Impact of process parameters on Mg-St content and tablet surface. wettability in the external lubrication method for a rotary tablet press

Transcription

1 1 2 Impact of process parameters on Mg-St content and tablet surface wettability in the external lubrication method for a rotary tablet press Takayuki Kamiya a,b, *, Hisami Kondo a, Hiroyuki Hiroma a, Kazunari Yamashita a, Tadashi Hakomori a, Kazuhiro Sako a, Yasunori Iwao b, Shuji Noguchi b Shigeru Itai b a Pharmaceutical Research and Technology Laboratories, Astellas Pharma Inc., 180 Ozumi, Yaizu, Shizuoka , Japan b Department of Pharmaceutical Engineering, Graduate School of Pharmaceutical Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka , Japan *Address correspondence to: Takayuki Kamiya Pharmaceutical Research and Technology Laboratories, Astellas Pharma Inc., 180 Ozumi, Yaizu, Shizuoka , Japan. Tel: , Fax: takayuki.kamiya@astellas.com 21 1

2 Abstract External lubrication is a new alternative used in compression processes of the pharmaceutical industry to minimize the negative effect of lubricant and to resolve sticking problems. This method can also prevent the deterioration of tablet properties; e.g., reduced tensile strength and a retarded rate of drug dissolution. The current study prepared tablets using the external lubrication method with varying potential critical process parameters and clarified the process of external lubrication via statistical analysis. In accordance with past results, tablets prepared using the external lubrication method had better tablet properties (higher hardness and more rapid disintegration) than those prepared using the internal lubrication method. Quantitative analyses of the magnesium stearate content and contact angle showed that the wettability of the tablet surface increased with the magnesium stearate content. Analysis of variance showed that all potential critical parameters are influential for the magnesium stearate content, but not for the disintegration of tablets. In addition, predictions of the magnesium stearate content and disintegration time of tablets made using a model equation correlated well with observed values. The external lubrication method can thus be applied by identifying the critical process parameters in the compression process during the development and manufacture of pharmaceutical products Keywords: External lubrication, tablet lubrication, rotary tablet press, lubricant concentration, magnesium stearate 2

3 Introduction Lubricants such as magnesium stearate (Mg-St) are essential additives for the formulation of tablets in the pharmaceutical industry. The main reason for adding lubricants to pharmaceutical ingredients is to improve the powder flowability and to prevent tableting problems associated with sticking, picking, capping and lamination. Although Mg-St is usually blended with a powder or granules, immediately prior to compression at a concentration ranging from 0.25% to 1.5%, 1,2 an excessive amount or excessive mixing time of Mg-St has been reported to deteriorate the tablet quality in terms of delayed tablet disintegration, slower drug dissolution and reduced tablet hardness. 1,3,4 This deterioration in quality might be explained by the surface coverage of a hydrophobic film of Mg-St on the powder or granules, which can reduce the wettability and water permeability of the material while weakening the binding interactions between particles. 5 8 In addition, the physicochemical properties of Mg-St can differ from batch to batch, 9 and the blending conditions and concentration of the lubricant should thus be carefully determined during the development of pharmaceutical formulation and manufacturing processes to ensure consistent product quality and a high level of process understanding. As an alternative to the internal Mg-St lubrication method, an external lubrication method using Mg-St has attracted attention because it reduces the manufacturing time by omitting lubricant blending, minimizes the negative effect of the lubricant and resolves sticking problems. A lubricant is mixed with compressed air at a constant rate and is continuously sprayed onto the surfaces of the punches and dies via a dedicated nozzle during the time interval between the discharge of tablets and the filling of the dies with powdered materials. 2,10 12 Any excess lubricant is removed using a vacuum dust collector near the spraying nozzle. The important advantage of the external lubrication method is that the 3

