Effect of High Shear Wet Milling (HSWM) of Recrystallized Ibuprofen Particles on its Flow Properties for Solid Dosage Pharmaceutical Formulation 1
2
1
1
3
Abdur Rashid , Ivan Marziano , Ted White , Tony Howes , Jim Litster and Lian X. Liu
1
1
School of Chemical Engineering, University of Queensland, St Lucia, Qld., 4072, Australia Email:
[email protected] 2
Material Sciences and Oral Products, Pfizer Global Research and Development, Kent CT13 9NJ, UK 3
School of Chemical Engineering, Industrial and Physical Pharmacy, Purdue University, West Lafayette, IN 479 07-2100, USA
ABSTRACT To design advanced pharmaceutical formulations it is essential to control the particle size of the active drug ingredients. This study investigates the particle size reduction by HSWM, (Silverson® mill) of recrystallized ibuprofen particles and how these particles influence its properties during process formulation. The procedures used are as follows: (1) preparing an ibuprofen solution in 30% water – ethanol mixture at 40 °C; (2) recrystallizing it by cooling the solution to 0 °C; (3) reducing the crystal size by in-situ Process Analytical Techniques (PAT) (Labmax® coupled with HSWM) with the particle size measured by a Lasentec Focused Beam Reflectance Measurement, FBRM®. Subsequently the flow properties of the milled ibuprofen particles and its binary mixtures with lactose were measured using a Schulze® RST-XS ring shear tester. Results show that the morphology of ibuprofen was changed from needle to hexagonal like crystals during recrystallization and crystal size was reduced dramatically by the HSWM for 1 hour. The flowability of milled ibuprofen powders is reduced significantly due to its reduced size and change of surface morphology. Mixing the HSWM ibuprofen powders with lactose enhanced its flow properties. However, the increase of the mixture flowability is less significant in comparison to the binary mixtures of lactose with ibuprofen without HSWM.
INTRODUCTION In the pharmaceutical industry, control of the particle size of an Active Pharmaceutical Ingredient (API) is essential for the design of a new drug formulation. For formulations with low aqueous solubilities, small particles with narrow size distribution have many benefits such as enhanced oral bioavailability (Kamahara et al., 2007). The reduction of particle size can often be achieved through dry or wet milling processes. However, dry milling has the drawback of dusting and noise (Lee et al., 2004) and therefore a wet milling method is often used for the size reduction of pharmaceutical ingredients. Direct compression is the most efficient process used in tablet manufacturing because it is fast, simple and comparatively inexpensive. Together with compression properties, the flowability of the powder mixture is one of the most important factors in a direct compression process. When the particle size of pharmaceutical powders is reduced through a milling
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster process, it is essential to look into the flowability of the processed powders if a direct compression process is to be used. Ibuprofen is a non-steroidal drug that is needle-shaped and widely used as an anti-inflammatory analgesic. While there are literature available on the changing of ibuprofen properties such as crystal shape and size through crystallization on its flow properties (e.g., Yu et al., 2010, Rasenack and Mu¨ller, 2002), no literature is found on how milling of ibuprofen affects its flowability. In this work, ibuprofen was used for recrystallization followed by high shear wet milling process. The flow characteristics of the wet milled ibuprofen as well as the binary mixtures of the milled ibuprofen with lactose were studied in detail. EXPERIMENTAL Materials Ibuprofen USP was purchased from Professional Compounding Chemists of Australia Pty Ltd (PCCA, Sydney). The ibuprofen powder particles are needle shaped and have a volume median particle size of 103 µm from Malvern size analysis. A full description of the ibuprofen powders can be found in the work published earlier (Liu et al., 2008). Two type of lactose powders were used for mixing with the HSWM ibuprofen powders. The first type (Lactose I) is a spray-dried lactose powder supplied by Murray Goulburn (Melbourne). The second lactose powder (Lactose II) has a larger particle sizewith a narrow size range (Lactose II). The particle size of the ibuprofen and Lactose powders are listed in Table 1. The HSWM ibuprofen has a similar volume median size to that of LI. Table 1. The particle size and density of the ibuprofen powders and lactose . Samples Commercial ibuprofen (CI) HSWM ibuprofen Lactose I (LI) Lactose II (LII)
d[4,3] (µ µm)
D[v,0.5] (µ µm)
d [v,0.9] (µ µm)
d[v,0.1] (µ µm)
Span
Density (kg/m3)
110.9
70.0
195
25.3
2.43
1118
44.3 52.0 185.8
33.4 34.4 179.0
95.1 82.9 252.0
7.63 10.3 128.1
2.63 2.11 0.692
1118 1579 1579
Crystal preparation The experiments were performed at the Pfizer research laboratories in Sandwich, Kent, UK. The Silverson® HSWM was attached integrally to a Process Analytical Techniques (PAT) level Labmax® crystallizer with internal ATR FTIR for solution analysis and a Lasentec® Focused Beam Reflectance Measurement (FBRM) for particle size monitoring. It was decided to prepare the HSWM sample after recrystallization so that the ibuprofen crystals start with a plate-like shape instead of the needle shape from commercial ibuprofen. A 30% w/w water ethanol mixture was used as the solvent and was saturated with commercial ibuprofen crystals at 25 oC. The decanted saturated solution was held in the crystallizer at 35 oC for 2 h to ensure complete dissolution, and then cooled to 0 oC overnight to give a crop of grown crystals. The crystal slurry was then circulated for 1 hour through the
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster in-line HSWM operating at 6000 rpm using a circular stator screen to give the milled product. Finally, the milled material was filtered and dried for flow testing.
