effect of magnesium stearate on bonding and porosity expansion of tablets produced from materials...

International Journal of Pharmaceutics 179 (1999) 107–115

Effect of magnesium stearate on bonding and porosityexpansion of tablets produced from materials with different

consolidation properties

K. Zuurman a, K. Van der Voort Maarschalk b, G.K. Bolhuis a,*a Groningen Institute for Drug Studies (GIDS), Department of Pharmaceutical Technology and Biopharmacy,

Uni6ersity of Groningen, Ant. Deusinglaan 1, 9713 AV, Groningen, The Netherlandsb AKZO Nobel Central Research, De6enter, The Netherlands

Received 2 September 1998; received in revised form 23 November 1998; accepted 27 November 1998

Abstract

The negative effect of magnesium stearate on tablet strength is widely known. This strength reduction is alwaysconsidered to be the result of reduction of interparticle bonding. It is also known that interparticle bonding affectsrelaxation of tablets. Relaxation increases with decreasing bonding. Microcrystalline cellulose is an example of amaterial with a high lubricant sensitivity, which effect is caused by its plastic deformation behavior duringcompression. This paper shows for microcrystalline cellulose that the porosity under pressure was equal forunlubricated tablets and for tablets containing 0.5% magnesium stearate. This points to equal densification properties.The lubricated tablets show, however, a much larger relaxation than the tablets without magnesium stearate. Thisdifference can be ascribed to the reduction of interparticle bonding by the lubricant, because a strong interparticlebonding counteracts tablet relaxation. In contrast to microcrystalline cellulose, aggregated g-sorbitol (Karion Instant)has a low lubricant sensitivity. Both porosity under pressure and tablet relaxation were found to be equal forlubricated and unlubricated sorbitol tablets. This phenomenon is caused by the particle structure of g-sorbitol. Duringcompression, a lubricant film will be destroyed by fragmentation of the sorbitol aggregates. For this reason,magnesium stearate will hardly affect the interparticle bonding between sorbitol particles and hence have only a smallor no effect on tablet relaxation. © 1999 Elsevier Science B.V. All rights reserved.

Keywords: Stored energy; Particle bonding; Relaxation; Tablet strength; Magnesium stearate; Microcrystallinecellulose; Sorbitol

* Corresponding author. Tel.: +31-50-363-3288; fax: +31-50-363-2500.

0378-5173/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.

PII: S 0378 -5173 (98 )00389 -5

K. Zuurman et al. / International Journal of Pharmaceutics 179 (1999) 107–115108

1. Introduction

Pharmaceutical tablets are normally composedof several ingredients. One of the ingredients isusually a lubricant that prevents the tablet fromsticking to the die and punches and minimizeswear of them. Magnesium stearate is the mostcommonly used pharmaceutical lubricant. In sev-eral papers, however, it has been shown thatmagnesium stearate can have an adverse effect onbonding between particles (Strickland et al., 1956;Bolhuis et al., 1975; De Boer et al., 1978; Dubergand Nystrom, 1982; Vromans and Lerk, 1988).

Additionally, it was recently demonstrated thatinteraction between particles affects the relaxationbehavior of tablets (Rees and Tsardaka, 1994;Van der Voort Maarschalk et al., 1996b). Tabletsproduced from materials with low interparticleattraction tend to suffer from more relaxationthan tablets made from materials where interpar-ticle attractions are large. Realizing that magne-sium stearate affects interparticle bonding andthat the latter has an effect on tablet relaxationimmediately after compression, it can be assumedthat magnesium stearate is a component that al-ters the relaxation properties of tablets.

The present paper discusses the effect of magne-sium stearate on bonding between particles in acompact and the consequenses of alterations inparticle attractions on tablet relaxation.

2. Materials and methods

The materials used were magnesium stearate,(Genfarma, Maarssen, The Netherlands), micro-crystalline cellulose (Pharmacel 102, DMV, Veg-hel, The Netherlands) with a mean particle size of100 mm (15% was smaller than 32 mm and 96%was smaller then 200 mm, determined using an AirJet Sieve), and spray-dried g-sorbitol with a parti-cle size between 106 and 212 mm. The sorbitolfraction was sieved from Karion Instant (Merck,Darmstadt, Germany) using an Air Jet Sieve(Alpine, Augsburg, Germany) equiped with USAStandard testing sieves (W.B. Tyler, Mentor, OH,USA).