4 concentration of lubricant incorporated into each tablet is less than that when using the internal Mg-St lubrication method 10,11. In fact, a thin layer of lubricant on the tablets has been reported to form when implementing the external lubrication method, thus preventing the sticking problem. 11 In addition, the external lubrication method has been demonstrated to reduce the capping tendency to a level comparable with that achieved with the internal lubrication method at a typical Mg-St concentration by determining the residual and maximum die wall pressures of a single punch tablet press. 13 However, previous studies using a rotary tablet press determined only relationships between a manufacturing condition and/or Mg-St content in the tablets and the tablet property. Thus, very little information is available regarding the quantitative analysis and practicality of the external lubrication method, and our overall level of process understanding therefore remains limited. Furthermore, no quantitative information on the wettability of the tablet surface subjected to the external lubrication system is available, although wettability studies are often conducted to evaluate the lubrication property of a tablet surface in the case of the internal lubrication method. Such studies would shed light on the mechanism of the external lubrication method compared with that of the internal method. The ICH Q8 document (International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use, Pharmaceutical Development) recommends that critical process parameters (CPPs) be specified and appropriately controlled according to the results of multivariate experiments. In this way, it is possible to understand the drug products and manufacturing processes according to a systematically associated functional understanding of the material properties and process parameters in terms of the critical quality attributes of the drug products. 14 We previously evaluated six process parameters for controlling the concentration of lubricant added by the 4

5 external lubrication system and demonstrated that the spray rate of the lubricant, air volume of the dust collector, and rotation speed of the rotary press were potential CPPs. 15 In the present study, we investigated the criticalities of these potential CPPs for the external lubrication method in terms of the characteristics of the resulting tablet products by evaluating the relationships between the wettability of the tablet surface, disintegration property or hardness and the lubricant concentration of the whole tablet. Furthermore, a simple model equation that predicts the amount of lubricant in tablets was established from the results of statistical analyses. 5

6 Material and method Reagents and chemicals Mg-St of normal grade was purchased from Taihei Chemical Industrial Co. Ltd. (Osaka, Japan). Lactose monohydrate (Di-lactose R) was purchased from Freund Corp. (Tokyo, Japan), and microcrystalline cellulose (Ceolus TM PH-101) was purchased from Asahi Kasei Chemicals Corp. (Tokyo, Japan). Nitric acid (60%; specific gravity, 1.38) and a standard solution of magnesium (Mg1000) were purchased from Kanto Chemical Co., Inc. (Tokyo, Japan) Overview of the external lubrication system Mg-St was dispersed into the external lubrication system (EXTALUB, Hata Iron Works, Kyoto, Japan) using compressed air, and the resulting material was transferred via a tube to a dedicated spraying nozzle ready to be sprayed onto the punches and dies used in the rotary tablet press (HT-X20, Hata Iron Works). The spray rate (g/min), which is the weight of lubricant consumed by the feeding of the nozzle per unit time, was measured using standard weighing equipment Monitoring of the degree of dispersion and concentration of lubricant The degree of dispersion of the lubricant was monitored using a laser sensor located at the point where the lubricant was dispersed into compressed air in the external lubrication system. The degree of dispersion can be expressed as 100 (1-I/I0)(%), where I is the intensity of the light transmitted by the lubricant-dispersed air and I0 is the intensity of the incident light. The light path length was 100 mm. The concentration of lubricant (mg/l) in 6

