ORIGINAL ARTICLES
College of Pharmacy1, University of Louisiana at Monroe, USA, North-West University2, Potchefstroom, South Africa, and Departments of Chemical Engineering, Chemistry, and Materials Science and Engineering3, The Pennsylvania State University, School of Pharmacy4, University of Wisconsin, Madison, USA
Increased dissolution and physical stability of micronized nifedipine particles encapsulated with a biocompatible polymer and surfactants in a wet ball milling process Ning Li1, M. D. DeGennaro1, W. Liebenberg2, L. R. Tiedt2, A. S. Zahn3, M. V. Pishko3, M. M. de Villiers4
Received September 5, 2005, accepted October 5, 2005 Melgardt M. de Villiers, Division of Pharmaceutical Sciences, School of Pharmacy, University of Wisconsin, Madison WI 53705, USA
[email protected] Pharmazie 61: 595–603 (2006)
Suspensions of nifedipine, a practically water-insoluble drug, were prepared in the presence of a biocompatible polymer, polyvinylpyrrolidone (PVP, K value 17), and three surfactants, sodium lauryl sulfate (SLS, anionic), cetyltrimethylammonium bromide (CETAB, cationic), polysorbate 80 (Tween 80, nonionic), by wet milling in ceramic ball mills. Nifedipine powders encapsulated with PVP and the surfactants were recovered from the suspensions after milling and evaluated for changes in particle size, morphology, sedimentation rate in aqueous suspensions, crystal form, and dissolution. Particle size analysis indicated that milling of suspensions in solutions of PVP and surfactants is an efficient method for reducing the particle size of nifedipine to below 10 mm. Furthermore, DSC and XPS analysis indicated that during milling the nifedipine crystals were coated with the PVP or surfactants and that milling with PVP stabilized the nifedipine crystal form during milling while nifedipine was gradually amorphisized when milled in a quaternary nifedipine/PVP/SLS/CETAB system. The decrease in particle size caused a significant decrease in sedimentation rate and increased the dissolution rate of nifedipine in simulated gastric fluid when compared to milled nifedipine and powder mixtures of the drug and the excipients.
1. Introduction Few materials used in the manufacturing of pharmaceuticals exist in the optimum size, and most materials must be reduced in size at some stages during the production of a dosage form. Milling is the mechanical process for reducing the particle size of solids. Fine milling, particles smaller than 50 mm, is often achieved by ball milling (Parrott 1976; Brain 1994). A ball mill consists of a horizontally rotating hollow vessel of cylindrical shape with the length slightly longer than the diameter. The mill is partially filled with balls of pebbles, which act as the grinding medium. At the optimum speed, the balls roll and cascade over one another to provide an attrition action while the balls are also carried up the side of the mill and fall freely onto the material with an impact action. Milling is therefore a combination of impact and attrition depending on the speed of the mill. The critical speed of a ball mill is the speed at which the balls just begin to centrifuge with the mill because at or above this speed no significant size reduction occurs. Normally ball mills are operated at 60–85% of the critical speed and in general lower speeds are for finer grinding. An empirical rule for the optimum speed of a ball mill is given by n ¼ 57 40 log D, where n is the speed in revoPharmazie 61 (2006) 7
lutions per minute and D is the inside diameter of the mill. For a given speed, smaller balls will give a slower but finer grinding (Voller 1983). Ball mills are operated either wet or dry. Normally with wet milling particles smaller than 50 mm can be produced from slurries containing 30–60% solids (Staniforth 2002). From the viewpoint of power consumption, wet grinding is more efficient than dry grinding. To produce even finer particles small amounts of grinding aids can be added to the mill (Apte et al. 2003). However, the use of grinding aids for pharmaceutical products is limited by the physiological and toxicological restrictions on medicinal products. For example, some dispersing agents have been added to wet milling where the addition of as little as 0.1% of a surface active agent may increase the production rate of a mall mill 20 to 40% (Parrott 1976). Grinding aids that have been used for grinding drugs include sodium chloride for dexamethasone, amorphous magnesium aluminosilicate for ketoprofen, indomethacin, naproxen and progesterone, sodium lauryl sulfate for ursodeoxycholic acid, polyvinyl pyrrolidone for sulfathiazole, chitin and chitosan for griseofulvin, phenobarbital, prednisolone, flufenamic acid, and indomethacin, and amorphous silicon dioxides for prednisone, digoxin and griseofulvin (Apte et al. 2003; Gupta et al. 2003; Chung et al. 595
ORIGINAL ARTICLES