The elastic modulus was measured by dynamic

mechanical analysis (DMA) with a RheometricsSolids Analyzer (Piscataway, NY, USA), at afrequency of 0.63 rad/s. The method has previ-ously been described in more detail (Van derVoort Maarschalk et al., 1996a).

Compaction of 500 mg powder into flat-facedtablets with a diameter of 13 mm was carried outon a compaction simulator (ESH, Brierley Hill,UK). The upper punch displacement profiles weresine waves with different amplitudes in order tovary the maximum compression pressures. Theaverage compaction rate was 3 mm/s, correspond-ing with the DMA frequency of 0.63 rad/s. Thelower punch was stationary during compression.The ejection time was always 10 s.

For the production of tablets containing mag-nesium stearate (‘lubricated tablets’), mixing ofthe filler-binder with magnesium stearate was per-formed with a Turbula-mixer, model 2P (W.A.Bachofen, Basle, Switzerland) at 90 rev/min for aperiod of 30 min. Before compression of thecompacts from unlubricated matarial, the die wasprelubricated with magnesium stearate.

Yield stress of the test materials was calculatedfrom the force displacement profiles according toHeckel (1961a,b). Corrections were made for elas-tic punch deformation as previously described(Van der Voort Maarschalk et al., 1996b).

Tablet dimensions were measured at least 16 hafter compaction with an electronic micrometer(Mitutoyo, Tokyo, Japan) and the tablets wereweighed on an analytical balance.

Crushing strengths of the tablets were measuredat least 16 h after compaction using the com-paction simulator as previously described (Vander Voort Maarschalk et al., 1997a). Tensilestrength of the tablets was calculated according toFell and Newton (1968).

3. Results and discussion

Fig. 1 shows the relation between tensilestrength and compaction pressure for tablets com-pressed from unlubricated microcrystalline cellu-lose and microcrystalline cellulose lubricated with0.5% magnesium stearate, respectively. It can beseen that the presence of the lubricant causes a

K. Zuurman et al. / International Journal of Pharmaceutics 179 (1999) 107–115 109

Fig. 1. Tensile strength of unlubricated () and lubricated (�) microcrystalline cellulose compacts.

strong decrease in tablet tensile strength. The highlubricant sensitivity of microcrystalline celluloseand other excipients that undergo plastic defor-mation under compression has been explained inprevious work by film formation of magnesiumstearate upon the substrate particles during themixing procedure (Bolhuis et al., 1975; De Boer etal., 1978). As shown in Fig. 2 (closed symbols)tablets containing magnesium stearate have ahigher porosity than unlubricated ones that werecompressed at the same compaction load. Thisobservation suggests that either the densification(that is the porosity reduction of the microcrys-talline cellulose/magnesium stearate blend underload) is more difficult, compared to the powderwithout magnesium stearate, or the lubricatedtablets show a larger increase in volume due to ahigher stress relaxation after compaction.

Powder densification can be studied by thedetermination of the yield strength of a material.A high yield strength is related with difficultdensification, i.e. larges pressures are necessary toreach a certain porosity. Table 1 shows that thepresence of 0.5% magnesium stearate results in asmall decrease in yield strength of this microcrys-talline cellulose mixture. This effect points to fa-

cilitating of powder densification by magnesiumstearate, just as has been found previously for anumber of directly compressible materials (Vro-mans and Lerk, 1988). The open symbols in Fig.2 show, however, that the minimal attainableporosity under pressure is simular for unlubri-cated and lubricated microcrystalline cellulosepowder, when pressed at a certain load. Based onthis observations one would expect about thesame porosities (after ejection) for unlubricatedand lubricated tablets when compressed at equalloads.

However, the closed symbols in Fig. 2 show ahigher porosity of lubricated tablets after relax-ation. The latter must hence be caused by a higherporosity expansion after compression of the pow-der blend. From previous data (Van der VoortMaarschalk et al., 1998) it became clear that allpore formation was a result of phenomena occur-ing after decompression (i.e. after punch release).These data demonstrate that the compacts pro-duced from a blend of microcrystalline celluloseand magnesium stearate suffer from more stressrelaxation than compacts made from microcrys-talline cellulose only.