7 the air flow was calculated from the spray rate (g/min) and the inlet dispersion air volume (L/min) Preparation of tablets using the external lubrication method The powder for compression was prepared by blending a 4:1 (w/w) mixture of lactose monohydrate with microcrystalline cellulose using a diffusion mixer (container mixer PM200, L.B. Bohle, Ennigerloh, Germany). The tablets were prepared using a rotary tablet press and the external lubrication system. The rotary tablet press HT-X20 used in this study had 20 stations, and its maximum rotation speed was 70 rpm. All 20 stations were used for preparation of tablets using the external lubrication method. The gravity feeder was used for providing the powder for compression on the turntable. The diameter and curvature radius of the punches used in the present study were 8.0 and 12.0 mm, respectively. The powder was compressed to form tablets of 180 mg in weight and 3.3 mm in thickness. The spray rate, flow air volume of the dust collector and rotation speed of the rotary tablet press were selected as variable parameters from the parameters of the external lubrication system and rotary tablet press, because these have been previously demonstrated to affect the amount of Mg-St in the tablets. 15 Their values are given in Table 1. For all samples, the tablets analyzed were those manufactured 10 min after initiating the compression process Preparation of tablets using the internal lubrication method The mixed powder (99 wt%) described above containing lactose monohydrate and microcrystalline cellulose was blended with 1 wt% of Mg-St for 10 minutes at 20 min 1 using a 20-L diffusion mixer (container mixer LM20, L.B. Bohle, Ennigerloh, Germany). The resulting blend was used to evaluate the compression process of the conventional internal 7

8 lubrication method. Tablets were prepared using the same method used for the preparation of tablets with the external lubrication method, except that only 5 of the 20 rotary press stations were used at a rotating speed of 30 min 1. A smaller gravity feeder was used to minimize the effect of and circulation on the turntable Preparation of tablets without lubricant A small portion (180 mg) of the mixed powder containing lactose monohydrate and microcrystalline cellulose was manually compressed to a thickness of 3.3 mm using a precision universal tester (Autograph, AGS-20KNG, Shimadzu Corporation, Kyoto, Japan) Analysis of Mg-St content Nitric acid solution (0.1 mol/l) was prepared by diluting 60% nitric acid with distilled water. Ten tablets were disintegrated in approximately 35 ml of 0.1 mol/l nitric acid using ultrasonic irradiation, and the resulting mixtures were placed in a water bath at 70 to 80 C for 20 min. After being cooled to room temperature, the samples were diluted to 50 ml with 0.1 mol/l nitric acid and centrifuged at 1900 g for 10 min. The concentration of Mg in the supernatant was then quantified using an atomic absorption spectrophotometer (Z-2000, Hitachi High-technologies Corporation, Tokyo, Japan) to determine the Mg-St content of the tablets. The frame method was used for this analysis, and the wavelength was set to nm. A calibration curve was established by preparing several magnesium solutions having different concentrations from Mg1000 and 0.1 mol/l nitric acid. A control solution was also prepared from the powder used for the compression but without Mg-St to determine the baseline response for each test solution

9 Analysis of tablet properties One hundred tablets were randomly sampled for weight analysis, and their individual weights were measured using an electric scale. The mean weight and relative standard deviation were then calculated. Thicknesses were measured for 10 randomly selected tablets using a thickness meter (Digital linear gauge DG-825, Ono Sokki, Kanagawa, Japan). Hardnesses were measured for 10 randomly selected tablets using a Schleuniger hardness meter (Dr. Schleuniger 8M, Pharmatron, Thun, Switzerland). A disintegration test was conducted using a disintegration time tester (NT-200, Toyama Sangyo Co., Ltd., Osaka, Japan) according to the description provided in the 16th Edition of the Japanese Pharmacopeia. Distilled water was used as the medium. In the analysis of tablets without any lubricant, 10 tablets were tested for weight and thickness, five tablets for hardness and six tablets for disintegration Contact-angle analysis The contact angle was measured for the sample tablets manufactured using the external lubrication method. The effect of lubricant added via the external lubrication system on the hydrophobicity of the surface of the tablet was then evaluated. Tablets manufactured using the internal lubrication method and tablets containing no lubricant were used as reference samples. The contact angles made by droplet of water on either upper or lower surface of the tablets were measured using the 2 method with a contact angle meter (Type CA-V, Kyowa Interface Science Co., Ltd., Saitama, Japan). The measurement was performed after 100 ms of placing the water droplet Calculation of Mg-St layer thickness 9

