Evaluation and Characteristics of a New Direct Compression Performance Excipient

Direct compression (DC) is a preferred manufacturing process as the continually modernizing pharmaceutical industry strives to improve its manufacturing output while reducing operating costs (1, 2). Compared with wet-granulation technology, direct compression offers the advantage of product stability, simplification of manufacturing process, and lower process cost (see Figure 1) (3). The tableting blend for a DC process contains the active pharmaceutical ingredient (API), a filler, a binder, a disintegrant, auxiliary excipients (e.g., glidants and solubilizers), and a lubricant. DC technology and the use of modern tableting machines demand that the excipients and API form a compressible mixture with excellent flowability and a low tendency of particle segregation.

Figure 1: Simplification of a manufacturing process using engineered excipients with direct-compression technology (API is active pharmaceutical ingredient, DC is direct compression, WG is wet granulation, and RC is roller compaction).

The choice of tableting process is highly influenced by the flowability and compressibility of the API-excipient mixture (see Figure 2) (1). The particle size/shape, density, moisture content, and composition of the excipients affect flowability and compressibility, which ultimately drives the tableting process (see Figure 3) (4–9). Traditional excipients have limited ability to provide flowability and compressibility in mixtures with cohesive and poorly flowable APIs. Tableting parameters, such as equipment geometry and energy input add to the complexity of the DC system when dealing with multiparticulate powder systems (10).

Figure 2: Powder blend compressibility and flowability requirements for various tableting technologies (DC is direct compression, WG is wet granulation, and RC is roller compaction).

To increase the use of direct compression in pharmaceutical tableting, novel excipients with enhanced flow and compressibility, which can accommodate API variability, are needed. Nonetheless, this is not an easy task to achieve as the more compressible a material is, the less flowable it will be. The most cost-effective methodology is to enhance certain functionality of an existing excipient by using novel processing techniques, or synergistically combining it with other commonly used excipients. For example, processing techniques, such as spray drying or granulation, have been used over time to improve microcrystalline cellulose (MCC) properties. By using combinations of excipients classified as generally recognized as safe (GRAS) and innovative processing techniques, particles can be engineered to provide desired properties for use in a DC process. The resulting engineered excipients are commonly named "coprocessed," "high functionality," "multifunctional," or "performance" excipients. This new class of excipients streamlines the process such that, in the end, the formulator will have the choice of preparing a DC blend consisting of API, "coprocessed excipient," auxiliary excipients, and lubricant. This approach may be acceptable to the industry and regulatory agencies because the novel excipients created are not considered new chemical entities (11).

Figure 3: Critical performance parameters for excipients to be used in direct compression (DC). (LOD is loss on drying.)

The current status of high-functionality excipients has been reviewed in various articles (12–15). This class of excipients should be very appealing to implementation in a pharmaceutical process governed by a quality-by-design (QbD) approach (16). Based on QbD principles, the quality of the drug product is a function of drug substance, excipients and, manufacturing process; thus using a performance excipient in a QbD formulation should simplify the scheme.

Taking into consideration the requirements of QbD and a DC process, a performance excipient, PanExcea MHC300G, based on MCC, hydroxypropyl methylcellulose (HPMC), and crospovidone (CPVD), has been developed. The performance excipient was designed to provide good flowability and compressibility by optimizing particle shape, size, porosity, density, composition, and surface roughness (17–18). The objective of this paper is to evaluate the physical properties and functionalities of PanExcea MHC300G and to correlate these properties with its performance in a high-speed tableting machine.

Experiment

Materials. The authors obtained the following materials: PanExcea MHC300G, Avantor Performance Materials (formerly Mallinckrodt Baker); microcrystalline cellulose (102 RanQ, RanQ Pharmaceuticals & Excipients); microcrystalline cellulose (Emcocel 90, JRS Pharma); magnesium (Mg) stearate (Product No. 2256-05, Mallinckrodt Chemicals); stearic acid (Hystrene, PMC Biogenics); colloidal silica (RxCipients GL100, J. M. Huber); hydroxypropyl methylcellulose, USP grade (Pharmacoat, grade 603, SHIN-ETSU Chemical); crospovidone (Polyplasdone XL-10, International Specialty Products (ISP)); and ibuprofen (respectively, Product grade: Albemarle 20, Albemarle 40 and Albemarle 70, Albemarle; and Ibuprofen 50, BASF).