Flowability measurement The method used for flowability measurement was the shear cell test. A ring shear tester (Dietmar Schulze model RST-XS, Wolfenbüttel) was used for all the measurement. The shear tester is designed to measure shear stress τ, at different values of normal stress, σ. A consolidation load of 2 kPa was used for all the flow tests. A full description of the method used can be found from the previous published work (Liu et al., 2008). For ibuprofen binary mixtures, the chosen lactose powder was mixed with the ibuprofen sample at different mass ratios (3;1, 1:1 and 1:3) in a SPEX CertiPrep 8000D mixer (SPEX CertiPrep, Inc., Metuchen, USA) for 10 minutes (no grinding medium was used). The flow properties of the binary mixtures were then measured using the RST-XS shear tester. RESULTS AND DISCUSSION Characteristics of the HSWM ibuprofen Figure 1a shows an optical image of the commercial ibuprofen crystals. As shown in Fig. 1a, the commercial ibuprofen crystals are needle shaped whereas the recrystallised ibuprofen crystals (Figure 1b) are plate-shaped. The HSWM ibuprofen (Figure 1c) is much smaller and the fragments are of more equi-dimensional shape.
(a)
(b)
(c)
Figure 1. Optical images of (a) commercial ibuprofen; (b) recrystallized ibuprofen before milling; (c) the milled ibuprofen. The scale bars are 50, 200 and 50 µm respectively. As mentioned, FBRM was used for measuring the ibuprofen crystal size in-line in the crystallizer. Figure 2 shows the frequency (count/s) versus the chord length of the recrystallised ibuprofen crystals before milling and after milling. Although FBRM does not give an absolute measure of particle size, it is an excellent tool for on-line monitoring of particle size. As shown in Figure 2, the ibuprofen crystal size is reduced significantly after milling. The corresponding length square weight mean chord length (from the FBRM in the crystallizer) is 28.9 µm, in comparison to the recrystallized ibuprofen before milling of 103.2 µm (Table 2). So the ibuprofen after HSWM process is of considerably smaller size (by about one third).
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster From Malvern size analyses, the commercial crystals had a volume median size of 70 µm with a span of 2.43 [(D90% – D10%) / D50%, where DX% is the size for which X % of the total by volume is smaller], while the HSWM product has a volume mean of 33.6 µm and a span of 2.64 (Table 1). Table 2. Statistical results of FBRM data before and after milling. Chord length Before milling (µm) After milling (µm) (HSWM) C10 35.9 10.22 C50 103.2 28.9 C90 196.8 73.2
Figure 2. In-line particle sizing of ibuprofen particles by FBRM.
Flowability of single compound Table 3 shows the flow results of both the HSWM and the commercial ibuprofen. The last row in Table 3 gives, as a percentage, the estimated 95% uncertainties (errors) on the average results of triplicated measurements. Generally the measurement errors are low. As can be seen from Table 3, both the commercial ibuprofen and the HSWM ibuprofen have flow functions between 2 to 4, which fall into the cohesive powder range as defined by Jenike (1964). Although the HSWM ibuprofen is rounder in shape in comparison to the needle shaped commercial ibuprofen, its flowability is similar to the commercial ibuprofen. This is because the particle size of the HSWM ibuprofen is smaller and the powder flowability reduces with the decrease of particle size (Liu et al., 2008). In addition, the surface morphology of the ibuprofen crystals may have been changed by the milling process. The flow properties of the two types of lactose powders are also listed in Table 3. Lactose II has a very high flow function of 14 (easy flow) due to its large particle size whereas the Lactose I is also cohesive in nature with a flow function of 3.1. Although Lactose I has a similar particle size to HSWM ibuprofen, it flows better than the HSWM ibuprofen, which is probably due to various other factors such as particle-particle adhesion forces, inter-particle
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster separation distances and the particle surface asperity (hr) as discussed in the previous work (Liu et al., 2008).
