The Relative Densities of

Pharmaceutical Powders, Blends, Dry Granulations, and Immediate-Release Tablets

Bruno C. Hancock,* Joshua T. Colvin, Matthew P. Mullarney, and Andrey V. Zinchuk

The absolute and relative densities of pharmaceutical solids play an important role in determining their performance (e.g., flow and compaction properties) in both tablet and capsule dosage forms. In this article, the authors report the densities of a wide variety of solid pharmaceutical formulations and intermediates. The variance of density with chemical structure, processing history, and dosage-form type is significant. This study shows that density can be used as an equipment-independent scaling parameter for several common drug-product manufacturing operations.
IMAGE 100

any physical responses of powders, granules, and compacts such as powder flow and tensile strength are determined largely by their absolute and relative densities (18). Although measuring these properties is a simple task, a review of the literature reveals that a combined source of density data that formulation scientists can refer to does not exist. The purpose of this article is to provide such a reference source and to give insight about how these critical properties can be measured for common pharmaceutical solids and how they can be used for monitoring common drugproduct manufacturing operations.

Density, solid fraction, and porosity relationships

The majority of pharmaceutical solids are initially presented in powder form. The relative densities of these powdered pharmaceutical materials generally increase as the materials are processed into solid dosage forms. This occurs irrespective of the processing pathway used or the type of manufacturing equipment chosen (see Figures 1 and 2). Densification is often necessary to enhance the handling properties of the solid materials (e.g., powder flow) and to permit more-efficient processing operations to be used (e.g., high-speed tablet compression). Densification also facilitates dosing the active ingredient to patients in acceptably sized dosage forms. Several terms are commonly used to describe the apparent densities of solid pharmaceutical materials such as specific volume, porosity, and solid fraction. The following relationships between these parameters are: density mass volume (units of g/mL or kg/m3) mass 1 density [1b] mass molecular [2a] mass envelope volume [2b] [1a]

Bruno C. Hancock, Joshua T. Colvin, Matthew P. Mullarney, and Andrey V. Zinchuk are all scientists in the solids development group at Pfizer Inc., Eastern Point Rd., Groton, CT 06340, tel. 860.715.2484, fax 860.441.3972, [email protected].
*To whom all correspondence should be addressed.

specific volume volume (units of mL/g or m3/kg) absolute density volume apparent density

true density

64

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

ple in terms of its minimum or maximum density or even its density heterogeneity.

Methods
True (absolute) density measurement. The true or absolute densities of finely divided solids is (tablet, ribbon, granule) most often measured by pycnometry. Such Assembly Assembly Compressed Assembly measurements work by using a displacement of similar of particles assembly of compacts fluid such as helium, air, mercury, or oil to penmolecules of particles etrate the voids between neighboring particles, True density Apparent Apparent density Apparent thereby providing an estimate of the volume calculated from density calculated from density of the solid part of a sample. The principles of molecular calculated from weight and calculated from weight and weight and bulk volume of bulk weight such measurements are described in Chapter unit-cell volume volume of compacted and volume 699 Density of Solids in the current United powder powder States Pharmacopeia (10). In this work, the Should Includes Measured by Includes true densities of powder samples ( 1 g) were correspond to interparticle calipers or interparticle and determined using a helium pycnometer (Quanhelium density spaces immersion intercompact tachrome Inc., Boynton Beach, FL) that was spaces operated according to the manufacturers recommended procedures. Calibration was perFigure 1: Schematic describing the densities of solid pharmaceutical samples. formed using standard stainless steel spheres of known mass and volume. The mean value of triplicate determinations is reported. This property of the roller-compacted ribbons and granulations was measured in 1.0 exactly the same way as it was for the powders, whereas tablets were lightly crushed using a mortar and pestle before testing. All samples were equilibrated at controlled ambient conditions (22 2 C, 40 5% RH) before testing to ensure reproducible results. Bulk (envelope) density measurement for powders and granules. The bulk and tapped densities of powdered excipients, drugs, blends, and granulations are commonly determined using the Powder Granulation Tablet methods described by the American Society for Testing and Materials (ASTM) (11,12) and in the United States Pharmacopeia Chapter 616 Bulk Density and Tapped Density (10). The results of these measurements can be easily affected by the choice 0 of equipment, operator technique, or measurement conditions; Applied pressure therefore, one of these standardized procedures (1012) should be closely followed (1315). All measurements for this study Figure 2: Relative density changes expected during the manufacture were conducted at controlled ambient conditions (22 2 C, of a solid dosage form. 40 5% RH) using a tapped density instrument (VanKel, Cary, NC) fitted with 100-mL glass volumetric cylinders. All samples were tapped (14.25-mm height; 300 taps/min) until they reached in which the apparent density is usually bulk or tapped, and the a terminal density (typically after 2000 taps). envelope volume is the space bounded by the exterior surface Bulk (envelope) density measurement for compacted samples. The of the sample. envelope volume of roller-compacted ribbons and immediaterelease tablets can be determined by direct measurement with relative density apparent density absolute density [3a] calipers or by using standard fluid-displacement techniques. Both methods were used in this study. The direct measurement solid fraction relative density [3b] of ribbon and tablet dimensions was achieved with a standard digital micrometer calibrated using National Institute of Stanpercent porosity 100 (1 relative density) [3c] dards and Technology traceable gauge blocks. The instrument used for fluid displacement measurements was a Geopyc 1360 Pharmaceutical scientists almost always refer to the mean instrument (Micromeritics Instrument Corp., Norcross, GA), density of a sample; however, evidence for density gradients in and the displacement fluid comprised graphite lubricated glass solid pharmaceutical samples does exist (9), and under some spheres of a small size (mean volume diameter 132 m). circumstances it may be more appropriate to describe a sam- Cylindrical sample chambers with diameters between 12.7 and
Particle Powder Compacted powder Bulk tablets Relative density 66