Methods. High-shear wet granulation (HSWG) of MCC (89%)/HPMC(2%)/CPVD (9%). In a 1-L stainless-steel bowl were placed 133.5 g MCC, 3.0 g HPMC and 13.5 g CPVD. The bowl was attached to a vector micro high-shear mixer/granulator (GMX.01, Vector). The dry mixture was mixed for 2 min at 870 rpm impeller speed and 1000 rpm chopper speed.

Drop by drop, 70 g of deionized water (i.e., the liquid binder) was added to the dry blend using a peristaltic pump at a dose rate of 12 g/min. During the liquid binder addition, the impeller speed was 700 rpm and the chopper speed was 1500 rpm. The wet massing time was 60 s, maintaining the same impeller and chopper speed as during liquid addition.

Following the granulation, the wet granular material was dried in a tray at 60 °C. The resulting granular material (moisture content 2.35%) was screened through a 30 mesh sieve. The yield of the granular material that passed through 30 mesh was 116.73 g (79.3 % referenced to dry starting materials and dry product).

Physical characterization. Particle-size analysis was performed using an air jet sieving instrument (Micron Air Jet Sieve, Hosokawa Micron Powder Systems). Angle of repose, aerated bulk density/tapped bulk density, and total flowability index were measured using a powder tester (Model PT-S, Hosokawa Micron Powder Systems). Particle morphology was assessed using an environmental scanning electron microscope (ESEM, XL30, FEI) using a voltage of 5 kV, spot size of 3, and SE detector. The samples were sputtered with iridium before SEM analysis (sputtering time 40 s).

Granule strength was determined using the following three methods. Method A: The particle-size distribution of 75–100 g of MHC300G was measured and the material was loaded in a 4-L V-blender (MaxiBlend, GlobePharma) and tumbled for two hours. After tumbling, the granular material was collected and analyzed again for particle-size distribution. Method B: 60 g of MHC300G were charged in a 1-L stainless-steel bowl that was attached to a vector micro high-shear mixer/granulator (GMX.01,Vector). The powder was processed by applying an impeller speed of 950 rpm and a chopper speed of 3600 rpm for 5 min. The particle-size distribution was measured before and after the high shear experiment. Method C: 100 g of MHC300G were passed through a mill (Quadro Comil model U3, Quadro Engineering) at 3000 rpm using US # 16 (1180 µm) screen.

Dilution potential of MHC300G was assessed by preparing blends of ibuprofen (IBU), MHC300G and silica. The blends were prepared in a V-blender (MaxiBlend, GlobePharma). Before blending, the powder mixture was passed through a 30 mesh sieve to break the IBU clumps. A blend containing IBU/MCC/HPMC/CPVD/silica was prepared in a similar manner. Powder characteristics as well as IBU content uniformity were analyzed for all blends.

Tableting studies. Unless otherwise specified, tablets were produced on an instrumented 10-station rotary tableting press (RIVA-Piccola, SMI). The tablet press was configured with 10-mm (0.3937 in.) round standard concave Tableting Specification Manual (TSM) "B" tooling. A compaction profile (i.e., the variation of tablet properties with increasing compression force at constant tablet weight and turret rotation speed) and a strain-rate study (i.e., the variation in tablet properties with turret rotation speed at constant tablet weight and compression force) were performed for powder blends containing: a) 88.75% MCC (RanQ 102), 2% HPMC, 9% CPVD, Mg stearate 0.25%; b) 99.75% PanExcea MHC300G, 0.25% Mg stearate; c) 30% ibuprofen, 61% MCC (Emcocel 90), 1.5% HPMC, 6.0% CPVD, 1% silica, 0.5% stearic acid; d) 30% ibuprofen, 68.5% MHC300G, 1% silica, 0.5% stearic acid.

Figure 4: Evaluation and characteristics of a new direct-compression performance excipient. Scanning electron microscope micrographs of A: MHC300G (engineered particles); B: EMB001 (engineered particles); C: EMB002 (traditional high-shear wet granulation); and D: microcrystalline cellulose. The scale bar in the images in the left column represents 100 µm, and the scale bar in the right column represents 20 µm.

The blends for the tableting studies were prepared in the following manner: all ingredients except the lubricant (Mg stearate or stearic acid), were mixed, passed through a mill (Quadro Comil model U3, Quadro Engineering) at 3000 rpm using US # 16 screen and then blended in a V-blender (MaxiBlend, GlobePharma) for 15 min at 20 rpm. The lubricant was added to the resulting mix in the V-blender and everything was blended for 3 min at 20 rpm. The final blend was discharged from the V-blender and transferred to the tableting machine.