A method to estimate the stress relaxationpropensity is calculation of the amount of elastic


Fig. 2. Tablet porosity under pressure (open symbols) and after relaxation (closed symbols) as a function of compaction pressure.Symbols as in Fig. 1.

energy stored during compression (Van der VoortMaarschalk et al., 1996a):

W=12

*s c

2

E(1)

in which: E is the elastic modulus; and sc the yieldstrength of the material.

Table 1 lists yield strengths, elastic moduli andcalculated stored elastic energy for microcrys-talline cellulose, the blend of microcrystalline cel-lulose with 0.5% magnesium stearate andmagnesium stearate. The data suggest that theamount of stored energy increases enormouslyupon addition of magnesium stearate. The reasonfor this large increase is the extremely small elasticmodulus measured for the lubricated material.

The elastic modulus of the non-porous com-pacts is calculated by extrapolation of data ob-tained from compacts with different porosities tozero porosity. The validity of this approach isquestionable, however. The stress relaxation ex-pressed by stored energy is a property of the bulkof the material: it expresses the propensity of theparticles to regain their original shapes. As al-ready indicated, mixing with magnesium stearate

covers the surface of microcrystalline celluloseand this phenomenon may have a large influenceon the bonding of the particles in the strips usedin DMA measurements. Table 1 suggests thatstrips without magnesium stearate are muchstiffer than strips with magnesium stearate. How-ever, there are no reasons to suppose that therigidity of the microcrystalline cellulose particlesin a strip changes upon the addition of a smallamount of lubricant. For this reason, the data forthe elastic modulus of strips containing magne-sium stearate must be considered as an artefact: itis more a measure of the stiffness of the bondingbetween the particles than a measure of the stiff-ness of the particles themselves (Van der VoortMaarschalk et al., 1996b). Consequently, simplemeasurement of the elastic modulus of compactsmade from binary mixtures may lead to erroneousvalues. Exactly the same consideration is applica-ble when explaining the neglectible effect of mag-nesium stearate addition on yield strength. Theyield strength is a measure of the force necessaryto deform a particle permanently. This value isclearly related with the bulk properties of thematerial. The small amount of magnesium


Table 1Yield strength, elastic modulus and stored energy of microcrystalline cellulose (mcc) without and with magnesium stearate (MgSt)

Yield strength (Mpa) Elastic modulus (GPa) Stored energy (kJ/m3)PRIVATE Material

21 12mcc (pure) 18892mcc+0.5% MgSt 19

1661mcc+1.0% MgSt 18–a –aMgSt (pure) 6

a Not measurable.

stearate will not have a significant effect in thistype of yield strength determination.

Another approach is to estimate the storedenergy on the basis of the stored energy of theindividual components. Assuming that the pres-ence of a second material does not affect theviscoelastic properties of the first material in amixture, the energy stored by deformation of theindividual particles can be calculated by:

Wmix=Wcell�8cell+Wmgst�8mgst (2)

with 8cell and 8mgst the volume fractions micro-crystalline cellulose and magnesium stearate, re-spectively.

The stored energies can be calculated as de-scribed in Eq. (1) using the yield strength andelastic moduli of the individual components.

So, Eq. (2) may be written as:

Wmix=12�

s c,cell2

Ecell

�8cell+12�s c,mgst

2

Emgst

�8mgst (3)

The volume fraction of magnesium stearate isalways very low relative to the fraction microcrys-talline cellulose. Moreover, the yield strength ofmagnesium stearate is low compared to that ofmicrocrystalline cellulose (Table 1). Unfortu-nately, it is impossible to determine the elasticmodulus of magnesium stearate, so the relativecontribution of the lubricant cannot be calculated.However, if the elastic modulus of magnesiumstearate is not extremely low, its contribution tothe amount of energy stored in a tablet uponcompression will be neglectible. As the volumefraction of magnesium stearate in a tablet is low,it is very plausible to conclude that the amount ofstored energy in a tablet with a small amount of

magnesium stearate is about the same as that in atablet without magnesium stearate.

The data discussed so far show that relaxationphenomena in terms of stored energy are equalfor tablets formed within or out of the presence ofmagnesium stearate. The different relaxation forlubricated and unlubricated tablets in terms ofvolume or porosity change implies that bondingphenomena play an important role here.

The latter is important, because it counteractstablet relaxation (Van der Voort Maarschalk etal., 1996a).

