Pharmaceutical Development and Technology, 10:423–437, 2005
Copyright D 2005 Taylor & Francis Inc.
ISSN: 1083-7450 print / 1097-9867 online
DOI: 10.1081/PDT-200054462
Impact of Solid-State Properties on Lubrication Efficacy
of Magnesium Stearate
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
K. Phanidhara Rao, Garima Chawla, Aditya M. Kaushal, and Arvind K. Bansal
Department of Pharmaceutical Technology (Formulations), National Institute of Pharmaceutical
Education and Research, Punjab, India
The advent of high-speed tableting and slug capsule-filling
machines has ushered in an increasingly important role for the
lubricants to enact during manufacturing of dosage forms.
Although lubricants help in processing, they can also adversely
affect the flow properties and dissolution profile of the drug. It is
thus critical to maintain a balance between these two behaviors,
by understanding the underlying mechanisms and using their
For personal use only.
optimum concentration in the formulation. The source and
manufacturing process inculcate different solid-state properties
to magnesium stearate, the most commonly used lubricant,
leading to variations in its lubrication efficacy. However, there
has been no complete study relating the lubrication efficacy of
magnesium stearate to various levels of solid state. Hence, this
study was aimed at comprehensively scrutinizing the role of
molecular, particle, and bulk level properties of solid state on the
lubrication efficacy of magnesium stearate. A method based on
net work done during compression using texture analyzer, was
developed and validated to analyze its performance. Particle and
bulk-level properties were studied using microscopy, particle
size analysis, and particle surface area determination, and
molecular level was characterized using thermal, spectroscopic,
and crystallographic methods. Interplay of solid-state character-
istics such as particle size, degree of agglomeration, and crystal
habit were found to markedly influence the lubrication potential
of magnesium stearate.
Keywords magnesium stearate, lubrication efficacy, net work
of compression, solid-state characterization
Received 6 October 2004, Accepted 13 December 2004.
Address correspondence to Arvind K. Bansal, Department of
Pharmaceutical Technology (Formulations), National Institute of
Pharmaceutical Education and Research, Sector—67, S.A.S.
Nagar, Punjab 160 062, India; Fax: 91-0172-2214692; E-mail:
akbansal@niper.ac.in
423
Order reprints of this article at www.copyright.rightslink.com
INTRODUCTION
Most tablet formulations contain an additive designed
to reduce the intergranular and die wall friction, called
lubricants. Magnesium stearate is the most common
choice among lubricants.[1] These additives function by
interposing a film of low shear strength at interface be-
tween the tableting mass and the die wall. Lubricants with
low shear strength but cohesive tendencies perpendicular
to shear plane serve the purpose legitimately.[2] Lubri-
cants, besides plummeting frictional forces during manu-
facturing of compressed solid dosage forms, can affect
many other quality parameters such as weight variation,
hardness, disintegration, time and dissolution rate.[3] The
role of lubricants has assumed greater significance with
the introduction of high-speed tableting machines, as
shorter dwell times and faster compression rate pose
greater demands on their performance.[4] Similarly, their
lubrication efficiency critically affects weight variation in
dosator- and tamping-based capsule-filling machines.[5]
Magnesium stearate improves the performance of adhe-
sive materials (lactose, calcium carbonate, etc.), prevents
material buildup on tampering faces and shafts, and
decreases the ejection force leading to reduced wear of
these automatic slug-forming equipments. However,
magnesium stearate is known to adversely affect the flow
properties and dissolution rate, both in tablets and
capsules, due to its hydrophobic nature. Higher levels of
magnesium stearate in capsules may lead to the formation
of hydrophobic powder beds that do not disperse after the
dissolution of capsule shell. Hence, there is a need to
control the levels of magnesium stearate in the formulation
to satisfy both the content uniformity and dissolution
acceptance criteria, by understanding the parameters
controlling the performance of magnesium stearate.