Director Software (SMI) was used to collect and analyze the tableting data.

Table I: Composition of PanExcea MHC300G, EMB001, EMB002, and EMB003 excipients.

Tablet characterization. A friability test was performed according to the United States Pharmacopeia (USP) recommendations for friability determination of compressed, uncoated tablets (Chapter <1216>) using a tablet-friability tester (Varian Friability Tester, with Varian drum, Varian). The hardness of the tablets was measured using a tablet-hardness tester (Benchsaver series, VK 200, Varian). Tablet thickness was measured using a micrometer. Tablet-disintegration tests were performed with a disintegration system (3100, Distek) using 900 mL deionized water at 37 ± 0.5 °C.

Table II: Bulk density, compressibility index, and Hausner ratio for various grades of microcrystalline cellulose (MCC) and coprocessed excipients containing MCC.

Results and discussion

Physical characterization of PanExcea MHC300G performance excipient. Composition and particle morphology. The SEM micrographs (see Figure 4) of particles obtained using several manufacturing processes and compositions (see Table I) show the particle morphology of all materials as being different. The MHC300G containing 89% MCC, 2% HPMC, and 9% CPVD has homogeneous spherical to quasi-spherical particles with a rough surface and some open pores (see Figure 4A). The increase in concentration of HPMC to 5% (EMB001) triggered a significant increase in the open pores (see Figure 4B). The difference in porosity is also reflected in changes in aerated bulk density and tapped density. The higher densities of MHC300G indicate denser particles as compared with EMB001 (see Table II). The composition of MHC300G was finalized to provide optimal particle shape, size, and porosity needed for direct compressible materials (see Figure 3). The SEM micrographs of the high-shear granulated material (EMB002) having similar composition to that of MHC300G show agglomerated particles (see Figure 4C), while the MCC shows typical needle-shaped particles (see Figure 4D) (19).

Table III: Particle-size distribution and physical characteristics of PanExcea MHC300G (three sample lots).

Particle-size distribution and flowability parameters. The average particle size and the particle-size distribution of a granular material is known to impact flowability, blending ability, wetting, drying, mechanical properties, and stability (4). The flowability of pharmaceutical blends (i.e., API plus excipients) is critical for effective use of direct compression. Poor flow can cause bridging, arching, surging, and enhanced movement of particles in the die cavity. There are few simple acceptable tests to evaluate the flowability of a powder, the Hausner ratio and measurement of total flowability index being just few of them (20, 21). The Hausner ratio is defined as the ratio of tapped to aerated bulk density, where values < 1.25 indicate good flow properties. The total flowability index, as described by Carr, is calculated based on a series of measurements of powder properties (e.g., angle of repose, aerated and tapped bulk density, and particle-size uniformity) using a Hosokawa powder tester (21, 22). The values of the flowability index range from 0 to 100, with 100 being the material with the best flow properties. A material with a total flowability index of > 80 has very good flowability.

Figure 5: Flowability index for EMB003 (powder blend), EMB002 (HSWG), MHC300G, and commercial grades of microcrystalline cellulose (MCC) (Avicel PH 102, Emcocel 90). Higher index indicates better flowability (see Reference 21).

The particle-size distribution and the physical characteristics of MHC300G are summarized in Table III. A total flowability index greater than 80, an angle of repose below 35°, and a Hausner ratio less than 1.25 indicate that MHC300G has very good flowability. When compared with various grades of MCC, EMB001, EMB002, and EMB003, MHC300G provides superior performance in terms of flowability (see Table II and Figure 5). For example, Avicel PH 102 has a Hausner ratio of ~ 1.33, a compressibility index of ~ 25% and a total flowability index of 72, and MHC300G has a Hausner ratio of 1.21, a compressibility index of ~ 17 % and a total flowability index of 87. These data suggest that the processing methodology used in MHC300G particle engineering improved flowability attributes significantly.

Table IV: Granules friability test* for the MHC300G using a V-blender. Comparison with a granular product of similar composition obtained by high-shear wet granulation.

Particle strength. The particles used in a DC process should be strong and not produce fines during blending or mixing operations. The MHC300G granules and a granular material of similar composition prepared by conventional HSWG (EMB002) were tested in a particle friability experiment. Upon tumbling samples of excipients in a V-blender for two hours, the percentage of fines was unchanged for MHC300G, but for EMB002 (see Table IV) they were doubled. The strength of the MHC300G particles was tested also in a high-shear process in which the impeller was run at 950 rpm for 5 min, or by passing the MHC300G excipient through a mill at 3000 rpm and a mesh size corresponding to US # 16 (1180 µm). All tests showed (see Table V) that the particles did not break during the process and the percentage of fines did not increase. The granule-strength studies and the SEM micrographs indicated that the MHC300G performance excipient consists of strong particles with robust bonding bridges between the granules' components, resulting in a unique structural morphology.