In previous work, it has been demonstratedthat the interaction between particles in a tabletcan be derived from the relationship between ten-sile strength and porosity (Van der VoortMaarschalk et al., 1996b). Fig. 3 shows profilesfor both unlubricated and lubricated microcrys-talline cellulose tablets and it can easily be seenthat the relationship between tensile strength andporosity changes dramatically as an effect of thepresence of magnesium stearate. The profiles arefitted according to the Ryshkewitch–Duckworthrelation (Duckworth, 1953):

lnSS0

= −k ·o (4)

in which S is the tensile strength, S0 is the tensilestrength at zero porosity, o is the porosity of thetablets, and k is a constant, in previous workreferred to as ‘bonding capacity’ (Van der VoortMaarschalk et al., 1996b).

Table 2 lists the parameters obtained by meansof Eq. (4). These calculated values indicate thatthe bonding capacity as quantified by k is largerin compacts containing magnesium stearate. Thisis in sharp contrast with any other observation. It


Fig. 3. Tensile strength as a function of tablet porosity. Symbols as in Fig. 1.

implies that the previous interpretation of k is notpossible when dealing with a system containingmore than one component. Addition of magne-sium stearate to microcrystalline cellulose turnsthe powder into a binary system. It is noted thataccording to the correlation coefficients (depictedin Table 2), the Ryshkewitch–Duckworth equa-tion fits the data acceptably.

In a binary system of components A and B,there are three types of particle interaction: acohesive interaction A–A, a cohesive interactionB–B and an adhesive interaction A–B. For abinary system of a filler-binder (A) and a lubri-cant (B), the cohesive attraction A–A is strong,

because otherwise the material would not be agood filler-binder for direct compaction. It is gen-erally known that adhesive interaction A–B be-tween a lubricant such as magnesium stearate anda filler-binder is poor (Bolhuis and Holzer, 1996).The cohesive interaction B–B between magne-sium stearate particles is also poor and has thesame order of magnitude as the interaction A–Bbetween magnesium stearate and a filler-binder(Rowe, 1988).

Which type of interparticle attraction domi-nates the tablet strength is determined by the typeof consolidation of the main component. With theexception for filler-binders which show completelyplastic deformation (e.g. starch) or completelyfragmentation under load (e.g. dicalcium phos-phate dihydrate), most filler-binders show variousextends of fragmentation, which means that parti-cle fracture will be a function of the pressureexerted (Van der Voort Maarschalk and Bolhuis,1998a,b). As an effect of fragmentation, the dom-inating attraction type will change. During mixingwith magnesium stearate, the filler-binder will becovered with the lubricant and attraction typesB–B and A–B will dominate. At increasing com-paction pressures, however, attraction type A–Awill become more and more important as an effect

Table 2Tensile strength at zero porosity and constant k calculatedfrom the Ryshkewitch–Duckworth relation of tablets com-pressed from microcrystalline cellulose and sorbitol, respec-tively, with and without magnesium stearate

R2ak (−)PRIVATE Material S0 (MPa)

19.1 6.5 0.979457mcc (pure)4.9mcc+0.5% Mgst 8.0 0.993529

Sorbitol (pure) 11.5 6.6 0.9150078.8 0.994256Sorbitol+0.5% Mgst 9.1

a R2=correlation coefficient.


Fig. 4. Tensile strength of unlubricated () and lubricated (�) sorbitol compacts. Data plotted by (*) reflect unlubricated sorbitolcompacts which show beginning of capping.

of particle fragmentation, which successively re-places the continuous network of the lubricant bya network of the filler-binder. The increase of theimportance of attraction type A–A and the de-crease of the importance of the attraction typesA–B and B–B at increasing compaction forcesmeans that the bonding capacity k in Eq. (4) willbe a function of the compaction pressure andhence of the tablet porosity. Just as could beexpected, Table 2 shows a dramatic decrease intensile strength at zero porosity for lubricatedtablets. It should be realized, however, that ex-trapolation to zero porosity leads to unreliableresults because the slope of the relation in Eq. (4)is k. Consequently, it must be concluded that theuse of both k and S0 leads to an erronous inter-pretation of particle bonding when the materialunder study is a binary system of a filler-binderand a lubricant.

