

tablets). Of the 18 sets of binders and 

binder addition methods in which the 

level was examined at both 3% and 

6%, 13 sets showed statistically 

significant increases in hardness, 2 

sets showed statistically significant 

decreases in hardness (METHOCEL 

A15P LV (solution) and METHOCEL 

E15P LV (dry), and 3 sets showed no 

statistically significant change 

(METHOCEL K3P LV and PVP (K90) 

in solution, and METHOCEL F4P LV 

in the dry state) with an increase in 

binder level. 

The significance of the binder 

addition method also may be 

examined. At the 3% binder level, 

only METHOCEL A15P LV, where the 

solution method gave harder tablets, 

and HPC-LF, where the dry method 

gave harder tablets, were statistically 

different. The METHOCEL E5P LV 

binder was "borderline" for p > 0.05. 

The remaining 6 binders did not 

exhibit a statistically significant 

difference in hardness due to the 

binder addition method. A much 

different case exists for the 6% binder 

level. Here, tablet hardnesses were 

different between the methods for 7 

of the 9 binders. Granulations 

prepared when the binder was 

prehydrated gave harder tablets for 

METHOCEL E15P LV, METHOCEL 

F4P LV, HPC-LF, and PVP (K29-32). 

On the other hand, granulations 

prepared where the dry binder was 

added to the drug/lactose mix gave 

harder tablets for METHOCEL A15P 

LV, METHOCEL E5P LV, and HPC- 

EF. There was no statistically 

significant difference in hardness with 

binder addition methods for 

METHOCEL K3P LV and PVP (K90) 

at the 6% level. The differences in 

friability as a function of binder 

addition method were very minor. 

High Shear Granulation: Vitamin C Model 

15

As stated on page 12, granulations 

were compressed at 1000, 2000, and 

3000 lbs. In addition, the friabilities 

were measured at 2 minutes, 4 

minutes, and 6 minutes. In general, 

tablet hardness increased with 

compression force between 1000 and 

2000 lbs., but then decreased 

between 2000 and 3000 lbs. for 

tablets containing 3% binder. At the 

6% binder level, the tablets continued 

to increase in hardness with 

compression force. For a given 

binder, binder level, and compression 

force, friability naturally increased with 

time. Interestingly, in some of the 

cases where the hardness at 3000 

lbs. > hardness at 2000 lbs., the 

friabilities of tablets compressed at 

2000 lbs. were less than those 

compressed at 3000 lbs. at each 

time interval. 

The ranges of thickness and weight 

variation were relatively narrow and 

very similar for the two binder 

addition methods. For weight 

variation, the percent relative 

standard deviations (%RSD) varied 

from 0.59% for 3% HPC-LF to 1.46% 

for 6% HPC-EF using the solution 

method, and from 0.56% RSD for 

METHOCEL E15P LV to 1.47% RSD 

for 3% HPC-EF using the dry 

method. The thickness variation 

ranges are as follows: solution 

method, 0.25% RSD for 6% 

METHOCEL E15P LV to 0.77% RSD 

for 3% METHOCEL A15P LV; dry 

method, 0.22% RSD for 6% PVP 

(K29-32) to 0.79% RSD for 

METHOCEL K3P LV. 

16 High Shear Granulation: Vitamin C Model


17

Tablet Dissolution Properties 

Just as was the case with the in vitro 

dissolution curves for acetaminophen 

shown earlier, the release profiles 

varied both as a function of the type 

and amount of binder in the 

formulation. A USP Type 2 apparatus 

at 50 rpm and 900 mL of pH 7.0 

phosphate buffer were used. A 

number of these dissolutions are 

given in Figures 28-34. There is no 

"USP specification" for Vitamin C 

tablets, but in order to facilitate 

comparisons with the acetaminophen 

data, limits of 80% released in 30 

minutes have been added. The time 

to reach 80% dissolved is designated 

by t 80% . 

The five METHOCEL products 

evaluated in this study are compared 

in Figures 28-31. In Figure 28, 

dissolution behavior for a binder level 

of 3% using the solution binder 

addition method is shown. Here, the 

METHOCEL K3P LV gave the fastest 

dissolution (t 80% of 12.4 min), 

followed closely by the E5P LV and 

F4P LV grades at about 14 min. The 

METHOCEL E15P LV, by 

comparison, had t 80% of 19.4 min, 

clearly showing the effect of 

molecular weight/viscosity of 

hypromellose 2910 in this formulation. 