Table V: Granules friability test for the MHC300G using high shear or mill.*

Tableting study of MHC300G. PanExcea MHC300G was designed to ensure its use in high-speed tableting machines. Its physical properties, functional characteristics (i.e., flowability and compressibility) and particle strength indicate its suitability for DC processes in high-speed tableting machines. A tableting study was undertaken to understand the compaction profile and the variation in tablet properties with tableting parameters. The effect of turret-rotation speed on tablet quality (i.e., the strain rate) was studied as it is very important for tableting process scale-up. Placebo tablets of MHC300G/Mg stearate (0.25%) and a powder blend of MCC/HPMC (2%)/CPVD (9%)/Mg stearate (0.25%), respectively, were obtained using a 10-station rotary tableting machine configured with 10-mm round standard concave TSM "B" tooling.

Figure 6: Tensile strength versus compression force profile for tablets (10 mm diameter) prepared from MHC300G/0.25%Mg stearate blend, and MCC/2%HPMC/9%CPVD/0.25%Mg stearate blend (MCC is microcrystalline cellulose, HPMC is hydroxypropyl methylcellulose, and CPVD is crospovidone).

Table VI: Ibuprofen formulations used in tableting studies.

A compaction profile was performed for various compression-force levels at constant turret-rotation speed (32 rpm) and tablet weight. Tablets were collected at each run to measure the average weight, thickness, breaking force (i.e., hardness) and disintegration times. The plots of tensile strength versus compression force (see Figure 6) show similar profiles and an increase in tensile strength with compression force for both MHC300G/Mg stearate and the blend of MCC/HPMC/CPVD/Mg stearate. MHC300G produces strong compacts (i.e., tablet radial-tensile strength of 1.2–4.5 MPa or tablet breaking forces of 95–260 N for 10-mm diameter compacts) in the compression-force range used in current tablet manufacturing (i.e., 5–15 kN) (see Figure 6). These values are lower than the ones corresponding to tablets prepared from MCC/HPMC/CPVD/Mg stearate blend. The results are not unexpected as it is well known that processing MCC via a wet-granulating technique results in some reduction of the compressibility of MCC, the mechanism of which is not well understood (23, 24). It has been suggested that the decrease in MCC compactibility after granulation is associated with the decrease in MCC primary particle porosity (23). The formulation containing 30% ibuprofen and PanExcea MHC300G, however, gives stronger compacts at compression forces up to 15 kN than the corresponding formulation containing MCC (see Table VI and Figure 7). Tablets prepared when using MHC300G are shown to retain the tensile strength (3.2 MPa for placebo and 3.0 MPa for 30% ibuprofen formulation at ~ 9 kN compression force) and the tensile strength of tablets prepared using the blend of MCC/HPMC/CPVD is reduced after mixing with ibuprofen (4.7 MPa for placebo and 2.5 MPa for 30% ibuprofen formulation at ~ 9kN compression force; see Figures 6 and 7). No significant tablet weight variation, capping or laminating tendencies were observed.

Figure 7: Tensile strength vs. compression force profile for tablets (10 mm diameter) prepared from 30%IBU/MHC300G/silica/stearic acid blend, and 30%IBU/Emcocel90/HPMC/CPVD/silica/stearic acid blend (IBU is ibuprofen, HPMC is hydroxypropyl methylcellulose, and CPVD is crospovidone).

Figure 8: Variation in tablet weight with turret rotation speed for tablets (10 mm diameter) prepared from MHC300G/0.25%Mg stearate blend, and MCC/2%HPMC/9%CPVD/0.25%Mg stearate blend (MCC is microcrystalline cellulose, HPMC is hydroxypropyl methylcellulose, and CPVD is crospovidone).