10 Compressed Mg-St was prepared from the approx. 200 mg of Mg-St with 8 kn pressure using 8 mm die, flat punches and a precision universal tester. Compressed density (0.967 mg/mm 3 ) was calculated from the actual weight (194 mg) and thickness (3.99 mm) of the compressed Mg-St. Total tablet surface areas of upper and bottom convex curve were mm 2. The partial volume of Mg-St could be calculated from the results of Mg-St contents and the compressed density. Thickness of Mg-St layer was calculated by dividing the volume by the surface area of tablet Statistical analysis of parameters of the external lubrication system and Mg-St content or disintegration time Analysis of variance (ANOVA) of the parameters in the external lubrication system and the Mg-St content was performed using version 9.8 of the software Unscrambler (CAMO Software Japan, Tokyo, Japan). A formula for predicting the Mg-St content was subsequently established from the result of a multiple regression analysis, which was calculated according to the manufacturing conditions and Mg-St quantification results of eight samples (i.e., A1 to A8). ANOVA of the parameters in the external lubrication system and the disintegration time of tablets was performed with same software according to the conditions and results of A1 A

11 Results and Discussion Physicochemical properties of tablets manufactured using external and internal lubrication methods The physicochemical properties of the tablets manufactured using the internal and external lubrication methods are given in Table 2. While the hardness and disintegration times were comparable among the tablets manufactured using the external lubrication method, the physicochemical properties differed between the tablets manufactured using the internal and external lubrication methods. All the tablets manufactured using the external lubrication method were approximately two-fold harder than those manufactured using the internal lubrication method, and the disintegration time of the tablets manufactured using the external lubrication method was five-fold shorter than that of the tablets manufactured using the internal lubrication method. The high hardness and rapid disintegration properties of the tablets manufactured using the external lubrication method can be explained in terms of the differences in the distribution of the lubricant relative to the surface of the granules. Because the lubricant was quickly sprayed only onto the punch and die in the rotary tablet press when using the external lubrication system, the possibility of a distribution of Mg-St within the tablets seems negligible. The slight presence of Mg-St within tablets prepared using the external lubrication method would contribute to higher hardness because the adhesion of granules in tablets is not prevented. In addition, the slight presence of hydrophobic Mg-St within the tablets would make it easier for water to penetrate the tablets and granules, resulting in rapid disintegration in less than 2 minutes, even though the tablets in this study do not contain any disintegrant. The hardness of tablets without lubricant was lower than that of tablets produced using the external lubrication method. This difference might be ascribed 11

12 to the difference in compression speed (rotary tablet press vs. precision universal tester) and, in fact, the tablets without lubricant are slightly thicker. It remains unclear whether lubricant amounts on the surface of tablets were quantitatively different between internal and external lubrication methods. Therefore, the Mg-St content and surface wettability of the tablets prepared using the external lubrication method under a variety of manufacturing conditions were determined as shown in Table 3. The quantities for the external lubrication method ranged from one-tenth to one-fiftieth of typical quantities (0.18 mg/tablet, 1 wt%) for the internal lubrication method, and the standard deviations were quite low, less than 2%. Next, Mg-St concentrations in the surface layer of tablets were compared for external and internal lubrication methods. Mg-St is most likely localized on the tablet surface in the external lubrication method. Since the measurement of thickness of Mg-St on the tablet surface is technically difficult, the thickness was calculated from the quantity of Mg-St, its compressed density and the surface area of the tablets, and could be estimated as around 1 μm. This value of thickness is close to the value for the external lubrication method reported by Yamamura et al., under 1 µm at the center of the tablet surface. 11 By the way, the mean particle sizes of lactose monohydrate and microcrystalline cellulose which were used for the tablet excipients are approximately 150 µm. In case Mg-St with 1 μm layer is adhered onto the surface of these excipients on the outmost layer of tablets, the total concentration of Mg-St can be calculated to wt% from the following equation C = A V W 2 V where 12