The methylcellulose product, 

METHOCEL A15P LV, produced the 

tablets that had the longest dissolution 

time, about 8 minutes longer than 

that of METHOCEL E15P LV. If one 

compares Figure 28 with Figure 29 

(which illustrates the same binders at 

the same level, except that the dry 

binder addition method was used), it 

is clear that binder addition method is 

insignificant in this aspect of the 

binder performance as well. The rank 

order of the binders remains the 

same, with the t 80% times just slightly 

longer for the dry method. Note that 

the order present here is the same as 

that observed for Vitamin C in the 

fluid bed granulation study (Part I), 

but once again shows a reversal of 

the relative performance of the 

18 High Shear Granulation: Vitamin C Model

METHOCEL A and METHOCEL K 

products compared to the 

acetaminophen model in both this 

study and the earlier fluid bed study. 

At the 6% binder level (Figures 30- 

31), the order of tablet dissolution is 

the same as at 3% binder. The 

METHOCEL K3P LV, E5P LV, and 

F4P LV all gave very similar 

dissolution profiles (t 80% ≈ 21 min) 

when the solution addition method 

was used, followed by METHOCEL 

E15P LV about 10 minutes later. The 

METHOCEL A15P LV gave an almost 

sustained-release dissolution, with 

t 80% not reached until 82 minutes. 

Turning to the dry addition method 

(Figure 31), the order and general 

performance of the binders was not 

different from that in the solution 

method. There was a slightly larger 

spread in the dissolution behavior of 

the K3P LV, E5P LV, and F4P LV 

products, and the A15P LV tablets 

achieved 80% release in about 72 

minutes. As has been consistently 

observed throughout these 

granulation studies, increasing the 

amount of binder gives longer tablet 

dissolution times and accentuates 

differences in the relative 

performance of the binders. 

The remaining figures in this section 

compare the dissolution behavior of 

METHOCEL K3P LV (which was 

essentially equivalent to that of the 

METHOCEL E5P LV and F4P LV 

products in this formulation) with the 

HPC products and the other 

competitive binders. In Figure 32, it is 

seen that there is practically no 

difference in the dissolution profiles 

for METHOCEL K3P LV, HPC-EF, 

and HPC-LF, regardless of the binder 

addition method, at the 3% binder 

level. At the 6% level (not shown), the 

two HPC products had almost 

identical dissolution profiles, again 

irrespective of the way in which the 

binder was placed in the granulations; 

the t 80% values were 27-29.5 minutes 

vs. 19 minutes for the METHOCEL 

K3P LV. 


19

The data for METHOCEL K3P LV 

and the two viscosity grades of PVP 

at 3%w/w binder are presented in 

Figure 33, which illustrates the 

unimportance of the method in which 

the binder was added. The dissolution 

of tablets containing PVP (K90) was 

about 6 minutes slower than those 

prepared with the METHOCEL 

cellulose ether, which was in turn 

about 3-4 minutes slower than that of 

tablets prepared with PVP (K29-32). 

The same general trends were 

evident at the 6% binder level, where 

the t 80% values were approximately 

15 minutes, 19 minutes, and 29 

minutes for PVP (K29-32), 

METHOCEL K3P LV, and PVP 

(K90), respectively. 

The final figure, Figure 34, compares 

the performance of METHOCEL 

K3P LV and acacia at 6% with that of 

the pregelatinized starch at the 10% 

level. (Recall that the acacia and 

pregelatinized starch were only 

utilized in the solution binder addition 

method.) The acacia tablets dissolved 

relatively rapidly, while the profile of 

tablets prepared with the modified 

starch was virtually indistinguishable 

from that of tablets made with 

METHOCEL K3P LV. 


Conclusions: Vitamin C Model 

Formulation 

The METHOCEL and HPC polymers 

formed good granules with good 

milling properties, both when added 

as aqueous solutions to the high 

shear mixer-granulator, or added as 

a dry powder to the unit followed by 

granulation with water. Good 

densities, comparable with the 

competitive binders, were obtained 

with both binder addition methods. 

The particle size distributions 

appeared to be completely 

satisfactory. The majority of the 

granules were retained on the 20, 

40, and 60 mesh screens; less than 

20% of the material was finer than 

140 mesh (with the exception of 

PVP (K90). 