Table 3. Flow results of single compounds.
Samples
CI HSWM LI LII Est % error
Consolidation stress
Unconfined yield stress
Flow Function
σ 1 (Pa)
FC (Pa) 1482 1886 1263 269 11
FCC 2.5 2.3 3.1 14 10
3727 4326 3916 3882 3
Angle of Bulk density internal from shear cell friction
φi (°)
ρb (kg/m3)
41 49 43 38 3
489 442 693 826 2
Flowability of binary mixtures of ibuprofen and lactose Table 4 summarises the flow test results for the binary mixtures of ibuprofen and the two types of lactose. Figure 3 plots the flow function of the binary mixtures of both the commercial ibuprofen and HWSW ibuprofen with Lactose I versus lactose content. The flow functions of the ibuprofen (0% Lactose I) and Lactose I (100%) are also included. Interestingly, when 25% of Lactose I (cohesive by itself) is mixed with the HSWM ibuprofen and commercial ibuprofen, the flow function for both of the mixtures increased significantly, which makes the mixtures fall into the easy flow region (flow function between 4 to 10). For the commercial ibuprofen, the highest measured flow function is reached at a lactose content of 50% and then it reduces significantly when the lactose content is increased to 75%. On the other hand for the HSWM ibuprofen, the highest measured flow function occurs at 25% of lactose and then gradually reduces as the lactose contents increases. The reason for the increased flowability when Lactose I is mixed with ibuprofen could be caused by two factors. Firstly, Lactose I has similar particle size to the ibuprofen particles, particularly to the HSWM ibuprofen. Therefore, potential segregation between ibuprofen and lactose is minimized. When there is no segregation between the two different powders, the flowability of the mixture is less likely to be determined by any one of the powders. Secondly, better cohesion force between ibuprofen and lactose particles could exist, which is equivalent to an increase of the mixture particle size. The reason that the maximum flow function is obtained at around 25% to 50% of lactose by mass is not clear as the volume ratio of the lactose to ibuprofen particles is 0.71 at the mass ratio of 1:1 based on the density of the two components. That is, the volume of lactose in the mixture is rather small (17.7%) at the mass content of 25%. It is also noted that, although the flow function between the commercial ibuprofen and the HSWM ibuprofen is similar statistically, the flow function increase after adding Lactose I is much smaller for the HSWM ibuprofen. This indicates that particle surface morphology must have played an important role in determining the powder flowability. Figure 4 plots the internal angle of friction of the binary mixtures of ibuprofen versus Lactose I content. The internal friction angle of a powder is a measure of its internal friction during
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster steady flow (or its ability of maintaining a constant flow) and the lower the angle the better the ability for constant volume flow. As shown in Figure 4, the lowest internal friction angle was obtained at the Lactose I content of 25%. This is consistent with the highest flow function value at around 25% of lactose content (Figure 3).
Table 4. Summary of the flowability test results for binary mixtures of ibuprofen and lactose.
Consolidation Unconfined stress yield stress
Flow Function
Angle internal friction
Bulk density from shear cell
σ 1 (Pa)
FC (Pa)
FCC
φi (°)
ρb (kg/m3)
3391
778
4.4
37
3708
1164
3.2
42
3878
1296
3.0
44
3478
1236
2.8
41
3729
951
3.9
40
3741
590
6.3
38
CI:LII (3:1)
3937
670
5.9
39
CI: LII (1:1)
4233
655
6.5
41
CI:LII (1:3)
4029
1409
2.9
46
Mixtures (Mass ratio)
HSWM:LI (3:1) HSWM:LI (1:1) HSWM:LI (1:3) HSWM:LII (3:1) HSWM:LII (1:1) HSWM:LII (1:3)
564
662 661
601 714 784
600 641 674
Figure 5 plots the flow function of mixtures of HSWM and commercial ibuprofen with Lactose II. It is seen that the flow function of the mixtures for both the commercial ibuprofen and the HSWM ibuprofen generally increases with the increase of lactose content. For the commercial ibuprofen the increase is almost linear with lactose content. However, the increase of the flow function for binary mixtures of the HSWM Ibuprofen with Lactose II is much slower initially. As the flow function only increases slightly with lactose content up to 50%, its flow function is still below 4 and within the cohesive powder region. Whereas the flow functions of all the measured binary mixtures of the commercial ibuprofen and lactose have flow function above 5, falling in the easy flow region. These results show that the recrystallization and HSWM process not only reduces the particle size, but also likely have changed the surface properties of the ibuprofen particles (such as roughness, electrostatic charge), which subsequently affects its interactions with lactose particles. Figure 6 plots the mixture internal friction angle at different mass content of Lactose II. The internal friction
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster angle of the mixture decreases with the increase of the lactose content, which is consistent with the flow function results. With the HSWM sample, it is noticed that the flowability of its binary mixtures with 25% of the free flowing Lactose II is worse than the mixture with 25% of Lactose I which has a much lower flowability. This could be caused by the large size difference between the lactose and ibuprofen particles. If so, industry should make sure that both the APIs and the excipients have similar particle sizes so that segregation of powders can be avoided.