13 C : Concentration of Mg-St in the 150 m layer (wt%), A : Mg-St content of tablet (mg/tab), W : Weight of tablet (180 mg/tab), V : Volume of whole tablet (mm 3 ), V : Volume of tablet of 150 m layer (mm 3 ), From this result, although the quantities of Mg-St for the external lubrication method ranged from one-tenth to one-fiftieth of typical quantities for the internal lubrication method, the concentration of Mg-St on the surface layer of tablet in each lubrication methods was almost same so that equivalent lubrication property could be attained. Although the quantity of Mg-St in the whole tablet is indispensable information with which to elucidate a unit formula from a regulatory point of view, the evaluation of tablet surface properties, such as wettability, in relation to the Mg-St content is valuable in understanding the function of tablets. The contact angle of the tablets produced without lubricant, 31.8, was approximately half that of sample A1 containing the lowest amount of lubricant of all the samples tested in the current study. This result shows that even the addition of a very small amount of Mg-St by the external lubrication system can notably reduce the wettability of the tablet surface. The contact angles of tablets prepared using the internal lubrication method were higher than those of the tablets prepared using the external lubrication method, probably because the content of Mg-St, 1.0 wt%, was higher than the contents on the surfaces of the tablets manufactured using the external lubrication method. The flattening of Mg-St during the blending process would also cause the higher contact angles in the internal lubrication method. 3 Mg-St has a tendency to cover surfaces by adherence followed by delamination, which leads to the formation of a discontinuous film around the particles. 16 The contact 13

14 angles increased with increasing Mg-St content, and the contact angles of the eight different tablets manufactured using the external lubrication system had a comparatively low standard deviation of less than 6. A pronounced positive correlation was observed between the contact angle and the Mg-St content of the tablets with a correlation coefficient value (R) of 0.953, as shown in Fig. 1. This result supports the idea that Mg-St was mainly added to the surface, not the interior, of the tablets produced using the external lubrication system and that the wettability can be controlled by varying the manufacturing conditions of the external lubrication system. These changes in wettability, however, did not affect the retarded disintegration property of the tablets Effect of manufacturing parameters of the external lubrication systems and rotary press on the Mg-St content of the tablets The results of the ANOVA using the manufacturing conditions and Mg-St content are presented in Table 4. All the main factors, namely the spray rate, flow air volume of dust collection and rotating speed of the rotary press, were determined to be significantly influential parameters (P < 0.05), although their interactions did not affect the Mg-St content. This confirmed that the Mg-St content can be simply expressed using these main factors only. The result of the ANOVA after pooling the insufficient interactions is also presented in Table 4. The smallest P-value of the spray rate among the three main factors indicates that the spray rate was the most influential manufacturing parameter in terms of its impact on the Mg-St content. More detailed parameters relating to the spray rate in the external lubrication system are given in Table 5. In the external lubrication system, the lubricant is dispersed into compressed air after being held in a high fluidity state to prevent aggregation. The spray rate of the lubricant can be controlled by the inlet dispersion air volume and the inlet purge air 14

15 volume. Although two different spray rates (1.0 and 2.0 g/min) were used in the current study, the inlet dispersion air volume was almost the same for each spray rate, and the degree of dispersion and Mg-St concentration in the spray were also very similar for each spray rate. These results indicate that the spray rate was mainly affected by the fluidity of the lubricant being dispersed into the inlet dispersion air system and that the degree of dispersion of the lubricant could be effectively controlled by the inlet dispersion air volume for the mixing of the lubricant. We previously reported that the spray rate, degree of dispersion of the lubricant and Mg-St content of the tablets remained stable during an 8-h period in which the rotary tablet press and external lubrication system were being operated. 15 This particular system is therefore unique and possesses a distinct advantage over existing external lubrication systems in that it allows for the spray rate and the degree of dispersion of a lubricant that can easily aggregate to be readily monitored in real time. A formula for predicting the Mg-St content of the tablets was subsequently established: y = (x x ) (x x ) (x x ) , (1) where x : spray rate (g/min), x : air flow volume of the dust collector (m 3 /min), x : rotation speed of the rotary tablet press (min 1 ), y: predicted Mg-St quantity (mg/tablet), x, x, x : average values of settings for each parameter