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

Table I: Densities, solid fractions, and porosities of various powdered and granular pharmaceutical solids.
Sample type Powders Description Excipients Sorbitol (bulk) Sorbitol (tapped) Lactose spray-dried (bulk) Lactose spray-dried (tapped) Microcrystalline cellulose (PH101) (bulk) (17) Microcrystalline cellulose (PH105) (bulk) Microcrystalline cellulose (PH105) (tapped) Dibasic calcium phosphate anhydrous (bulk) Dibasic calcium phosphate anhydrous (tapped) Cross-linked poly(vinylpyrrolidone) (bulk) Stearic acid (bulk) Glyceryl behenate (bulk) Drug substances Bulk drug P (bulk) Bulk drug P (tapped) Bulk drug Q (bulk) Bulk drug Q (tapped) Bulk drug R (bulk) Bulk drug R (tapped) Direct compression Formulation A (bulk) Formulation A (tapped) Formulation B (bulk) Formulation B (tapped) Formulation C (bulk) Formulation C (tapped) Formulation D (bulk) Formulation D (tapped) Formulation E (bulk) Formulation E (tapped) Formulation E (bulk) (made from 0.60 solid-fraction ribbons) Formulation E (tapped) (made from 0.60 solid-fraction ribbons) Formulation E (bulk) (made from 0.80 solid-fraction ribbons) Formulation E (tapped) (made from 0.80 solid-fraction ribbons) Microcrystalline cellulose (PH101) (bulk) (13 roller compacted) (17) (2 roller compacted) (17) (3 roller compacted) (17) (5 roller compacted) (17) (10 roller compacted) (17) True density (g/mL) 1.48 1.48 1.55 1.55 1.56 1.57 1.57 2.80 2.80 1.21 0.99 1.00 1.29 1.29 1.31 1.31 1.42 1.42 1.52 1.52 1.93 1.93 1.91 1.91 1.56 1.56 1.76 1.76 1.76 1.76 1.76 1.76 Apparent density (g/mL) 0.59 0.73 0.53 0.67 0.33 0.25 0.46 0.70 1.43 0.29 0.59 0.57 0.14 0.36 0.43 0.69 0.25 0.44 0.57 0.64 0.60 0.72 0.73 0.84 0.59 0.70 0.43 0.58 0.48 0.69 0.58 0.90 Relative density (solid fraction) 0.40 0.49 0.34 0.43 0.21 0.16 0.29 0.25 0.51 0.24 0.60 0.57 0.28 0.38 0.33 0.53 0.18 0.31 0.38 0.42 0.31 0.37 0.38 0.44 0.38 0.45 0.24 0.33 0.27 0.39 0.33 0.51 Porosity (%) 60 51 76 67 79 84 71 75 49 76 40 43 72 62 67 47 82 69 62 58 69 63 62 56 62 55 76 67 73 61 67 49