2. Investigations, results and discussion 2.1. Milling process In this study two ceramic ball mills, 16 cm diameter with 20 cm length (optimum speed 68 rpm) and 8 cm diameter with 10 cm length (optimum speed 80 rpm) respectively were used. The ceramic round balls used in the large ball mill had a diameter of 3.5 cm and in the small ball mill 1.5 cm. Each ball mill was filled to about 30% with grinding media. The large ball mill was operated at optimum speed and the small mill at the optimum speed and 40 rpm (50% of the optimum speed). Aqueous suspensions of nifedipine powder, mean volume particle size 596
76.5 mm, with and without PVP and/or surfactants, were milled for up to 48 h. 2.2. Changes in particle size caused by milling
10
100
8
80
6
60
4
40
2
20
0
Cummulative undersize (%)
Frequency (%)
To test the applicability of the chosen speeds of the ball mills, the decrease in particle size of nifedipine when milled in the small mill at 40 rpm and 80 rpm was measured. The optimum speed for this mill was calculated to be 80 rpm; 40 rpm represents 50% and 80 rpm represents 100% of the optimum speed. The results show that for the 0.5% nifedipine suspensions used in this study the increase in speed did not significantly change the rate at which and the mean size obtained after milling for 48 h. Secondly, to determine the effect of the size of the ball mill on the milling process, the decrease in nifedipine particle size when milled in a small (500 ml) versus a large (4000 ml) amount was determined. The results show that when these mills were rotated at 100% of the optimum speed there was not a significant difference in the particle size obtained after milling for 48 h. The mean volume particle sizes of nifedipine after milling in the small and larger ball mill were 46.6 1.5 and 48.3 15.4 mm, respectively. For the combination of nifedipine and PVP or surfactant, Tween 80 or SLS or CETAB, the mean volume particle size in small and large mill are 4.2 1.2 and 7.8 6.6, 4.4 0.98 and 2.7 1.6, 4.2 0.87 and 2.1 1.1, 2.7 0.7 and 2.4 1.1 respectively. Based on these results, further experiments were performed in the small mill rotated at 80 rpm.
0 10
100 Particle sice (µm)
16 Mean volume pa rticle size (µm)
2002; Boldyrev at al. 1994; Yang et al. 1979). However, although all these adjustments can increase the efficiency of wet milling, flocculation restricts the lower limit to approximately 10 mm. Another drawback of milling pharmaceuticals is the change in crystal form upon milling (Sato et al. 1997; Itoh et al. 2003; Gupta et al. 2003). In one study by Sato et al. (1997) the pharmaceutical properties of ground mixtures of nifedipine with casein, magnesium silicate, and cellulose acetate phthalate in a vibrational ball mill was reported. The X-ray powder diffraction patterns and DSC data suggested that nifedipine was present in its amorphous form in these ground mixtures. Grinding increased the wettability, the solubility in water and bioavailability after oral administration of nifedipine. However, the authors did not look at the stability of the amorphous drug particles. This study added to results reported earlier that showed roll mixing of nifedipine with PVP produced amorphous drug with increased dissolution (Nozawa et al. 1986). The biggest drawback of this approach is that amorphous nifedipine is not stable (Caira et al. 2003). In another study, Itoh et al. (2003) reported that when the poorly water-soluble drug nifedipine was ground in a dry process with polyvinylpyrrolidone (PVP) and sodium dodecyl sulfate (SDS) different crystallinity, predominantly amorphous, behavior was shown in the ternary drug/PVP/ SDS system. However, when the ternary drug/PVP/SDS ground mixture was added to distilled water, crystalline nanoparticles, which were 200 nm or less in size, were formed that had excellent stability. Zeta potential measurement suggested that the nanoparticles had a structure where SDS was adsorbed onto the particles that were formed by the adsorption of PVP on the surface of drug crystals. The stable existence of crystalline nanoparticles was attributable to the inhibition of aggregation caused by the adsorption of PVP and SDS on the surface of drug crystals and to the electrostatic repulsion due to the negative charge of SDS on a shell of nanoparticles. However, the yield of nanoparticles produced by this process was low (500 mg/ml). In an effort to increase the yield of stable, crystalline, micronized nifedipine particles this study evaluated wet ball milling in the presence of polyvinylpyrrolidone, which was used to either stabilize or wet the drug solids, and a series of ionic and nonionic surfactants, cetyltrimethyl ammonium bromide (CETAB), sodium lauryl sulfate (SLS), and polysorbate 80 (Tween 80), which were introduced as wetting agents to improve the milling efficiency. Evaluation of the reduction in particle-size, the coating process, the physical stability of micronized drug crystals, and the dissolution rates of the nifedipine ball-milled products were used to determine the efficiency of the milling process.