A strain-rate study was performed on both MHC300G/Mg stearate (0.25%) and a blend of MCC/HPMC (2%)/CPVD (9%)/Mg stearate (0.25%). The tablet properties and the tableting parameters were monitored when the turret-rotation speed was increased at constant compression force and tablet weight. Increasing the turret rotation speed from 13 rpm to 99 rpm led to very small change in tablet weight (~ 2.0 %) for MHC300G/Mg stearate, indicating that MHC300G shows a very good weight control throughout the full range of the tablet press (see Figure 8). The consistent weight control resulted in little compression-force variation. The tablet strength (i.e., breaking force) of the MHC300G compacts decreases with ~ 20% from 13 to 50 rpm showing a moderate strain-rate sensitivity at relatively low turret speeds, and levels off showing no strain-rate sensitivity for turret speeds ranging from 50 to 99 rpm (see Figure 9). Conversely, the blend of MCC/HPMC/CPVD/Mg stearate (0.25%) shows overall higher strain-rate sensitivity as evidenced by a 34% decrease in tablet weight (see Figure 8) and a 35% decrease in tablet strength (i.e., breaking force) when varying the turret speed from 13 to 99 rpm (see Figure 9). This is consistent with MCC being a ductile material that undergoes plastic deformation (25).

Figure 9: Variation in tablet breaking force with turret rotation speed for tablets (10 mm diameter) prepared from MHC300G/0.25%Mg stearate blend, and MCC/2%HPMC/9%CPVD/0.25%Mg stearate blend (MCC is microcrystalline cellulose, HPMC is hydroxypropyl methylcellulose, and CPVD is crospovidone).

The strain-rate studies for formulations containing 30% IBU (see Table VI) indicated that both blends containing MHC300G and MCC show strain-rate sensitivity, with blends based on MHC300G having less strain-rate sensitivity. The decrease in average tablet weight when increasing turret-rotation speed from 13 rpm to 99 rpm (see Figure 10) was 3% for the IBU/MHC300G blend and 6% for IBU/MCC/HPMC/CPVD blend, respectively. This trend is paralleled by the change in tablet-breaking force (see Figure 11) as illustrated by an overall 40% loss of tablet strength for IBU/MCC/HPMC/CPVD/silica/stearic acid and ~ 25% loss of tablet strength for IBU/MHC300G/silica/stearic acid when increasing turret-rotation speed from 13 to 99 rpm. The tablets prepared in these tableting studies have excellent disintegration times, the tablet weight loss during the friability tests being negligible.

Figure 10: Variation in tablet weight with turret rotation speed for tablets (10 mm diameter) prepared from 30%IBU/MHC300G/silica/stearic acid blend, and 30%IBU/Emcocel90/HPMC/CPVD/silica/stearic acid blend (IBU is ibuprofen, HPMC is hydroxypropyl methylcellulose, and CPVD is crospovidone).

MHC300G dilution potential. The dilution potential can be defined as the amount of an active ingredient (API) that can satisfactorily be compressed into tablets with the excipient mixture.

Figure 11: Variation in tablet breaking force with turret rotation speed for tablets (10 mm diameter) prepared from 30%IBU/MHC300G/silica/stearic acid blend, and 30%IBU/Emcocel90/HPMC/CPVD/silica/stearic acid blend (IBU is ibuprofen, HPMC is hydroxypropyl methylcellulose, and CPVD is crospovidone).

The dilution potential of MHC300G was studied using ibuprofen, which is an API that is difficult to tablet by DC. A 62.5% ibuprofen formulation, corresponding to the current commercial dosage, was prepared by using different grades of ibuprofen (20, 40 and 70 µm average particle size, respectively), MHC300G and 0.5% silica. The flowability of these blends was good as indicated by total flowability indices ranging from 76 to 85 and percent-compressibility indices ranging from 17 to 25. A similar blend prepared by using a mixture of MCC, HPMC (2%), CPVD (9%) and ibuprofen (20 µm) had poor flowability with a percent-compressibility index of 31, and a total flowability index of 62 (see Table VII). The API content uniformity for the blends containing PanExcea MHC300G was excellent, and the IBU recovery was in the range of 98–102% with %RSD < 2 (RSD is relative standard deviation).

Table VII: Bulk density, compressibility, and flowability index of IBU/MHC300G/SiO2 and IBU/MCC/HPMC/CPVD/ SiO2.

Conclusion

PanExcea MHC300G was engineered as a performance excipient containing the three main components of a solid-dosage formulation: filler, binder, and a disintegrant. MHC300G is composed of homogeneous spherical particles in which the individual components cannot be distinguished from one another. Its physical properties and the consistency achieved by maintaining good control over the manufacturing process make this performance excipient ideal for use in drug-product manufacturing under the QbD principles. Placebo tableting studies by direct compression on a 10-station tableting machine showed very good weight control throughout the full range of the tablet press. The consistent weight control resulted in steady compression force when the turret speed was increased to 100 rpm. The tableting of MHC300G requires very little lubricant. Under the conditions used in this study, the resulted compacts were strong, with no capping tendency and excellent disintegration times. MHC300G has a very good dilution potential allowing the tableting of blends of up to 62.5% ibuprofen. Extended tabletability studies by direct compression of API/MHC300G blends with various percent content API will be reported in a follow-up paper.