Blends

Granulations

1.56 1.56 1.56 1.56 1.56

0.44 0.51 0.53 0.56 0.59

0.28 0.32 0.34 0.36 0.38

72 68 66 64 62

50.8 mm were used so that a variety of sample types and sizes could be analyzed, and fluid pressures from 0.07 to 0.22 MPa were used depending upon the size of the sample chamber. To permit the calculation of the bulk density of each sample using equation 2b, the samples were weighed using standard analytical techniques.

Materials
The densities of a wide range of pharmaceutical actives, excipients, and formulations (placebo and active) were measured during this study (see Tables IIV and Figure 3). Powdered excipients and active pharmaceutical ingredients (APIs) were evaluated for their true, bulk, and tapped densities to determine the normal range of powder densities that might be encountered during the development of solid dosage forms. These materiwww.phar mtech.com

68

Pharmaceutical Technology

APRIL 2003

Table II: Densities, solid fractions, and porosities of various roller-compacted and tableted pharmaceutical samples.
Sample type Roller-compacted ribbons Description Microcrystalline cellulose (PH105) (140 psi roll pressure) Microcrystalline cellulose (PH105) (355 psi roll pressure) Microcrystalline cellulose (PH105) (640 psi roll pressure) Formulation E ( 0.1-kg scale production) Formulation E ( 1-kg scale production) Formulation F ( 0.1-kg scale production) Formulation F ( 1-kg scale production) Formulation F ( 10-kg scale production) Placebo tablet formulation X (calipers) Placebo tablet formulation X (Geopyc) Placebo tablet formulation Y (calipers) Placebo tablet formulation Y (Geopyc) Placebo tablet formulation Z (calipers) Placebo tablet formulation Z (Geopyc) Direct compression Formulation A (5 kP; 100-mg weight tablet; eccentric press) (15 kP; 800-mg weight tablet; eccentric press) (6 kP; 100-mg tablet weight; rotary press) (20 kP; 800-mg tablet weight; rotary press) Formulation B (6 kP; 100-mg tablet weight; rotary press) (20 kP; 800-mg tablet weight; rotary press) Formulation C (6 kP; 100-mg tablet weight; rotary press) (20 kP; 800-mg tablet weight; rotary press) Formulation D (5 kP; 100-mg tablet weight; rotary press) (15 kP; 800-mg tablet weight; rotary press) (20 kP; 800-mg tablet weight; rotary press) Dry-granulated (roller-compacted) Formulation E (7 kP; 100-mg tablet weight; rotary press) (12 kP; 100-mg tablet weight; rotary press) (17 kP; 800-mg tablet weight; rotary press) Formulation F (12 kP; 160-mg tablet weight; rotary press) (18 kP; 640-mg tablet weight; rotary press) (23 kP; 640-mg tablet weight; rotary press) Commercial tablets Tums (10 kP, 1319-mg tablet weight) Alka-Seltzer (9 kP, 325-mg tablet weight) Motrin (5 kP, 390-mg tablet weight) Generic aspirin (8 kP, 377-mg tablet weight) Generic acetominophen (20 kP, 555-mg weight) Dimetapp tablet(16 kP, 470-mg tablet weight) True density (g/mL) 1.57 1.57 1.57 1.76 1.76 1.50 1.50 1.50 1.55 1.55 1.55 1.55 2.03 2.03 Apparent density (g/mL) 0.77 0.94 1.13 1.06 1.14 1.11 1.09 1.09 1.11 1.11 1.22 1.16 1.37 1.38 Relative density (solid fraction) 0.49 0.60 0.72 0.60 0.65 0.74 0.73 0.73 0.72 0.72 0.79 0.75 0.68 0.68 Porosity (%) 51 40 28 40 35 26 27 27 28 28 21 25 32 32