Magnesium stearate most closely resembles an ideal
lubricant with low friction coefficient, large ‘‘covering
 424 K. P. Rao et al.
potential,’’ and a high melting point.[6] However, it is
reported to suffer from source, manufacturer-to-manufac-
turer, and batch-to-batch variation in their physicochem-
ical properties that are mainly attributable to their solid-
state properties.[7 – 9] Solid-state properties of a substance
can be classified into three levels—molecular (crystalline/
amorphous forms), particle (habit, size distribution), and
bulk level (density, porosity). The relationship between
the three levels is such that an intentional or unintentional
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
alteration at one level may significantly influence the
other two, thereby affecting the physicochemical and
physicotechnical performance. This could also translate
into severe processing problems, keeping in view the
multiple functions performed by the lubricant. Attempts
have been made to understand the critical contributors to
the performance of lubricant, and there are numerous
contradictory reports interrelating surface area, particle
size,[10,11] and lamellar structure[12] to the lubrication
efficacy. However, the previous reports do not provide a
collective picture of all solid-state levels and how change
at one level may be interrelated to the effect seen at the
other level. Therefore, there is a need to assess the
lubrication efficacy by a suitable method and to be aware
For personal use only.
of the critical parameters contributing to variation in
lubrication efficacy, thus allowing selection of optimum
grade and quantity of the lubricant.
In this study, a simple method based on the net work
done using texture analyzer has been developed to
evaluate the lubrication efficacy, and an attempt has been
made to characterize the ‘‘best’’ and the ‘‘worst’’
performers, and relate their performance to solid-state
properties. Compaction involves two distinct stages—
compression (reduction in bulk volume due to displace-
ment of gaseous phase) and consolidation (increase in
mechanical strength due to cold welding and fusion
bonding between the particles).[2] Any lubricant needs to
perform during both the stages. This method based on use
of texture analyzer mimics the initial compression stage
of tableting and during slug formation in dosator- and
tamping-type, slug-forming, capsule-filling machines. Six
commercially available samples of magnesium stearate
were studied for their performance as a lubricant. All
three levels of solid-state properties viz. molecular,
particular, and bulk were investigated systematically to
identify the parameters critical for the lubricant efficacy
of magnesium stearate.
EXPERIMENTAL
Materials
Lactose (Meggle, Germany), microcrystalline cellu-
lose (PH 102, FMC International, Little Island, Cork,
Wallingstown, Ireland), dicalcium phosphate (Engranule,
Enar Chemie Pvt. Ltd., Gujarat, India), and six commer-
cial samples from different manufacturers (designated as I
to VI) of magnesium stearate were used for the study. Sam-
ple I was of British Pharmacopoeia (BP) grade; samples II,
IV, and V were of National Formulary (NF) grade; and
samples III and VI were of Indian Pharmacopoeia (IP) grade.
METHODS
Development of Method for Testing
Lubrication Efficacy
Settings of Texture Analyser
All experiments were done with a 50-kg load cell
using a 10-mm, flat-faced, die punch set, and texture
analyser (TA-XT 2i, Stable Microsystems, Godalming,
Surrey, UK) was calibrated for both the distance and the
load cell before starting an experiment. The following two
types of methods were used.
Method 1. The force required to travel a particular
distance (4 mm) was measured. Net work done was
calculated from area under the curve. Initial position of the
upper punch was kept constant in all the experiments using
the calibration scale. The other parameters were fixed as-
punch travel speed of 2 mm/sec before and after the test,
0.1 mm/sec during the test, and a trigger/stimulus was set
to ‘‘activate after traveling a distance of 1 mm.’’
Method 2. Distance traveled by the punch to attain
a particular force (475 N) was determined. Net work done
was calculated as in method 1. Stimulus was set to 0.05 N
(‘‘just after touching the bed’’), which triggered the
recording. The remaining texture analyser settings were
similar to that of method 1.
Selection of Suitable Excipient
Three diluents—dicalcium phosphate, lactose, and
microcrystalline cellulose—were tested for their suitabil-
ity for use in further experiments. Accurately weighed
quantities of dicalcium phosphate, lactose, and microcrys-
talline cellulose were compressed either neat or with 1%
(w/w) magnesium stearate (grade II) using texture
analyser, according to method 1. The diluents were mixed
in geometric progression with magnesium stearate and
then blended for 5 min in a double cone blender (VDM
4SP, Kalweka, Ahmedabad, India) at 150 rpm prior to
compression. Excipient showing minimum variation
(%RSD) based on 10 experiments was selected for fur-
ther experimentation.
