A comparative study of roll compaction of free-flowing and cohesive pharmaceutical powders

https://doi.org/10.1016/j.ijpharm.2012.02.033Get rights and content

Abstract

Roll compaction is widely adopted as a dry granulation method in the pharmaceutical industry. The roll compaction behaviour of feed powders is primarily governed by two parameters: the maximum pressure and the nip angle. Although the maximum pressure can be measured directly using pressure sensors fitted in the rolls, it is not a trivial task to determine the nip angle, which is a measure of the size of the compaction zone and hence the degree of compression. Thus a robust approach based upon the calculation of the pressure gradient, which can be obtained directly from experiments using an instrumented roll compactor, was developed. It has been shown that the resulting nip angles are comparable to those obtained using the methods reported in literature. Nevertheless, the proposed approach has distinctive advantages including (1) it is based on the intrinsic features of slip and no-slip interactions between the powder and roll surface and (2) it is not necessary to carry out wall friction measurements that involve plates that may not be representative of the roll compactor in terms of the surface topography and surface energy. The method was evaluated by investigating the effect of roll speed for two pharmaceutical excipients with distinctive material properties: microcrystalline cellulose (MCC) and di-calcium phosphate dihydrate (DCPD). It was found that the maximum pressure and nip angle for DCPD, which is a cohesive powder, decrease sharply with increasing roll speed whereas they are essentially independent of roll speed for MCC, which is an easy flowing powder. The roll compaction behaviour of MCC–DCPD mixtures with various compositions was also investigated in order to evaluate the effect of flowability. It was found that the nip angle and maximum pressure generally increased with improved flowability of the feed powders.

Introduction

Roll compaction is a continuous agglomeration process first developed in the late 19th century to produce coal briquettes (Simon and Guigon, 2003). It has been adopted as a dry granulation process in the pharmaceutical industry for more than 50 years (Kleinebudde, 2004). The technique produces ribbons or flakes that are then milled to form granules (Bennett and Cole, 2003). The purpose of this process is mainly to increase bulk density, flowability and uniformity of formulation blends for producing high quality tablets with low dosage and weight variations.

During roll compaction, powders are fed, gripped in the decreasing gap between the counter-rotating rolls and compressed. Johanson (1965) proposed that there are three distinctive regions, i.e. slip, nip and release regions (Fig. 1). In the slip region, particles slip at the roll surfaces and are rearranged, so that there is a small degree of densification and small compression pressures are developed. In the nip region, the velocity of the powder adjacent to the rolls is equal to that of the roll surfaces, and the powder undergoes traction induced compaction due to the decreasing roll gap and the friction along the surfaces of the rolls. The compacted powder finally enters the release region after passing through the minimum roll gap. A nip angle that defines the angular location of the onset of the nip region was introduced to specify the transition from slip to no-slip wall boundary conditions. The maximum compression pressure and nip angle are two important parameters since they govern the extent of powder densification.

Roll compaction is a complex process that depends on a number of factors, such as the system layout, processing conditions and feed powder properties (Guigon et al., 2007, Miller, 1997). The system layout includes the feeding method (Guigon and Simon, 2003, Inghelbrecht and Remon, 1998b), sealing systems (Funakoshi et al., 1977, Miguélez-Morán et al., 2009) and powder de-aeration method (Spinov and Vinogradov, 1967). Inghelbrecht and Remon (1998b) investigated the roll compaction of lactose with two smooth-surface rolls (L83 Chilsonator, Fitzpartrick, USA) instrumented with a screw-feeding system consisting of a vertical screw and horizontal screw. The powders were pre-densified and transported to the nip of the rolls by a vertical screw having a horizontal screw feed. The speeds for both screws were adjustable. The compacted ribbons were milled using a granulator (MG624, Frewitt, Switzerland) in order to evaluate their quality by sieve analysis and friability testing (Inghelbrecht et al., 1997) of the corresponding granules. It was found that the speeds of the horizontal and vertical screw feeders have a significant influence on the product quality. They argued that the horizontal screw speed controlled the amount of feeding powder. The effects of the vertical screw speed were mainly on the pre-densification and de-aeration of the feed powder. Guigon and Simon (2003) carried out experiments with lactose monohydrate, alumina and sodium chloride using a laboratory roll compactor (B100QC, K.R. Komarek, USA) having a vertical roll configuration and a horizontal screw feeding system. They concluded that the speed of the screw feeder is the only dominating factor for the throughput in the case of screw feeding. They also found that the roll compression stress applied to the ribbons exhibited periodic fluctuation with frequency that were similar to that of the rotation of the single screw feeding system. Spinov and Vinogradov (1967) examined the roll compaction of a copper powder at a constant roll speed and investigated the effect of air entrainment. They pointed out that aeration resulted in discontinuities of the density across the entire width of the ribbon, formation of large voids on the ribbon surface and a decrease in the width of the ribbons. They introduced a vacuum deaeration facility and showed that it was able to minimise the pores and enhance the uniformity and quality of the products. Powder leakage during roll compaction is a problem leading to insufficient compression. In order to solve this problem, sealing systems such as cheek plates (Miguélez-Morán et al., 2009) or a rim on one of the rolls (Funakoshi et al., 1977) are used. Miguélez-Morán et al. (2009) characterised the density of MCC ribbons produced using a laboratory scale roll compactor with cheek plates sealing. They found that non-uniform powder feeding was caused by the friction between the feeding powders and the cheek plates, which resulted in a non-uniform density distribution across the ribbon width. Funakoshi et al. (1977) designed a concavo-convex pair of rolls to prevent leakage of powders and thus improve the uniformity of the pressure across the width of the roll surfaces. The uniformity index, Pmax/Pmin, which is the ratio of the maximum and minimum pressures across the roller surface width, was combined with the amount of leaked powder to evaluate the efficiency of the sealing system. They examined the uniformity index and the amount of leaked powder at various rim wall slope angles, and found that 65° was the optimal concavity.