Increasing the amount of binder in 

the granulations caused an increase 

in hardness in 72% of the cases 

studied, caused a decrease in 11%, 

and caused no change in 17%. At the 

3% level, most of the binders gave no 

difference in hardness as a function 

of binder addition method, whereas at 

the 6% level there generally was a 

difference, but there was no 

consistency in which method gave 

the harder tablets. The binders that 

consistently gave the hardest tablets 

were HPC-LF, HPC-EF, and PVP 

(K90). The binders that routinely gave 

the least friable tablets were 

METHOCEL E15P LV, HPC-LF, and 

METHOCEL A15P LV. Povidone 

(K29-32) gave tablets that were 

relatively soft and friable. 

Comparing the cellulosic binders at 

the 3% level shows very little 

difference in the dissolution 

performance between METHOCEL 

K3P LV, F4P LV, E5P LV, and both of 

the HPC products. METHOCEL E15P 

LV produced tablets with a slightly 

longer dissolution profile, while the 

profile for tablets utilizing METHOCEL 

A15P LV was somewhat longer yet. 

At the 6% binder level, METHOCEL 

K3P LV, F4P LV, and E5P LV are 

recommended; the HPC products and 

METHOCEL E15P LV gave t 80% of 

about 30 minutes. The methylcellulose 

product produced tablets with 

dissolution times of over 1 hour. 

There were only minor differences 

resulting from the dry and solution 

binder addition methods. The PVP 

(K29-32) and acacia gave tablets with 

the most rapid dissolution, but the 

tablets prepared from both were 

relatively soft and friable. Tablets 

containing a pregelatinized starch 

used at the 10% level had essentially 

the same dissolution behavior as 

tablets containing METHOCEL 

K3P LV at 6%.


3: Methazolamide 


In this final section of the study, the 

following formulations were used: 

7.1 % Methazolamide 

87.4% Filler 

3.0% Binder 



7.1% Methazolamide 

84.4% Filler 

6.0% Binder 



This formulation, identical with that 

used in the fluid bed granulation 

study, is significantly different than the 

previous two. Methazolamide is a 

thiadiazoline sulfonamide which acts 

as a carbonic anhydrase inhibitor and 

is used in the treatment of glaucoma; 

Table 9 

BINDERS DRY 

Apparent Tap 


3% A15P LV 0.78 0.93 16 

6% A15P LV 0.72 0.89 19 

3% E5P LV 0.83 0.95 13 

6% E5P LV 0.85 0.97 12 

3% E15P LV 0.79 0.95 16 

6% E15P LV 0.76 0.92 17 

3% F4P LV 0.82 0.93 11 

6% F4P LV 0.86 0.99 13 

3% K3P LV 0.80 0.97 17 

6% K3P LV 0.85 0.97 12 

3% HPC-EF 0.79 0.89 11 

6% HPC-EF 0.79 0.89 11 

3% HPC-LF 0.73 0.90 12 

6% HPC-LF 0.82 0.93 12 

3% PVP (K29-32) 0.86 0.98 12 

6% PVP (K29-32) 0.85 1.02 16 

3% PVP (K90) 0.78 0.93 16 

6% PVP (K90) 0.81 0.99 19 

normal dosage is 50 mg. The 

compound is slightly soluble in water, 

and is therefore a good model for a 

low dose, low solubility drug. 

With this model formulation, only the 

dry binder addition method was 

utilized due to the equivalence of the 

two methods in the acetaminophen 

and Vitamin C models. The filler used 

in these formulations was a 50/50 

w/w mixture of Fast-Flo Lactose 316 

from Foremost Whey Products and 

terra alba (hydrous CaSO 4 , N.F.) from 

U. S. Gypsum. The granulations were 

prepared by adding the appropriate 

amount of filler, dry binder powder, 

and drug to the Fuji and mixing for 60 

seconds at 200 rpm, chopper speed 

high. The granulating liquid (which for 

this model system was water only) 

was then added by means of the 

attached funnel, and mixing was 

continued at 200 rpm, chopper speed 

high until a proper wet mass was 

achieved. The granulate was 

discharged and wet milled in a CoMil 

(model 197S) using a square hole 

Table 10 

BINDERS DRY 

screen (2A-3750037/63) and an 

impeller (2A-1607-086L) at 2665 rpm. 