 Solid-State Properties and Magnesium Stearate
Validation of Texture Analyser-Based Method
Intraday and interday variability were determined, for
both methods 1 and 2, using the work of compression
keeping the maximum distance traveled by upper punch
constant (4 mm) and measuring the distance traveled to
attain a fixed force (475 N), respectively. Reproducibility
and repeatability were checked by performing the
experiment on 3 successive days and thrice on a day.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
Comparative Performance of Various Samples of
Magnesium Stearate
The effect of different magnesium stearate samples
on intergranular friction, and on both the intergranular and
die wall friction, was determined by different methods.
Method A for Intergranular Friction. Method 1
was followed where the maximum distance traveled was
4 mm during the compaction to study the intergranular
friction. A weight of 580± 5 mg (weight enough to fill
a 10-mm die) of each blend having different samples of
magnesium stearate (1% w/w) was filled into the die
For personal use only.
cavity and compressed. First, five compressions were done
to coat the die wall with magnesium stearate. Using the
same blend, an additional six compressions were given to
determine the average of net work done required to travel a
distance of 4 mm for that blend. The amount of net work
done needed to compress only dicalcium phosphate was
used as the control while switching over to the other blend
with different magnesium stearate sample.
Method B for Both Intergranular and Die Wall
Friction. This experiment was done to study the effect
of different magnesium stearate samples on the combined
effect of intergranular and die wall frictions. The ex-
periment was done using method 2, where the maximum
force to be achieved was set at 475 N. One percent (w/w)
blends of different magnesium stearate samples with
dicalcium phosphate were prepared in the similar fashion
as in method A. The method of experimentation was
identical to method A, but without any preliminary
compression cycle to coat the die wall. The die wall was
cleaned with cotton swabs dipped in isopropyl alcohol to
remove a layer of magnesium stearate deposited from
previous experiment.
Flowability Test
Density Measurements
The effect of different magnesium stearate samples
on the improvement of flow properties was studied by
425
plotting bulk density and tapped density curves against
the number of taps. Tapped density was plotted against
number of taps by performing the tap density experiment
on blends filled in a graduated measuring cylinder. The
cylinder was subjected to 2000 taps at a rate of 300 taps/
min on Electrolab1 tap density apparatus (ETD-1020,
Mumbai, India). The volume was recorded at 2, 4, 8, 16,
50, 100, 300, and 750 taps. The profiles were compared
with that of neat dicalcium phosphate.
Determination of Flow Rate
Flow rate of the blend was determined using Erweka
Granulate tester (Type: GT; Erweka, Heusenstamm,
Germany). Time taken for 100 g quantity of blend to
flow through a 6-mm orifice was recorded in triplicates.
Solid-State Characterization
Solid-state characterization of commercially sourced
samples was studied at each of the three levels of solid state.
Particle and Bulk Level
Microscopy. Microscopic studies were done using
optical microscope (Leica DMLP, Leica, Wetzlar,
Germany) attached to an imaging system (DC-300,
Leica). In optical microscope, images were acquired at
a magnification of 630
(63
10) using the Image
Manager (IM50, V1.20 Twain module, Leica, Germany)
software. Crystal habit, state of agglomeration, and
relative size were observed.
Particle Size Analysis. Particle size analysis was
carried out using laser particle sizer (Analysete 22, Fritsch
GmbH, Idar Oberstein, Germany) having a measuring
range of 0.18 to 102.89 mm, with a helium-neon laser
source with less than 5 mW power and Version-C software
to compute the particle size distribution. Around 20 mg of
the sample was dispersed in water, ultrasonicated with
stirring at 75 rpm to disrupt the agglomerates and pumped.
The equipment works on the principle of diffraction
patterns (Fraunhofer rings) produced when a monochro-
matic laser light strikes the boundary edge of the two
media with a different refractive index. The diameter of
the particle is inversely proportional to the radius of the
first circle of the light produced due to diffraction and is
calculated using the following equation:
Particle diameter ¼ 1:81
f
l=Ro
where f, l, and Ro are the focal length of the imaging lens,
wavelength of laser light, and radius of the Fraunhofer
ring, respectively.