16 A high correlation coefficient, R = 0.994, was observed for the predicted values and the actual analytical values of A1 A8, as shown in Fig. 2. The slope of the spray rate is positive. This means a larger quantity of Mg-St in tablets can be applied with a higher spray rate. In the case that the spray rate increased by 0.1 g/min, the actual amount of consumed Mg-St increased by mg/tablet under a rotation speed of the rotary press of 30 min 1 (100 mg / (20 stations 30 min 1 )). However, the model equation indicates an increase of mg/tablet in this case, which is around 3% of the sprayed amount. This is consistent with the results of our previous study that showed that 4.3% of consumed Mg-St was added to the tablets as calculated from the total mass balance, 15 although this percentage would depend on the size of the punch, size of the turntable and number of stations or punches in the rotary press for an identical air flow volume of the dust collector and rotation speed. The slopes of the air flow volume of the dust collector and the rotation speed are negative. This means that a smaller quantity of Mg-St would be included in tablets with higher settings of these parameters. For example, in the case that the rotation increases from 30 to 31 min 1 at a spray rate of 1.5 g/min (75 mg/min/punch), the consumed amount of Mg-St per tablet decreases from 2.50 to 2.42 mg/tablet. As described above, it is estimated that 3% of the sprayed Mg-St is added to the tablets, and the corrected estimation of the decrease is mg/tablet, that is close to the coefficient of rotation speed in the model equation. The negative coefficient of the rotating speed in the model equation appears reasonable because a higher rotation speed reduces the amount of Mg-St sprayed onto the punches and dies and therefore reduces the amount of Mg-St included in the tablets. Although it is difficult to clarify quantitatively the meaning of the coefficient for the air flow volume of the dust collector, the balance of the air velocity between the air sprayed to apply Mg-St and the dust collector appears important and a negative relationship between the Mg-St content and air flow volume 16

17 is reasonable. The establishment of this equation makes it possible to configure the manufacturing conditions of the external lubrication system such that tablets can be prepared with a desired Mg-St content Effect of the manufacturing parameters of the external lubrication systems and rotary press on the disintegration time The ANOVA for the manufacturing conditions and disintegration time showed that none of the main factors or their interactions were determined to be significantly influential parameters (P> 0.05). Also, although ANOVA was re-performed after three interactions were pooled, P-value of the main factor (spray rate) was still high, indicating that no correlation between manufacturing conditions and disintegration time was observed Validity of the model equation To confirm the validity of model equations, an additional sample was prepared under conditions of a spray rate of Mg-St of 1.0 g/min, a flow air volume of the dust collector of 0.2 m 3 /min and a rotation speed of the rotary tablet press of 30 min 1, and their Mg-St content and disintegration times were evaluated as shown in Table 6. The quantified value of Mg-St content was 89% of the value predicted from model equation (1), confirming the high reliability of the equation Conclusions The properties of tablets manufactured using the external lubrication method (such as hardness and the disintegration time) deteriorated less than those of tablets prepared using the internal lubrication method, with the differences being due to the distribution of the lubricant 17

18 to the surfaces of the granules. Analysis of the contact angle showed clear relationships between wettability and Mg-St content of the tablet surface for both the external lubrication method and internal lubrication method. Most influential parameters for Mg-St content, namely the spray rate, rotation speed of the tablet press and the flow air volume of the dust collector, were statistically verified. None of the interaction effects of these variables were critical to the outcome of the tableting process. The process parameters for the disintegration of tablets were not so influential as demonstrated by statistical analysis. The results of this study also revealed good correlation between the predicted Mg-St content, which was calculated from the most influential parameters of an external lubrication system using a simple model equation, and actual experimental values. The disintegration time predicted with the model equation correlated with actual experimental values. These results indicate that the process of the external lubrication method can be understood by the statistical analysis of manufacturing parameters. Taken together, the results of this study show that the external lubrication method is a more practical approach for the development and manufacture of pharmaceutical drug products than past approaches, allowing efficient production, the prevention of manufacturing problems and the creation of new high-performance drug products Declaration of interest The authors have no conflicts of interest. The authors alone are responsible for the content and writing of this article. 18