Without currently being able to quantify inter-particle attraction of binary systems, the data inthis paper indicate that relaxation of compactedmaterial increases when interparticle attraction isdisturbed. There are no reasons to accept that thedriving force for relaxation changes because it is a

bulk property of the particles. Consequently theincreased relaxation is purely a result of decreasedinterparticle bonding.

Fig. 4 depicts the relationship between tensilestrength and compaction pressure of tablets pro-duced from pure sorbitol and sorbitol lubricatedwith 0.5% magnesium stearate. Data plotted byan asterix reflect unlubricated sorbitol tabletswhich do not fit the corresponding curve. Thisdiscrepancy must be ascribed to beginning cap-ping of the tablets, according phenomena re-ported by, for example Marshall et al. (1993). Ascompared with microcristalline cellulose (Fig. 1),the lubricant has a much smaller effect on thetensile strength. Fig. 5 shows the relationshipbetween tablet porosity and compaction pressurefor unlubricated and lubricated sorbitol tablets,respectively. Porosities below zero proves densifi-cation of the material under pressure (Van derVoort Maarschalk et al., 1997b). Just as found formicrocrystalline cellulose tablets, the porosity un-der compression was similar for unlubricated andlubricated sorbitol tablets, which effect points to asimilar amount of stored elastic energy. In con-trast to microcrystalline cellulose tablets, however,


Fig. 5. Tablet porosity under pressure (open symbols) and after relaxation (closed symbols) as a function of compaction pressure.Symbols as in Fig. 4.

the relaxation of lubricated sorbitol tablets wasabout the same as that of unlubricated sorbitoltablets, which is in agreement to the small differ-ence in tensile strength of lubricated and unlubri-cated tablets.

The relatively small effect of admixing magne-sium stearate can easily be understood when theparticle structure of the material is known. Spray-dried g-sorbitol particles consist of aggregates ofvery small primary particles (Bolhuis andChowhan, 1996). Mixing the powder with magne-sium stearate during 30 min, the aggregates willcompletely be covered by the lubricant. The factthat the powder particles are aggregates impliesthat they have a brittle fracture behavior. Duringcompaction a lot of fresh, new surfaces will beformed. This means that the interparticle attrac-tion will predominantly be of type A–A. Appar-ently, fracture occurs already in the early stages ofcompression, which leads to isolated patches ofmagnesium stearate rather than a continuous net-work of it (Riepma et al., 1993).

The strong interparticle attraction between sor-bitol particles which counteracts relaxation is al-most similar for both lubricated and unlubricated

sorbitol tablets. Consequently there is only asmall effect of magnesium stearate on tabletstrength.

In conclusion, this paper shows that a part ofthe decrease in tablet strength by mixing pharma-ceutical powders such as microcrystalline cellulosewith magnesium stearate is caused by a moreextensive relaxation of the lubricated tablets.

As the amount of stored elastic energy wasfound to be simular for lubricated and unlubri-cated tablets, the increased relaxation of the lubri-cated tablets will be caused by their smallerparticle attraction and the consequent reducedresistance against relaxation. The increase oftablet relaxation as an effect of mixing withmagnesium stearate is smaller when using materi-als which fragment during the compaction pro-cess.

4. Nomenclature

k ; bonding capacityS (S0); tensile strength (at zero porosity) (Pa)o ; porosity


W(mix); calculated amount of stored energy in amixture (J/m3)W(cell); calculated amount of stored energy inmicrocrystalline cellulose (J/m3)W(mgst); calculated amount of stored energy inmagnesium stearate (J/m3)8cell; volume fraction of microcrystalline cellu-lose8mgst; volume fraction of magnesium stearatesc,cell; yield strength of microcrystalline cellulose(Pa)sc,mgst; yield strength of magnesium stearate(Pa)Ecell; elastic modulus of microcrystalline cellu-lose (Pa)Emgst; elastic modulus of magnesium stearate(Pa)

References

Bolhuis, G.K., Chowhan, Z.T., 1996. Materials for directcompaction. In: Alderborn, G., Nystrom, C. (Eds.), Phar-maceutical Powder Compaction Technology. MarcelDekker, New York, pp. 419–500.

Bolhuis, G.K., Holzer, A.W., 1996. Lubricant sensitivity. In:Alderborn, G., Nystrom, C. (Eds.), Pharmaceutical Pow-der Compaction Technology. Marcel Dekker, New York,pp. 517–560.