Liliana A. Mîinea is a senior research chemist, Rajendra Mehta is a principal chemist, Madhu Kallam is an applications engineer, James A. Farina is a group leader of research and development (R&D), and Nandu Deorkar* is a director of R&D, all at Avantor Performance Materials (formerly Mallinckrodt Baker), 1904 J. T. Baker Way, Phillipsburg, NJ 08865, tel. 908.213.6720, fax 908.859.6932, nandu.deorkar@avantormaterials.com.

*To whom all correspondence should be addressed.

Submitted: Aug. 18, 2010. Accepted: Nov. 19, 2010.

References

1. D. McCormik, Pharm. Technol. 5, (4) 52–62 (2005) .

2. N.A. Armstrong, Encyclopedia of Pharmaceutical Technology, Informa Healthcare USA, Inc. (New York, Third Edition, 2006).

3. H.G. Kristensen and T. Schaefer, Drug Dev. and Ind. Pharm. 13, 803–872 (1987).

4. A.J. Hlinak, et al., J. of Pharm. Innov. 1, 12–17 (2006).

5. M. Jivraj, L.G. Martini, C.M. Thomson, Pharm. Sci. and Technol. Today, 3, 58–63 (2000).

6. R.J. Roberts, R.C. Rowe, Int. J. of Pharm. 36, 205–209 (1987).

7. K. van der Voort Maarschalk, G.K. Bolhuis, Pharm. Technol. 23, 34–46 (1999).

8. K. van der Voort Maarschalk et al., Int. J. of Pharm. 151, 27–34 (1997).

9. G.K. Bolhuis, N.A. Armstrong, Pharm. Dev. and Technol. 11, 111–124 (2006).

10. G. Bubb, Tablets and Capsules, 7, (2007).

11. IPEC–Americas, Frequently Asked Questions web page, ipecamericas.org/about/faqs, accessed Feb. 16, 2011.

12. N. Deorkar, Tablets and Capsules, 8, 22–26 (2008).

13. L. Mîinea, J. Farina, M. Kallam, N. Deorkar, Pharm. Processing, Feb. 2009, www.pharmpro.com/Articles/2009/03/A-New-World-of-Excipients-for-Oral-Solid-Dosage-Formulation/, accessed Feb. 16, 2011.

14. S.K. Nachaegari, A.K. Bansal, Pharm. Technol. 28 (1) 52–64 (2004).

15. M.C. Gohel, P.D. Jogani, J. of Pharm. and Pharm. Sci. 8, 76–93 (2005).

16. C. Moreton, "Functionality and Performance of Excipients in a Quality-by-Design World: Parts 1–6," a series of papers published in each issue of vol. 12 of American Pharmaceutical Review (2009).

17. N. Deorkar, J. Farina, L. Mîinea, S. Randive, "Directly Compressible High Functionality Granular Microcrystalline Cellulose Based Excipient, Manufacturing Process and Use Thereof" PCT/US2008/011555 (Oct. 2008).

18. Mallinckrodt Baker, product profile for PanExcea MHC300G, www.mallbaker.com/Literature/Documents/PanExcea%20IR%20Product%20Profile%20Lit7001.pdf, accessed Feb. 16, 2011.

19. L. Mîinea, R. Mehta, J. Farina, N. Deorkar, "Granular Excipients Based on Microcrystalline Cellulose: DOE study of a lab-based scale high shear granulation process," poster presentation at "Excipient Fest Americas", (Puerto Rico, April 2008).

20. USP 34–NF 29 General Chapter <1174> "Powder Flow."

21. Carr, R. L. Chem. Eng. 72,163–168, 1965.

22. Hosokawa Micron, product profile for Hosokawa Micron Powder Characteristics Tester PT-S, www.hosokawa.co.uk/powdtest.php, accessed Feb. 16, 2011.

23. S.I.F. Badawy, D.B. Gray, M.A. Hussain, Pharm. Res. 23, 634–640 (2006).

24. T. Suzuki et al., J. of Pharm. and Pharmacol. 53, 609–616 (2001).

25. C.K. Tye, C. Sun, G.E. Amidon, J. of Pharm. Sci. 94, 465–472 (2005).