7 6 Flow function
5 4 3 2
Commercial ibuprofen
1
HSWM ibuprofen
0 0
20
40
60
80
100
Lactose content (%)
Figure 3. Flow function of binary mixtures versus Lactose I content.
50
Internal friction angle (°)
46
42
Commercial ibuprofen
38
HSWM ibuprofen
34
30 0
20
40
60
80
100
Lactose content (%)
Figure 4. Internal friction angle of binary mixtures versus Lactose I content.
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster This work illustrates the complex relationship between the flowability of mixtures and their individual components. The flowability of a binary mixture is not only dependent on the particle size of each individual compound (and therefore the flowability of each single compound), but also depends on the interactions between the particle-particle bonds between the two compounds as well as the content of each compound.
Figure 5. Flow function of binary mixtures versus Lactose II content.
Figure 6. Internal friction angle of binary mixtures versus Lactose II content.
CONCLUSIONS
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster In this work, ibuprofen size reduction by recrystallization followed by a high shear milling process was carried out. The flow properties with and without lactose were studied. The results show that the morphology of ibuprofen was changed from needle to hexagonal like crystals during recrystallization and crystal size was reduced dramatically from a volume mean size of 70 µm to 33 µm by HSWM for 1 hour. Flow test results on the HSWM ibuprofen showed that its flowability is reduced significantly due to its reduced size and possibly change of surface morphology. Mixing the HSWM ibuprofen powders with both a fine and a coarse lactose powder enhanced its flow properties but the increase with the fine lactose powder is much higher than that with the coarser lactose powder. The maximum measured mixture flowability with the fine lactose was achieved at a mass content of 25%. However, the increase of the mixture flowability is less significant in comparison to the binary mixtures of ibuprofen without HSWM. ACKNOWLEDGEMENT This research is supported by Pfizer Pty. Ltd. whose support is gratefully acknowledged. REFERENCES Kamahara, T., Takasuga, M., Tung, H. H., Hanaki, K., Fukunaka, T., Izzo, B., Nakada,J., Yabuki, Y. and Kato, Y. (2007). Generation of fine pharmaceutical particles via controlled secondary nucleation under high shear environment during crystallization, Process Development and Scale-up, Organic Process Research & Development , 11, 699-703. Jenike, A.W. (1964). Storage and Flow of Solids, Bull. No. 123, Eng. Exp. Station, Univ. Utah, Salt Lake City. Lee, I., Variankaval, N., Lindemann, C. and Starbuck, C. (2004). Rotor-stator milling of APIs - Empirical scale-up parameters and theoretical relationships between the morphology and breakage of crystals, Am. Pharm. Rev., 7(5) 120-123. Liu, L.X., Marziano, I., Bentham, A.C., Litster, J.D., White, E.T. and Howes, T. (2008). Effect of particle properties on the flowability of ibuprofen powders, International J. Pharmaceutics, 362, 109-117. Rasenack, N. and Mu¨ller, B.W. (2002), Properties of Ibuprofen Crystallized UnderVarious Conditions: A Comparative Study,Drug Development and Industrial pharmacy, 28 (9),1077– 1089. Yu, W., Muteki, K., Zhang, L. and Kim, G. (2011), Prediction of bulk powder flow performance usingcomprehensive particle size and particle shape distributions, Journal of Pharmaceutical Sciences, 100, 285-293.
A.Rashid, L.X.Liu, I. Marziano, E.T. White, T. Howes and J.D.Litster
BRIEF BIOGRAPHY OF PRESENTER Dr. Lian Liu Dr. Lian Liu graduated with a PhD in Chemical Engineering from the University of Queensland. Her main research interest is in the areas of particle technology, particularly in granulation and compaction. At present she is a senior research fellow in the School of Chemical Engineering , The University of Queensland.