12
8
4
0 0
10
20
30
40
50
Time (hours)
Fig. 1: Particle size distribution of nifedipine (top) without milling, (bottom) changes in particle size upon milling for 1–48 h in the presence of PVP and surfactants ^ Nifedipine þ PVP; ~ Nifedipine þ Tween 80; & Nifedipine þ SLS; * Nifedipine þ CETAB; ~ Nifedipine þ PVP þ Tween 80; & Nifedipine þ PVP þ SLS; * Nifedipine þ PVP þ CETAB
Pharmazie 61 (2006) 7
ORIGINAL ARTICLES
The nifedipine powder used in this study consisted of crystalline particles with a mean volume particle size of 76.5 6.6 mm. The particle size distribution of this powder (Fig. 1) was multi-modal with a mixture of large particles (100–200 mm) and very small particles (5–20 mm). Milling a 0.5% suspension in water of this water-insoluble powder for up to 48 h in the small ball mill at 80 rpm only reduced the mean particle size to 50 mm and changed the particle size distribution to bi-modal distribution around two means of 20 and 70 mm. The addition of 1% PVP to this suspension significantly reduced the particle size because after 48 h, the mean volume size was 4.2 mm as shown in Fig. 1. The milling was very efficient because even after an hour the mean size was already reduced to about 7 mm. The decrease in particle size followed first order kinetics with a rate constant of 3.8 h1 (Table 1). Milling in the presence of SLS, CETAB and Tween 80 (0.5%) also significantly reduced the particle size by a first-order process. The speed by which the particle size was reduced in the presence of these surfactants (Table 1) was slower (mean 2.37 h1) than when PVP was added. The combination of PVP and the surfactants sped up the particle size reduction process for CETAB and Tween 80 but not for SLS. These results showed that the addition of the grinding aids significantly increased the efficiency of the wet milling process and that there was not a significant difference in the effect on the particle size with regard to the addition of a single grinding aid versus combinations of PVP and the surfactants. In addition, although in this study milling was performed for up to 48 h, results showed that there was not a significant difference in the mean size obtained after 24 h compared to 48 h. Comparison of the particle size distribution obtained after 48 h for nifedipine milled in the presence of the surfactants or combinations of PVP and surfactants showed that for SLS there was not a significant difference in the mean size. In addition, both processes produced uni-modal size distributions with very few particles below 1 mm. These results are in contrast with the results reported by Itoh et al. (2003) that showed for a dry milling process nifedipine particles below 1 mm was produced when the milled powder was suspended in water. When Tween 80 was added the particle size distribution was uni-modal after milling for 48 h but when PVP and Tween 80 was added a bi-modal distribution was produced with a significant number of particles below 5 mm. This effect was even more pronounced when CETAB or PVP plus CETAB was added to the suspensions because bimodal distributions Table 1: Kinetic parameters describing the first-order decrease in the particle size of nifedipine during wet ball milling in the small ball mill Product
Initial size, b (mm)
Final size, a (mm)
Rate, c (h1)
R2
Nifedipine Nifedipine þ PVP Nifedipine þ Tween 80 Nifedipine þ SLS Nifedipine þ CETAB Nifedipine þ PVP þ Tween 80 Nifedipine þ PVP þ SLS Nifedipine þ PVP þ CETAB
76.30 71.92 69.36 71.42 71.92 71.38 72.02 72.82
48.94 4.58 7.09 5.07 4.57 5.12 4.48 3.68
0.67 3.76 2.32 2.22 2.57 3.25 2.25 3.44
0.917 0.999 0.990 0.999 0.997 0.997 0.999 0.999
Data obtained by fitting a first-order decay equation, y ¼ a þ becx, using Table Curve 2 D v.4 (Systat Software, Inc., Point Richmond, CA, USA)
Pharmazie 61 (2006) 7
Fig. 2: SEM pictures of un-milled and 0.5% nifedipine suspensions milled for 48 h in the small ball mill at 80 rpm without and with adding 1% PVP
with significant numbers of particles below 2 mm was observed. In this study, it was not possible to explore the properties of the small particles because the particle-sizing instrument used in this study has a lower size limit of 597
ORIGINAL ARTICLES
Fig. 3: SEM pictures of 0.5% nifedipine suspensions milled for 48 h in the small ball mill at 80 rpm without and with adding 0.5% surfactant and 1% PVP
0.5 mm. Not withstanding this limitation it seemed as though it was possible with the addition of Tween 80, and especially CETAB combined with PVP, to produce significant populations of very small nifedipine particle,