This combination of screen and 

impeller worked very well for wet 

milling for all binders with the 

exception of PVP (K90), which had 

very poor wet milling properties. After 

tray drying at 110°F overnight to a 

moisture content of < 1%, the 

granulations were then dry milled with 

the CoMil using a round hole gratertype 

screen (2A-079G031/23120) and 

impeller (2A-1601-173), again at 

2665 rpm. The addition of disintegrant 

and lubricant were performed as 

described on page 2. 

All of the binders used in the previous 

two model systems were evaluated, 

with the exception of acacia and 

pregelatinized starch, which have 

tended to be effective only when 

placed in solution. 

The granulations were compressed 

into 0.5 inch tablets (700 mg) at 

1000, 2000, and 3000 lbs. total 

compression force. 


(SCU) ± sd (%) 

3% A15P LV 8.2 ± 1.2 0.29 

6% A15P LV 8.0 ± 0.6 0.32 

3% E5P LV 7.5 ± 0.4 0.38 

6% E5P LV 8.0 ± 0.7 0.31 

3% E15P LV 6.1 ± 0.7 0.44 

6% E15P LV 5.6 ± 0.7 0.39 

3% F4P LV 8.3 ± 0.6 0.35 

6% F4P LV 7.5 ± 1.0 0.33 

3% K3P LV 18.6 ± 1.8 0.25 

6% K3P LV 12.6 ± 0.9 0.24 

3% HPC-EF 9.2 ± 0.9 0.27 

6% HPC-EF 13.0 ± 0.7 0.20 

3% HPC-LF 9.5 ± 0.7 0.26 

6% HPC-LF 12.1 ± 0.6 0.28 

3% PVP (K29-32) 13.3 ± 1.0 0.32 

6% PVP (K29-32) 18.1 ± 1.6 0.33 

3% PVP (K90) 21.8 ± 2.6 0.20 

6% PVP (K90) 25.9 ± 2.8 0.24 

High Shear Granulation: Methazolamide Model 

21


The apparent density, tapped density, 

and compressibility index of each 

granulation was determined as 

before, and is given in Table 9 

(densities in g/cm 3 ). 

There is no clear trend in either the 

apparent or tap densities as a 

function of binder concentration. 

Despite having > 40% CaSO 4 , the 

granulations prepared here did not 

have densities that were appreciably 

greater than those of the 

acetaminophen or Vitamin C 

granulations. The values for the 

compressibility index I also varied 

over a narrow range; all of the 

granulations appeared to flow well on 

the tablet press. 

The particle size distributions of the 

five METHOCEL products break into 

3 groups. The first is comprised of 

METHOCEL A15P LV (Figure 35) 

and E15P LV. These have relatively 

little material on 20 mesh, and about 

30% on both the 40 mesh and in the 

fines at the 3% binder level. On the 

60-140 mesh screens the quantity of 

granules is nearly the same at the 3% 

and 6% binder levels; only in the fines 

is the amount at 6% significantly 

greater than the amount at 3%. There 

is no shift to larger granules as the 

amount of binder is increased. 

The second group is made up of 

METHOCEL E5P LV (Figure 36) and 

F4P LV. Here, roughly 5% of the 

material is present > 20 mesh, about 

15% is present as fines, and relatively 

little material is present 80-140 mesh; 

these amounts do not appreciably 

change with binder level. At the 3% 

binder level, > 40% is retained on the 

40 mesh screen; increasing binder 

causes a decrease at this granule 

size and a large increase on the 60 

mesh screen. 

METHOCEL K3P LV was somewhat 

unique. This binder (shown in Figure 

37) produced more of the largest 

granules than the other METHOCEL 

22 High Shear Granulation: Methazolamide Model

products; the amount of fines was 

about 25% at the lower binder level. 

Increasing binder causes a small 

decrease in the 20 mesh material, 

but a significant increase in the 40 

and 60 mesh granules at the expense 

of the fines. In other words, 

METHOCEL K3P LV showed a 

somewhat more classical shift in PSD 

with binder level. 

The PSD of the HPC products were 

very similar (Figure 38). Here, 

increasing binder concentration 

caused a decrease in the quantity of 

granules < 80 mesh, but gave a very 

large fraction that was retained on 40 

mesh screen. 

The performance of PVP (K29-32) 

(Figure 39) was unusual in that upon 

increasing the binder level, all of the 

sieve fractions except that retained on 

20 mesh and the fines remained the 

same. At 6%, there was less material 

at 20 mesh and more fines than at 

3% binder. The PVP (K90) also 

showed an increase in fines at the 

higher binder level. Finally, Figure 40 

compares the PSD for METHOCEL 

K3P LV and PVP (K29-32), two 

binders that will be further compared 

later in this section. 