Tablets

1.52 1.52 1.52 1.52 1.93 1.93 1.91 1.91 1.56 1.56 1.56

1.23 1.28 1.35 1.37 1.49 1.51 1.53 1.57 1.39 1.40 1.42

0.81 0.84 0.89 0.88 0.77 0.78 0.80 0.82 0.89 0.90 0.91

19 16 11 12 23 22 20 18 11 10 9

1.76 1.76 1.76 1.50 1.50 1.50 1.84 1.90 1.42 1.42 1.29 1.34

1.50 1.55 1.37 1.37 1.29 1.37 1.55 1.59 1.21 1.31 1.16 1.04

0.85 0.88 0.78 0.91 0.86 0.91 0.84 0.84 0.85 0.93 0.90 0.78

15 12 22 9 14 9 16 16 15 7 10 22

70

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

sidered in the remainder of this article. First, a search of Material a proprietary database of deMicrocrystalline cellulose, NFa velopment tablet formulaLactose spray-dried, NFb tions was used to provide Dibasic calcium phosphate anhydrous, comparative data for various USPc material types, which were Croscarmellose sodium, NFd 3.00 3.00 3.00 3.00 broadly classified as drug Magnesium stearate, NFe 0.50 0.50 0.50 2.00 substances, excipients, blends, roller-compacted ribbons, a Avicel PH200, FMC Corporation (Newark, DE) b granulations (wet and dry), Fast-Flo, Foremost Farms (Rothschild, WI) c A-Tab, Rhodia Inc. (Chicago, IL) and tablets. The mean and d Ac-Di-Sol, FMC Corporation ranges of these data are sume Vegetable-derived, Malinkrodt (St. Louis, MO) marized in Figure 3. Second, specific examples of each of these types of materials were selected for more-detailed Table IV: Roller-compacted formulations (values in %). consideration, and their properties are presented in Table Formulation Formulation I and II. Finally, some commercial tablet samples were Material E F characterized, and their properties are presented in Table Drug substance 2.9 7.3 II. Following is a sequential discussion of these data, Microcrystalline cellulose, NFa 62.4 59.5b from raw materials through in-process samples to final Dibasic calcium phosphate anhydrous, 31.2 0.0 tableted products. USPc Drug substance and excipient powders. The true densiMannitol, NFd 0.0 29.7 ties of solid pharmaceutical samples should directly Croscarmellose sodium, NFe 3.0 3.0 reflect the fundamental chemical properties (e.g., moleMagnesium stearate, NFf 0.5 0.5 cular weight, molecular formula, and unit-cell dimena sions) of the materials concerned. Because drug subAvicel PH200, FMC Corporation b Added half as intragranular and half as extragranular stances and excipients are primarily organic materials, c A-Tab, Rhodia Inc. their true densities did not vary greatly for the most part, d 2080 Granular, SPI Pharma (Newcastle, DE) with most materials exhibiting values that fell within the e Ac-Di-Sol, FMC Corporation range of 1.21.6 g/mL (see Table I) (16). Notable exf Vegetable-derived, Malinkrodt ceptions to this behavior were observed for materials als were stored at controlled ambient conditions (22 2 C, 40 with very different chemical structures such as inorganic ex5% RH) and used as they were received from their suppli- cipients (e.g., dibasic calcium phosphate) (2.8 g/mL) and waxy ers. Active and placebo blends were made from the powdered materials (e.g., glyceryl behenate, stearic acid; 1.0 g/mL and materials using either a low-shear V-blender (Blendmaster, Pat- 0.99 g/mL, respectively). terson-Kelley, East Stroudsburg, PA) or a high-shear mixer (TurIn general, the powdered APIs had the lowest relative densibula, Glen Mills Inc., Clifton, NJ), and their true, bulk, and ties. All samples had bulk relative densities 0.4, and some samtapped densities were determined. Several of the excipients and ples had bulk relative densities 0.2. This finding suggests that formulations (active and placebo) were roller compacted with the majority of APIs must be densified before manufacturing TF-mini or TF-156 roller compactors (Vector Corp., Marion, the final dosage form so that sufficiently high doses can be adIA). Roll speeds, pressures, and ribbon thicknesses were 24 ministered in a reasonably sized dosage form. Most of the exrpm, 100700 psi, and 13 mm, respectively. Both smooth and cipients that were evaluated had densities that were more suited serrated rollers were used, and ribbon samples were taken once to processing directly into solid oral dosage forms, and the range steady-state operating conditions had been reached. Lastly, of bulk relative densities for 48 excipients was between 0.5 and placebo and active tablets of a variety of sizes were manufac- 0.2 (see Table I and Figures 3 and 4). Several excipients such as tured from the blends and granulations at the laboratory scale sorbitol and dibasic calcium phosphate anhydrous exhibited using either a single-station tablet press (F-press, Manesty, tapped relative densities 0.4; therefore, they can be considKnowsley, Merseyside, UK) or a nine-station rotary tablet press ered well suited for use in direct-compression tablet formula(T-100 model, Kilian & Co. Inc., Horsham, PA). Operating con- tions in which no additional densification step is performed. ditions included the use of gravity or force-feeding systems Blends. Combinations of materials should have true densiand production rates of 500060,000 tablets/h. The formula- ties that are intermediate to their component materials, which tions were lubricated by low-shear blending with 0.5 or 1.0% was the case for all formulations (blends, granulations, tablets) magnesium stearate. considered in this work (see Tables I and II). The formulations, including blends, with true densities 1.6 g/mL almost always Results and discussion contained some significant amount of a dense inorganic exTo draw applicable conclusions, several data sources are con- cipient such as calcium phosphate.
Formulation A 48.25 48.25 0.00 Formulation B 48.25 0.00 48.25 Formulation C 0.00 48.25 48.25 Formulation D 47.5 47.5 0.00
72