19 References [1] T.A. Miller, P. York, Pharmaceutical tablet lubrication. Int. J. Pharm. 41 (1988) [2] T. Jahn, K.J. Steffens, Press chamber coating as external lubrication for high speed rotary presses: lubricant spray rate optimization. Drug Dev. Ind. Pharm. 31 (2005) [3] G.K. Bolhuis, C.F. Lerk, H.T. Zijlstra, A.H. DeBoer, Film formation by magnesium stearate during mixing and its effect on tableting. Pharm. Weekblad. 110 (1975) [4] W.A. Strickland, E. Nelson, L.W. Busse, T. Higuchi, The physics of tablet compression IX. Fundamental aspects of tablet lubrication. J. Am. Pharm. Assoc. Sci. Ed. 45 (1956) [5] G.K. Bolhuis, A.J. Smallenbroek, C.F. Lerk, Interaction of tablet disintegrants and magnesium stearate during mixing. I. Effect on tablet disintegration. J. Pharm. Sci. 70 (1981) [6] C.F. Lerk, G.K. Bolhuis, A.J. Smallenbroek, K. Zuurman, Interaction of tablet disintegrants and magnesium stearate during mixing II. Effect on dissolution rate. Pharm. Acta. Helv. 57 (1982) [7] D.S. Desai, B.A. Rubitski, S.A. Varia, A.W. Newman, Physical interactions of magnesium stearate with starch-derived disintegrants and their effects on capsule and tablet dissolution. Int. J. Pharm. 91 (1993) [8] M.J., Mollan Jr. M. Çelik, The effects of lubrication on the compaction and post-compaction properties of directly compressible maltodextrins. Int. J. Pharm. 114 (1996) 1 9. [9] J. Barra, R. Somma, Influence of the physicochemical variability of magnesium stearate on its lubricant properties: Possible solutions. Drug. Dev. Ind. Pharm. 22 (1996)

20 [10]P. Gruber, V.I. Glasel, W. Klingelholler, T. Liske, Direct lubrication of tablet tools, a contribution to the optimization of tablet manufacture. Drugs. Made. Ger. 34 (1991) [11] T. Yamamura, T. Ohta, T. Taira, Y. Ogawa, Y. Sakai, K. Moribe, K. Yamamoto, Effects of automated external lubrication on tablet properties and the stability of eprazinone hydrochloride. Int. J. Pharm. 370 (2009) 1 7. [12]Y. Oneda, M. Kubota, N. Kitamura, K. Fujita, H. Suzuki, Development of external lubrication system for tableting and its application a wide field of industry. J. Jpn. Soc. Pharm. Mach. & Eng. 18 (2009) [13]H. Takeuchi, S. Nagira, M. Aikawa, H. Yamamoto, Y. Kawashima, Effect of lubrication on the compaction properties of pharmaceutical excipients as measured by die wall pressure. J. Drug. Deliv. Sci. Technol. 15 (2005) [14]International conference on harmonization; ICH Harmonized Tripartite Guideline Pharmaceutical Development Q8(2006) [15]T. Kamiya, H. Kondo, H. Hiroma, S. Nakajima, M. Watanabe, K. Yamashita, K. Sako, M. Uemura, H. Hashizume, Development of the Novel Functional In-line Lubrication System and Applications for Pharmaceutical Industry. J. Soc. Powder. Technol. Japan 49 (2012) [16] A.C. Shah, A.R. Mlodozeniec, Mechanism of surface lubrication: influence of duration of lubricant-excipient mixing on processing characteristics of powders and properties of compressed tablets. J. Pharm. Sci. 66 (1977)

21 Figure Legends Figure 1. Relationship between the Mg-St content and contact angle Figure 2. Correlation between the predicted and observed Mg-St contents. Adjusted coefficients of determination was and root mean square error was