The performance of the feed powders in roll compaction can also be affected by the process parameters, such as roll pressure (Inghelbrecht and Remon, 1998b, Parrot, 1981), roll speed (Petit-Renaud et al., 1998, Yusof et al., 2004) and roll gap (Bindhumadhavan et al., 2005). Inghelbrecht and Remon (1998b) stated that the hydraulic pressure is the most important parameter that affects the size distribution and friability of the resulting granules. Large hydraulic pressures resulted in sufficient compression and hence strong granules that had a high fraction in the required size range. Petit-Renaud et al. (1998) found that an increase in roll speed of an instrumented compactor (model B-100QC, K.R. Komarek, USA) resulted in a linear increase of the mass throughput but a decrease of the normal pressure and nip angle at a constant roll gap. Yusof et al. (2004) showed that the throughput increased linearly with increasing in roll speed for maize as a feed powder. Bindhumadhavan et al. (2005) roll compacted MCC powders (Avicel PH 102) using a laboratory scale instrumented roll compactor with gaps in a range of 0.9–2 mm. They reported that the compression pressure decreased with increasing roll gap, and the values of the pressure were consistent with the predictions obtained using Johanson theory (Johanson, 1965). Parrot (1981) examined various pharmaceutical powders such as acetaminophen, precipitated calcium carbonate, DCPD and lactose at five different compression pressures using a concave–convex roll compactor. Their results showed that the bulk densities of the produced ribbons were linearly proportional to the logarithm of the pressure applied for all the samples.

In order to increase the efficiency and robustness of roll compaction, it is necessary to improve the current understanding of the influence of the system design and process conditions on particular feed powders and also develop a knowledge base of the influence of the properties of the feed formulations on those of the resulting granules. This has attracted increasing interest (Bacher et al., 2007, Chang et al., 2008, Endale et al., 2008, Inghelbrecht and Remon, 1998a, Kleinebudde, 2005, von Eggelkraut-Gottanka et al., 2002). In particular, the properties of the ribbons and granules made from various raw materials, and the tablets made from the granules have been investigated intensively. It has been shown that the tensile strength of the ribbons and tablets, the bulk density of the ribbons and granules, and the flow properties of the granules strongly depend on the formulation, particle size, moisture content, morphology and friction coefficient of the feed powders (Bacher et al., 2007, Chang et al., 2008, Grulke et al., 2004, Gupta et al., 2005, Herting et al., 2007). Herting et al. (2007) measured the tensile strength and dissolution properties of the tablets made from granules, which were obtained from roll compaction, in order to evaluate the performance of four different dry binders based on binary mixtures and dicalcium phosphate (DCP). It was shown that the dissolution behaviour of the tablets containing copovidones of different particle sizes was comparable, but the tensile strength of the tablets with a binder of smaller particle size was greater. Chang et al. (2008) examined the roll compaction of mixtures of a ductile API and excipients (i.e. mannitol and lactose) and the particle size of the corresponding granules. It was found that the granules containing lactose were generally smaller than those with mannitol due to the brittleness of lactose. Grulke et al. (2004) investigated the roll compaction of MCC (Avicel PH 101), lactose monohydrate (LM Granulac 200) and DCPD (Emcompress) and 1:1 mixtures of these powders at two specific ‘compaction rates’ of 2 and 7 kN/cm. They showed that the bulk and tapped densities of the granules had a similar trend to the values for the raw materials, implying that the densities of the feed powders play an important role in roll compaction. Gupta et al. (2005) explored the influence of ambient moisture on the roll compaction behaviour of MCC powder (Avicel PH 200), and showed that the presence of moisture resulted in a decrease in the strength of the ribbons. They also introduced a critical density of powders at which the densification from particle arrangement reaches a maximum value before the onset of particle deformation and found that the critical density was independent of the moisture content of the feed powder. Bacher et al. (2007) investigated the roll compaction behaviour of calcium carbonate powders with various morphological forms and sorbitols of different particle sizes. The ribbons were ground into granules. The improvement of the flowability, compactibility (Jørn, 2006) and compressibility characterised by the Walker coefficient (Walker, 1923) were calculated in order to evaluate the quality of the products. It was shown that particle morphology and particle size were the most influential factors in determining the properties of the compacts. A relationship between wall friction angle and maximum compression pressure was obtained by Bindhumadhavan et al. (2005), who investigated the roll compaction of lubricated MCC (Avicel PH 102) with a roll gap of 1.2 mm at a roll speed of 2 rpm. They found that the maximum pressure increased with increasing wall friction between the powders and stainless steel roll surfaces.