Table III: Direct-compression placebo tablet formulations (values in %).

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

of the powdered-drug, excipient, and blended formulation samples (see Tables I and II and Figure 3). Bulk For one model excipient (microcrystalline cellulose, Tapped Avicel PH105, FMC BioPolymer, Philadelphia, PA) 0.8 the relative density of the roller-compacted ribbons was between 0.49 and 0.72 (depending upon the 0.6 compaction conditions used), compared with a value between 0.16 and 0.29 for the powdered material (see Tables I and II). This amounts to a two- or threefold 0.4 increase in the materials density upon roller compaction and clearly illustrates the significant effect 0.2 that this processing operation can have upon the bulk properties of common pharmaceutical raw materials. For active Formulation E, the relative density of 0.0 Drug Excipients Blends Ribbons GranuTablets the roller-compacted samples also was increased by substances (n = 48) (n = 15) (n = 15) lations (n = 45) two- to threefold more than that of the blended for(n = 12) (n > 400) mulation, thereby indicating that this finding is not unique to single excipient samples. Notably, the range Figure 3: Relative density averages and ranges for solid pharmaceutical of ribbon relative densities that was recorded for 15 development samples. roller-compacted excipient and formulation samples was narrower than that of the powdered drug substances, excipients, and blends, which suggests that processing these maSorbitol (bulk) terials by roller compaction causes them to have more-consisSorbitol (tapped) tent relative densities (see Figure 3). The relative densities of the roller-compacted microcrysLactose spray-dried (bulk) talline cellulose samples varied measurably with changes in the Lactose spray-dried (tapped) roller compaction conditions such as roll pressure (see Table II). Because relative density is an intrinsic sample property, it Microcrystalline cellulose (PH101) (bulk) may be a very useful parameter to track during scale-up and process optimization experiments. Relative density would be Microcrystalline cellulose expected to reflect the overall mechanical performance of roller(PH105) (bulk) compacted ribbons and to be sensitive to changes in raw maMicrocrystalline cellulose terial properties, major process modifications, and even batch(PH105) (tapped) to-batch processing variations. Measuring this property using Dibasic calcium phosphate calipers and a fluid-displacement method resulted in similar anhydrous (bulk) results and identical relative rankings of the three placebo formulations (see Table II and Figure 5). The data show that eiDibasic calcium phosphate anhydrous (tapped) ther method could be used as an in-process test to track changes in ribbon properties or in process performance once the apCross-linked propriate calibration and methods validation experiments have polyvinylpyrrolidone (bulk) been performed. Stearic acid (bulk) Screened bulk granulations. The relative densities of more than 400 wet and dry development granulations fell within a very Glyceryl behenate (bulk) wide range. On average, the densities were not significantly dif0.0 0.2 0.4 0.6 0.8 1.0 ferent from those of a similar set of powder blends (see Figure Figure 4: Mean relative densities of some pharmaceutical excipient 3). Although granulating usually increases the apparent densipowders. ties of many materials, a significant number of occasions exist The relative densities of the blends were very similar to those in which the granulation process has no effect or even the opof the single excipient samples, which is not surprising because posite effect. One can speculate that this finding is due to the these materials were used in the majority of the blends. The use of noncompressive granulation operations such as fluidrange of relative densities for blends was narrower than that bed granulation or that it is the result of suboptimal granulaof both the drug substance and excipient powders, which sug- tion, milling, and screening operations being used. Another gests that simply blending these powdered raw materials in a possible explanation is that the density of some individual agformulation may lead to a more consistent range of densities. glomerates is increased by the granulation process, but the overRoller-compacted ribbons. The relative densities of the roller- all bulk density of these granulations is not increased because compacted ribbons from typical development formulations of changes in the particle packing efficiency induced by shifts ranged from 0.6 to 0.8 and were markedly higher than those in the particle-size distribution (e.g., altered fines content).
1.0 Relative density 74