The hardness and friability values of 

the methazolamide tablets 

compressed at 2000 lbs. total 

compression force are given in Table 

10. For both measurements, 20 

tablets were used; the reported 

friabilities are those after 4 minutes in 

a Roche friabilator. 

Acceptably hard tablets were 

produced with all binders, with the 

possible exception of METHOCEL 

E15P LV, having values ≈ 6 SCU (but 

note the % friability). For the 

METHOCEL products, the hardest 

tablets by a fairly wide margin were 

produced when METHOCEL K3P LV 

at 3% was utilized. In fact, for 3 of the 

5, the tablets at 3% were softer from 

a statistical standpoint; only for 


23

METHOCEL E5P LV was there a 

statistically significant increase in 

hardness at the higher binder level. 

There was no statistically significant 

difference in the hardness for 3% and 

6% levels of METHOCEL A15P LV. 

The HPC products did give slightly 

harder tablets relative to the 

METHOCEL products, with the 

exception of METHOCEL K3P LV as 

noted above. With 

hydroxypropylcellulose, hardness did 

increase with binder concentration. 

Povidone clearly produced quite hard 

tablets, especially the K90 viscosity 

grade; this polymer, as did the HPC, 

produced the more expected 

relationship between hardness and 

binder level. 

The friabilities were all < 0.5%, even 

for the softest of tablets. Even though 

some of the cellulosic binders 

produced tablets that were several 

SCU softer than the tablets produced 

using PVP, the friabilities were usually 

comparable. For example, note in 

Table 10 the values of hardness and 

friability for 6% METHOCEL K3P LV, 

6% HPC-EF, and 6% PVP (K90). 

The span of tablet weight variation 

(n=20) was from 1.53% RSD (relative 

standard deviation, standard 

deviation/mean x 100) for PVP (K90) 

at 3% to 0.56% RSD for HPC-EF at 

the 6% use level. This range is 

slightly more than the range observed 

for the acetaminophen granulations, 

but slightly less than the range 

observed for the Vitamin C 

granulations. The RSD of tablet 

thickness variations was also quite 

uniform, ranging from 0.59% (3% 

E5P LV) to 0.23% (for both 3% 

METHOCEL A15P LV and PVP 

(K29-32). 

Table Dissolution Properties 

The in vitro dissolution curves for the 

methazolamide tablets were recorded 

under USP conditions (Type 2 

apparatus, 100 rpm, 900 mL of pH 

4.5 acetate buffer) and are given in 

Figures 41-46. The USP limits are 

24 High Shear Granulation: Methazolamide Model

75% drug release in 45 minutes; the 

actual time that this amount of drug 

was liberated is denoted by t 75% . 

As was done in previous sections, the 

relative performance of the 5 

METHOCEL products will be 

examined first. In Figure 41, it can be 

clearly seen that there is essentially 

no difference in the time for 

dissolution when the binders were 

used at 3% w/w; the tablets prepared 

with METHOCEL A15P LV were just 

slightly faster to dissolve. When the 

binder is increased to 6% w/w (Figure 

42), there is virtually no shift in the 

dissolution behavior of tablets 

prepared with METHOCEL A15P LV. 

The tablets prepared with the 

METHOCEL E5P LV, E15P LV, and 

F4P LV required 6-8 minutes more 

time for dissolution with t 75% ≈ 25 

minutes. The 6% level of METHOCEL 

K3P LV as binder caused the tablets 

to require 12 additional minutes 

compared to the 3% level, but this 

was still one-third less than the 

specified 45 minutes. 

In Figures 43 and 44, comparisons 

are shown between two METHOCEL 

products and the two HPC products. 

At 3% binder (Figure 43), HPC-EF 

and HPC-LF perform in a manner 

similar to the hypromellose materials 

shown in Figure 41. Considerably 

more variety is exhibited at the 6% 

level (Figure 44). The lower molecular 

weight hydroxypropylcellulose gave 

t 75% of 37 minutes, while HPC-LF 

gave t 75% of 51 minutes, or 6 minutes 

longer than the USP specification. 