The flowability of the feed powder is an important factor in the performance of roll compaction since it affects the compression behaviour (Mansa, 2006) and the cohesion of the granules (Chang et al., 2008). von Eggelkraut-Gottanka et al. (2002) argued that the poor flowability of feed powders with a small particle size might result in the fluctuation of the powder filling even when a screw feeder was used. This phenomenon might cause aeration during roll compaction, which could facilitate the formation of large voids in compacted ribbons and therefore limited densification as shown by Spinov and Vinogradov (1967). How powder flowability affects on the feeding and subsequent compression behaviour was explored by Miguélez-Morán et al. (2008), who examined the effect of lubrication using magnesium stearate (MgSt) on the roll compaction of MCC (Avicel PH 102). The results showed that the compaction pressure and homogeneity of the ribbon density was dominated by the way that the powders were fed into the compaction zone, which was determined by the flowability of the feed powders. Mansa (2006) examined the roll compaction of MCC (Comprecel M101), DCPA (Anhydrous Emcompress) and mixtures of these two powders with various ratios and found that an increase in the cohesion of the feed powder resulted in an increase of the nip angle. Chang et al. (2008) found that the granules with lactose as a feed powder were less cohesive than those with mannitol, which was consistent with the better flow behaviour of lactose.

Although roll compaction has been investigated intensively in the last few decades, it is still not well understood primarily due to the diversity in the controlling factors and material properties. For example, although previous studies highlighted that the materials properties of the feed powders, particularly the flow properties, play an important role in roll compaction, it is still unclear how the dominating compression parameters (i.e. compaction pressure and nip angle) depends upon the properties of the feed powder. This was the objective of the current work, especially the influence of the flow properties and roll speed. For this purpose two pharmaceutical excipients MCC and DCPD with distinctive mechanical and flow properties were used. A detailed analysis of the mechanical response of these two powders was performed, for which a robust method for determining the nip angle was established and the advantage of this method will also be discussed. In addition, the affect of roll speed on the maximum pressure and nip angle for these powders will be explored.

Section snippets

Materials

Two commonly used pharmaceutical excipients were selected: microcrystalline cellulose (MCC) of Avicel grade PH 102 (FMC Biopolymer, USA) and dibasic calcium phosphate dihydrate (DCPD) of Calipharm D grade (Rhodia, France). MCC is a crystalline powder (crystallinity > 78%) with needle-shaped particles (see Fig. 2a). DCPD (Calipharm D) is also a crystalline powder but with shale-like particles (Fig. 2b).