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

1.0 Mean relative density 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Formulation X Formulation Y Formulation Z Relative density 0.9 Calipers Fluid displacement

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Tums Alka- Motrin Generic Generic DimeSeltzer aspirin pain tapp reliever

Figure 5: Comparison of the relative densities for roller-compacted ribbons of Formulations X, Y, and Z determined using the caliper and fluid-displacement methods (standard deviation 0.01 in all cases).

Figure 7: Mean relative densities of some commercial pharmaceutical tablets (standard deviation 0.02 in all cases).

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0x 1x 2x 3x 5x 10x Number of roller compaction repetitions

0.8 Time (min) or strength (kP) 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.70

Relative density

Disintegration time _ eccentric press Disintegration time _ rotary press Crushing strength _ eccentric press Crushing strength _ rotary press

0.75

0.80 Solid fraction

0.85

0.90

0.8 Time (min) or strength (kP)

Figure 6: The influence of repeated roller compaction on the relative density of microcrystalline cellulose (Avicel PH101) (17).

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.0

To address this unanticipated result in more detail, granules of Formulation E were obtained by gently milling the rollercompacted ribbons using a cone mill (Mini or 193S model Comill, Quadro Engineering Inc., Waterloo, ON, Canada). Data for dry-granulated samples of microcrystalline cellulose were also obtained from a recent article (17). The bulk and tapped solid fractions of Formulation E granules varied according to the solid fractions of the ribbons from which they were produced, with the most-dense ribbon samples producing the mostdense granules (see Table I). A similar pattern of behavior was observed with the microcrystalline cellulose samples that were roller compacted on multiple occasions, with the material that had been compacted to the greatest extent producing the mostdense granule samples (see Table I and Figure 6) (17). For Formulation E, the granules had as much as a twofold lower density than the ribbons that they were produced from, but they still had a greater density than the uncompacted blend of the same formulation. Similarly, the roller-compacted microcrystalline cellulose granules were 3080% more dense than the ungranulated excipient. These two data sets support the widely
76

Disintegration time _ eccentric press Disintegration time _ rotary press Crushing strength _ eccentric press Crushing strength _ rotary press

4.0

8.0

12.0

16.0

20.0

24.0

28.0

Compression stress (MPa)

Figure 8: (a) Mean tablet-crushing strength and disintegration time as a function of solid fraction for Formulation A (error bars denote one standard deviation); (b) Mean tablet-crushing strength (kP) and disintegration time (min) as a function of compression stress for Formulation A (error bars denote one standard deviation).