Figures 45 and 46 compare the 

METHOCEL A and K products with 

the PVP polymers. At 3% binder (see 

Figure 45), the K90 grade of PVP 

shows a somewhat longer drug 

release, although still well within 

limits. At 6% binder, shown in Figure 

46, the lower viscosity povidone has 

a dissolution profile that is nearly 

intermediate between that of the 

METHOCEL polymers, while the 

PVP (K90) material gave t 75% of 

52 minutes. 


25

Conclusions; Methazolamide 


In terms of densities and 

compressibility indices, there were no 

real differences in the polymers 

evaluated here. The particle size 

distributions were all reasonable, but 

several of the binders exhibited the 

behavior where more fines were 

present at the higher binder level than 

at the lower; METHOCEL K3P LV 

and the HPC products were 

exceptions to this observation. For the 

cellulose ether materials, the tablets 

produced from this formulation were 

slightly softer than those produced 

from the Vitamin C and significantly 

softer than the acetaminophen 

formulations, granulated in a high 

shear mixer-granulator. In particular, 

however, METHOCEL K3P LV and 

the PVP polymers were more 

effective binders with respect to tablet 

hardness in the methazolamide 

formula. Comparing identical 

methazolamide formulations 

granulated by fluid bed and by high 

shear reveals that the use of 

cellulosic polymers in the fluid bed 

produces significantly harder tablets. 

Even though the tablets were slightly 

softer, the friabilities were still quite 

low. Friabilities were comparable to 

the acetaminophen high shear work, 

lower than the Vitamin C high shear 

work, and comparable with the 

methazolamide fluid bed work. The 

binders that were best overall 

considering both hardness and 

friability were 3% and 6% 

26 High Shear Granulation: Methazolamide Model 

METHOCEL K3P LV, 6% HPC-EF, 

and the povidones. Considering the 

dissolution characteristics of the 

cellulose ether materials, all gave 

t 75% of < 20 minutes when used at 

the 3% level, with METHOCEL A15P 

LV being just slightly the fastest. 

Consistent with the other model 

systems, increasing the binder level 

resulted in longer dissolution times. At 

6%, METHOCEL A15P LV was the 

fastest, with METHOCEL E15P LV, 

E5P LV, and F4P LV at t 75% of 25 

minutes, K3P LV at 30 minutes, HPC- 

EF at 37 minutes, and HPC-LF at > 

45 minutes. Considerable differences 

were evident with the two viscosity 

grades of PVP. While the K29-32 

grade gave rapid release at the 3% 

level, the K90 grade produced the 

longest dissolution time of any of the 

tested binders at the lower binder use 

level. At 6%, the K90 grade produced 

tablets having t 75% > 45 minutes. All 

factors considered, the best binders 

for this formulation were METHOCEL 

K3P LV at 3%, PVP (K29-32), 

METHOCEL A15P LV, and the HPC 

products at 3%.

For more information, complete literature, and product samples, 

you can reach a Dow representative at the following numbers: 

From the United States and Canada: ............................call 1-800-447-4369 

............................fax 1-989-832-1465 

In Europe: ................................................................toll-free +800 3 694 6367 † 

................................................................call +32 3 450 2240 

................................................................fax +32 3 450 2815 

From Latin America and Other Global Areas: ..........call 1-989-832-1560 

..........fax 1-989-832-1465 

† Toll free from Austria (00), Belgium (00), Denmark (00), Finland (990), France (00), Germany (00), 

Hungary (00), Ireland (00), Italy (00), The Netherlands (00), Norway (00), Portugal (00), Spain (00), 

Sweden (00), Switzerland (00), and the United Kingdom (00). 

Or you can contact us on the Internet at www.methocel.com 

NOTICE: No freedom from any patent owned by Seller or others is to be inferred. Because use conditions and applicable laws may differ from one 

location to another and may change with time, Customer is responsible for determining whether products and the information in this document are 

appropriate for Customer’s use and for ensuring that Customer’s workplace and disposal practices are in compliance with applicable laws and other 

governmental enactments. Seller assumes no obligation or liability for the information in this document. NO WARRANTIES ARE GIVEN; ALL 

IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE ARE EXPRESSLY EXCLUDED. 

Published July 2002 

Printed in U.S.A. *Trademark of The Dow Chemical Company Form No. 198-01170-0702AMS 

*

Granulation with Dow Cellulosic Polymers II. High Shear Granulation

Create successful ePaper yourself

Delete template?

Save as template?