Binary mixtures of MCC and DCPD of various compositions, i.e. with 25%, 50%, 75% and 90% MCC

Characterisation of powders

The material properties of the feed powders are summarised in Table 1. The mean particle size (D50 = 8.12 μm) of DCPD, which has a higher true density, is more than an order magnitude less than that of MCC (D50 = 90.65 μm). The values of the flow function, ffc, which represents the flowability, indicate that DCPD is a cohesive powder (i.e. 2 < ffc < 4) while MCC is an easy-flowing powder (i.e. 4 < ffc < 10). Although the effective angles of internal friction ϕe of DCPD and MCC are similar, the angle of wall

Transition angle

For both MCC and DCPD powders, the maximum pressures were not obtained at the minimum roll gap (i.e. θ = 0), but corresponded to a finite value of the angular position (Fig. 4), which was also found in numerical simulations (Cunningham, 2005) and the analysis of experimental data (Chekmarev and Vinogradov, 1963, Guigon et al., 2007, Lecompte et al., 2005). Radchenko (1974) suggested that a neutral angle could be introduced to represent the angular position for the maximum pressure with reference

Conclusions

The roll compaction behaviour of two common pharmaceutical powders with distinctive material properties (MCC and DCPD) and their mixtures was investigated. It has been shown that the maximum pressure decreased with increasing roll speed for DCPD, while that for MCC is almost constant. A robust method for determining the nip angle based on the intrinsic features of slip and no-slip interactions between the powder and roll surface was developed. It has an advantage in determining the nip angle

Acknowledgments

The authors would like to thank AstraZeneca for funding the project and providing sample materials. The characterisation equipments used were obtained, through Birmingham Science City: Innovative Uses for Advanced Materials in the Modern World, with support from AWM and part funded ERDF.

References (42)

  • A. Michrafy et al.

    Analysis of strain stress state in roller compaction process

    Powder Technol.

    (2011)
  • A.M. Miguélez-Morán et al.

    Characterisation of density distributions in roller-compacted ribbons using micro-indentation and X-ray micro-computed tomography

    Eur. J. Pharm. Biopharm.

    (2009)
  • A.M. Miguélez-Morán et al.

    The effect of lubrication on density distributions of roller compacted ribbons

    Int. J. Pharm.

    (2008)
  • B.A. Patel et al.

    Predicting the pressure distribution during roll compaction from uniaxial compaction measurements

    Chem. Eng. J.

    (2010)
  • K. Schönert et al.

    Shear stresses and material slip in high pressure roller mills

    Powder Technol.

    (2002)
  • O. Simon et al.

    Correlation between powder-packing properties and roll press compact heterogeneity

    Powder Technol.

    (2003)
  • C.Y. Wu et al.

    Roller compaction of moist pharmaceutical powders

    Int. J. Pharm.

    (2010)
  • B. Bennett et al.

    Pharmaceutical Production: An Engineering Guide

    (2003)
  • Bindhumadhavan, G., 2004. Title. PhD Thesis, University of Birmingham,...
  • C.K. Chang et al.

    Roller compaction, granulation and capsule product dissolution of drug formulations containing a lactose or mannitol filler, starch, and talc

    Aaps Pharmscitech

    (2008)
  • A.P. Chekmarev et al.

    Investigation on specific pressure, specific friction, and the coefficient of friction during metal powder rolling

    Soviet Powder Metal Ceram.

    (1963)
  • Cited by (33)

    • The effect of excipient particle size on the reduction of compactibility after roller compaction

      2022, International Journal of Pharmaceutics: X
      Citation Excerpt :

      Von Eggelkraut-Gottanka et al., 2002) Additionally, powder feeding of very fine material might cause aeration during roller compaction, which could facilitate the formation of large voids in compacted ribbons and therefore limit the densification.( Yu et al., 2012; Miguélez-Morán et al., 2008) Using starting material with smaller particle size can however also be beneficial. Herting and Kleinebudde(Herting and Kleinebudde, 2007) for example, reported that a finer particle size of microcrystalline cellulose (MCC) and theophylline can be beneficial for the tensile strength of tablets produced after dry granulation.

    • Continuous dry granulation

      2022, How to Design and Implement Powder-to-Tablet Continuous Manufacturing Systems
    • Relationship between powder properties and uniformity of ribbon property using feeding guider designs with thermography (PAT) in roller compaction

      2022, Powder Technology
      Citation Excerpt :

      The more powder compressed by the rollers; the more heat generated (higher temperature) of the ribbon. A similar trend was observed by Wiedey and Kleinebudde [15], Krok and Wu [17], Michrafy, Diarra, Dodds and Michrafy [19], Yu, Gururajan, Reynolds, Roberts, Adams and Wu [27], Olaleye, Wu and Liu [28], Perez-Gandarillas, Perez-Gago, Mazor, Kleinebudde, Lecoq and Michrafy [29], Souihi, Reynolds, Tajarobi, Wikström, Haeffler, Josefson and Trygg [30] and Miguelez-Moran, Wu, Dong and Seville [21]. Also, there is a robust correlation between roller force and temperature distribution in the ribbon.

    View all citing articles on Scopus
    View full text