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

held belief that granulation normally increases the density of pharmaceutical powders, at least when using a compressive granulation process such as roller compaction. Further research is required to determine the specific circumstances that could lead to a reverse trend and could result in no significant degree of bulk densification occurring upon granulation. Tablets. The tablet samples had the highest relative densities of all the material types considered (mean 0.84) (see Figure 3 and Table II), and the range for typical development tablet formulations was between 0.66 and 0.93. For commercially available tablets the average relative density was practically identical to this, and the range of tablet relative densities was also very similar (see Table II and Figure 7). No obvious difference was noted between the relative densities of the directcompression, dry-granulated, or commercial tablets evaluated in this study. One should note that the relative density of conventional pharmaceutical tablets is such that they have a significant porosity and thus they cannot be expected to behave as if they were solid bodies, which is an important consideration for those interested in studying the physical and mechanical properties of pharmaceutical tablets. The tablets of the direct-compression placebo Formulations A, B, C, and D (see Table III) had relative densities that somewhat depended on the type of tablet press used and the tablet size, although the greatest differences were due to changes in the formulation composition (see Table II). The direct-compression formulations containing dibasic calcium phosphate formed tablets that were less dense at any given tablet-crushing strength, even though this particular excipient has a high true density. The effect of increasing the lubricant level from 0.5 to 2.0% in the direct-compression microcrystalline celluloselactose blends (Formulation A versus D) was to slightly increase the degree of densification required to achieve a given tablet-crushing strength. Tablets of the dry-granulated Formulation E had relative densities that were two- to fourfold greater than their starting blends, two to three times higher than their granulations, and 1020% greater than their roller-compacted ribbons (see Table II). When compared with the starting bulk drug substance, these tablets were more than four times more dense. Scale-up. The choice of tablet press operating conditions and tablet characteristics such as size and shape clearly influences the relative density of the tablets that are produced (see Table II), and this parameter could thus be used as a parameter to discriminate between samples made using different conditions. Because the relative density can be determined for tablets of all shapes and sizes (and also for powders, blends, granulations, and roller-compacted ribbons), it is a very attractive parameter to use when comparing various solid pharmaceutical samples. Instead of using equipment-dependent variables such as compression pressure as a common condition for comparing samples, one can compare them at a common relative density. For fundamental mechanical property testing of pharmaceutical compacts, several authors have advocated collecting data

from a range of relative densities and extrapolating to the theoretical zero-porosity point to permit such material comparisons (18,19). This approach requires more material and time than comparisons at a single experimentally accessible relative density and does not consider the properties of the materials in their normal operating state. Routinely comparing the properties of pharmaceutical compacts at a standardized relative density (i.e., 0.9 [10% porosity]) as advocated by several researchers (16,20) seems to be the most reasonable approach for formulation development and process scale-up experiments. The data in Figures 8a and 8b provide an example of the usefulness of solid-fraction measurements during tablet formulation development and process scale-up studies. During the early stages of tablet development, the supply of the API for Formulation A was limited, thus prototype tablets required to evaluate the stability of the formulation were manufactured using an instrumented single-station eccentric tablet press. As the project progressed, more of the API was synthesized, and the manufacture of sufficient tablets for clinical studies required the use of a small rotary tablet press. Because the compression characteristics and instrumentation of eccentric and rotary tablet presses are quite different, the solid fraction of the tablets was used as the parameter to gauge the equivalence of tablets made using two types of tablet presses (see Figure 8a). From the data, one can observe that using this approach resulted in tablets with equivalent mechanical properties and in vitro disintegration performance, whereas comparisons made on the basis of the measured compression forces of the two types of presses, as is done in many cases, would not have resulted in equivalent tablets being manufactured for the stability evaluations and clinical studies (see Figure 8b).

Conclusions
The authors have reported the absolute and relative densities of a wide range of pharmaceutical solids and probed some of the influences of chemical structure, processing history, and dosage-form type on these properties. Monitoring the relative density of solid pharmaceutical materials (e.g., API, excipients, blends, ribbons, granules, tablets, etc.) can be useful during the design, optimization, and scale-up of manufacturing processes for solid pharmaceutical dosage forms and may help achieve robust drug-product manufacturing processes. For example, the bulk densities of excipient and API powders will clearly indicate whether a need exists for a densifying unit operation such as dry granulation before the manufacture of tablets. Similarly, when attempting to manufacture uniform tablet dosage forms using various tablet presses, monitoring the compact relative density instead of a remote instrument parameter should provide a more robust means of ensuring consistent product performance.