22 Table Table 1. Experimental settings of the external lubrication system Sample no. A1 A2 A3 A4 A5 A6 A7 A8 Spray rate of Mg-St (g/min) * Flow air volume of dust collector (m 3 /min) Rotation speed of rotary tablet press (min -1 ) * Spray rates were g/min or g/min, which were determined by the calculation of 3 minutes moving average including the sampling point. 5 1

23 Table 2. Comparison of tablet properties Sample no. Weight (mg) Thickness (mm) Hardness (N) Disintegration time (s) A ± ± ±4.2 80±11 A ± ± ±3.9 76±6 A ± ± ±4.4 79±15 A ± ± ±4.6 91±17 A ± ± ±6.1 76±9 A ± ± ±5.8 86±11 A ± ± ±6.4 80±17 A ± ± ± ±7 Internal lubrication 179.6± ± ± ±24 method Tablets without any lubricant 181.4± ± ±5.4 47±7 Weight (n = 100), thickness and hardness (n = 10), and disintegration time (n = 6) are presented for A1 A8 and tablets of the internal lubrication method. Weight and thickness (n = 10), hardness (n = 5), and disintegration time (n = 6) are presented for tablets without lubricant. Results are presented as a mean ± standard deviation. 13 2

24 Table 3. Mg-St contents of tablets and comparison of contact angles Sample no. Mg-St (mg/tab) * Contact angle ( ) Tablets without any lubricant 31.8±4.4 A ± ±5.6 A ± ±3.9 A ± ±3.9 A ± ±4.9 A ± ±4.1 A ± ±3.8 A ± ±2.9 A ± ±2.9 Ref (Internal lubrication method) 95.2±3.2 * Mean ± standard deviation (n = 3) 10 tablets tested for each sample except Tablets without any lubricant (n = 5). Results are the mean ± standard deviation. Not tested. 21 3

25 Table 4. P-values calculated from ANOVA (Mg-St content) Item P-value P-value after pooling the interactions Spray rate of Mg-St Flow air volume of dust collector Rotation speed of rotary tablet press Spray rate air flow volume Spray rate rotation speed Flow air volume rotation speed *Settings and results of the eight samples (A1 A8) selected for ANOVA 4

26 Table 5. Parameters of the external lubrication system related to the spray rate of Mg-St Sample no. A1 A2 A3 A4 A5 A6 A7 A8 Spray rate of Mg-St (g/min) Inlet dispersion air volume (L/min) Inlet purge air volume (L/min) Degree of dispersion by laser light in pipe (%) Concentration of Mg-St in sprayed air (mg/l)

27 Table 6. Prediction of Mg-St content and disintegration time using the statistical model Predicted Experimental Deviation (%) Mg-St content (mg/tab) ± *Process parameters: spray rate of Mg-St (X 1 ) = 1.0 g/min, flow air volume of the dust collector (X2) = 0.2 m 3 /min, rotation speed of the rotary tablet press (X3) = 30 rpm 6

28 Fig. 1 Kamiya et al. 90 Contact angle ( ) R= External lubrication No lubricant Mg-St content (mg/tab)

29 Fig. 2 Kamiya et al R= R = Actual Mg-St content (mg/tab) Predicted Mg-St content (mg/tab) Adjusted coefficients of determination (R 2 ) is and RMSE is

30 Fig. 3 Kamiya et al. 120 Actual disintegration time (sec) R= Predicted disintegration time (sec) Adjusted coefficients of determination (R 2 ) is and RMSE is 5.7.

31 Graphic Abstract Kamiya T et al. Surface wettability of tablet by examined by contact angle method 90 Schematic image of Mg-St distribution Contact angle (degrees) R= external lubrication No lubricant External lubrication 40 no lubricant Mg-St content (mg/tab) Internal lubrication Mg-St Good correlation between predicted and actual Mg-St content 0.16 R = Actual Mg-St content (mg/tab) Predicted Mg-St content (mg/tab)