Bolhuis, G.K., Lerk, C.F., Zijlstra, H.T., De Boer, A.H., 1975.Film formation by magnesium stearate during mixing andits effect on tabletting. Pharm. Weekbl. 110, 317–325.

De Boer, A.H., Bolhuis, G.K., Lerk, C.F., 1978. Bondingcharateristics by scanning electron microscopy of powdersmixed with magnesium stearate. Powder Technol. 20, 75–82.

Duckworth, W.H., 1953. Discussion of Ryshkewitch paper byWinston Duckworth. J. Am. Ceram. Soc. 36, 68.

Duberg, M., Nystrom, C., 1982. Studies on direct compressionof tablets 6: evaluation of methods for the estimation ofparticle fragmentation during compaction. Acta Pharm.Suec. 19, 421–436.

Fell, J.T., Newton, J.M., 1968. The tensile strength of lactosetablets. J. Pharm. Pharmacol. 20, 657–658.

Heckel, R.W., 1961a. Density-pressure relationships in powdercompaction. Trans. Met. Soc. AIME 221, 671–675.

Heckel, R.W., 1961b. An analysis of powder compactionphenomena. Trans. Met. Soc. AIME 221, 1001–1008.

Marshall, P.V., York, P., Maclaine, J.Q., 1993. An investiga-tion of the effect of the punch velocity on the compactionproperties of ibuprofen. Powder Technol. 74, 171–177.

Rees, J.E., Tsardaka, K.D., 1994. Some effects of moisture onthe viscoelastic behaviour of modified starch during pow-der compaction. Eur. J. Pharm. Biopharm. 40, 193–197.

Riepma, K.A., Vromans, H., Lerk, C.F., 1993. A coherentmatrix model for the consolidation and compaction of anexcipient with magnesium stearate. Int. J. Pharm. 97, 195–203.

Rowe, R.C., 1988. Interaction of lubricants with microcrys-talline cellulose and anhydrous lactose—a solubilityparameter approach. Int. J. Pharm. 41, 223–226.

Strickland, W.A., Nelson, E., Busse, L.W., Higuchi, T., 1956.The physics of tablet compression 9: fundamental aspectsof tablet lubrication. J. Am. Pharm. Ass. Sci. Ed. 45,51–55.

Van der Voort Maarschalk, K., Bolhuis, G.K., 1998a. Improv-ing properties of materials for direct compression, Part 1.Pharm. Technol. Europe 10 (9), 30–35.

Van der Voort Maarschalk, K., Bolhuis, G.K., 1998b. Improv-ing properties of materials for direct compression, Part 2.Pharm. Technol. Europe 10 (10), 28–36.

Van der Voort Maarschalk, K., Vromans, H., Bolhuis, G.K.,Lerk, C.F., 1996a. The effect of viscoelasticity and tablet-ting speed on consolidation and relaxation of a viscoelasticmaterial. Eur. J. Pharm. Biopharm. 42, 49–55.

Van der Voort Maarschalk, K., Zuurman, K., Vromans, H.,Bolhuis, G.K., Lerk, C.F., 1996b. Porosity expansion oftablets as a result of bonding and deformation of particu-late solids. Int. J. Pharm. 140, 185–193.

Van der Voort Maarschalk, K., Zuurman, K., Vromans, H.,Bolhuis, G.K., Lerk, C.F., 1997a. Stress relaxation ofcompacts produced from viscoelastic materials. Int. J.Pharm. 151, 27–34.

Van der Voort Maarschalk, K., Zuurman, K., Van Steenber-gen, M.J., Hennink, W.E., Vromans, H., Bolhuis, G.K.,Lerk, C.F., 1997b. Effect of compaction temperature onconsolidation of amorphous copolymers with glass transi-tion temperatures. Pharm. Res. 14, 415–419.

Van der Voort Maarschalk, K., Vromans, H., Bolhuis, G.K.,Lerk, C.F., 1998. Influence of plasticizers on tablettingproperties of polymers. Drug Dev. Ind. Pharm. 24, 258–261.

Vromans, H., Lerk, C.F., 1988. Densification properties andcompactibility of mixtures of pharmaceutical excipientswith and without magnesium stearate. Int. J. Pharm. 46,183–192.

.

effect of magnesium stearate on bonding and porosity expansion of tablets produced from materials...

Documents