Acknowledgments
The authors thank Cindy Oksanen, Beth Langdon, Dauda Ladipo, Glenn Carlson, and Tracy Orlinksi for their active support of this study.

78

Pharmaceutical Technology

APRIL 2003

www.phar mtech.com

References
1. S. Pedersen and H.G. Kristensen, Change in Crystal Density of Acetylsalicylic Acid During Compaction, STP Pharm. Sci. 4 (3), 201206 (1994). 2. S.N. Tsu, On the Relationship between Specific Volume and Crystallinity of Starch, Starch 31 (3), 7375 (1979). 3. R.M. Spriggs, Expression for the Effect of Porosity on Elastic Modulus of Polycrystalline Refractory Materials, Particularly Aluminium Oxide, J. Am. Ceram. Soc. 44, 625629 (1961). 4. E.C. Abdallah and D. Geldart, The Use of Bulk Density Measurements as Flowability Indicators, Powder Technol. 102, 151165 (1999). 5. J.M. Newton and F. Bader, The Prediction of the Bulk Densities of Powder Mixtures and Its Relationship to the Filling of Hard Gelatin Capsules, J. Pharm. and Pharmacol. 33, 621626 (1981). 6. A.C.Y. Wong, Characterisation of the Flowability of Glass Beads by Bulk Densities Ratio, Chem. Eng. Sci. 55 (18), 38553859 (2000). 7. C. Cohard et al., A Correlation between Flowability and Tapping Variation Rates of Powders, Int. J. Pharm. Technol. Prod. Manufac. 6 (3), 1016 (1985). 8. F. Nicklasson and G. Alderborn, Analysis of the Compression Mechanics of Pharmaceutical Agglomerates of Different Porosity and Composition Using the Adams and Kawakita Equations, Pharm. Res. 17 (8), 949954 (2000). 9. G. Nebgen et al., 1H-NMR Microscopy of Tablets, J. Pharm. Sci. 84 (3), 283291 (1995). 10. United States Pharmacopeia, USP 25NF 20 (United States Pharmacopeial Convention Inc., Rockville, MD, 2002). 11. ASTM, Standard Test Method for Tap Density of Powders of Refractory Metals and Compounds by Tap-Pak Volumeter, (American Society for Testing and Materials, West Conshohocken, PA, 1985). 12. ASTM, Standard Test Method for Mechanically Tapped Packing Density of Formed Catalyst and Catalyst Carriers (American Society for Testing and Materials, West Conshohocken, PA, 1988). 13. H.G. Brittain et al., Changes in Material Properties Accompanying the National Formulary Identity Test for Microcrystalline Cellulose, Pharm. Res. 10, 6167 (1993). 14. H.G. Brittain, Physical Characterization of Pharmaceutical Solids, Pharm. Res. 8, 963973 (1991). 15. H.G. Brittain, Functionality Testing of Excipient Materials, Pharm. Technol. July (Contempo Issue), 6668 (1993). 16. A.H. Kibbe, Handbook of Pharmaceutical Excipients, American Pharmaceutical Association & Pharmaceutical Press, Washington, DC & London, UK, 2000. 17. J.M. Bultman, Multiple Compaction of Microcrystalline Cellulose in a Roller Compactor, Eur. J. Pharm. Biopharm. 54, 5964 (2002). 18. A.B. Mashadi and J.M. Newton, Assessment of the Mechanical Properties of Compacted Sorbitol Instant, J. Pharm. Pharmacol. 39, 67P (1977). 19. B.C. Hancock, K. Christensen, and S.-D. Clas, Microscale Measurement of the Mechanical Properties of Compressed Pharmaceutical Powders, Part 1: The Elasticity and Fracture Behavior of Microcrystalline Cellulose, Int. J. Pharm. 209, 2735 (2000). 20. E.N. Hiestand, Mechanics and Physical Principles for Powders and T Compacts (SSCI, West Lafayette, IN, USA, 2002). P

80

Pharmaceutical Technology

Circle/eINFO 56
APRIL 2003

www.phar mtech.com