Particle Surface Area Determination. Specific
surface area was determined for the selected samples
 426 K. P. Rao et al.

using Coulter surface area determination apparatus RESULTS AND DISCUSSION

(Coulter SA 3100, version 2.11; Fullerton, CA, USA),
which works on the BET (Brunauer, Emmett, and Teller) Method Development for Testing
principle. The experiment was done using both single- Lubrication Efficacy
point and multipoint methods. Weighed quantity (about
200 mg) of sample was degassed at 40
C for 1 hr and the Selection of Suitable Excipient
weight of sample after degassing was taken as the initial
weight. Amount of nitrogen adsorbed was calculated by The %RSD of maximum force achieved on the punch
the difference in weight before and after adsorption. The to travel a certain distance showed that dicalcium
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12

specific surface area was calculated using the
BET principle.
Molecular Level
Thermogravimetric Analysis. Thermogravimetric
analysis (Mettler Toledo, TGA/SDTA 851, Greifensee,
Switzerland, Mettler STAR, version 5.1) was done for
each sample placed in alumina crucible, from a
temperature range of 30
C to 210
C at the rate of 5
C/
min under nitrogen purge (20 mL/min).
Differential Scanning Calorimetry. Differential
scanning calorimetry (DSC; 821 Mettler Toledo, Grei-
fensee, Switzerland, operating with STAR, version 5.1)
was done between temperatures  10
C to 210
C at a
For personal use only.
phosphate showed minimum degree of variation and
was therefore selected for further experimentation. The
results are shown in Figure 1.
Validation of Texture Analyser-Based Method
The validation exercise confirms that the method is
suitable for use and produces results that are accurate and
precise (i.e., the system works as it ought to work). The
texture analyser-based methods were validated for
performance by checking interday and intraday variations.
Both methods 1 and 2 showed a %RSD of < 3%; interday
%RSD 2.52 and 1.01, respectively, and intraday %RSD
2.15 and 0.75, respectively.

heating rate of 5
C/min and nitrogen purge of 100 mL/

min. Samples (2 – 5 mg) were tested in sealed aluminum Comparative Performance of Various Samples of
pans with a pinhole, and an empty pan was used as a Magnesium Stearate
reference. The temperature axis and cell constant of
DSC were previously calibrated with a high-purity Lubrication efficacy was studied using method A (for
standard of Indium. intergranular friction) and method B (for both intergran-
Hot Stage Microscopy. Optical microscope was ular and die wall friction) to select the magnesium
attached with a hot stage (Leica LMW 50, Leica, Ger- stearate samples (‘‘most effective performer’’ and ‘‘least
many) to observe the thermal events. The samples were
observed with and without silicone oil, covered with a
glass cover slip and heated. Images were acquired using the
Image Manager (IM50, V1.20 Twain module) software.
FTIR Studies. Fourier-transform infrared (FTIR)
spectra were recorded using the KBr pellet method in the
region 400 to 4000 cm 1 on PerkinElmer spectropho-
tometer (FTIR multiscope/synthetic monitoring system,
PerkinElmer, Buckinghamshire, UK), equipped with
Spectrum v3.02 software. Background spectrum was
collected under identical conditions. The spectrum for
each sample (an average of 16 coadded scans) was
recorded at a resolution of 4 cm 1, Fourier transformed
and ratioed against background interferogram.
Powder XRD Studies. Powder x-ray diffraction
Figure 1. Relative standard deviation (n = 10) of maximum
(XRD) patterns of samples were recorded at room
force achieved on the punch to travel a certain distance by
temperature on x-ray powder diffractometer (Seifert, different excipients and excipient – magnesium stearate blend.
Iso-Debyeflex 2002, Ahrensburg, Germany) using Cu LAC, lactose; MCC, microcrystalline cellulose; DCP, dicalcium
Ka radiation (l =1.54184’) at 30 kV, 20 mA passing phosphate; LAC + MS, lactose – magnesium stearate blend;
through nickel filter. Data were collected in a continuous MCC + MS, microcrystalline cellulose – magnesium stearate
scan mode with a sweep of 1.2
/min over an angular blend; DCP + MS, dicalcium phosphate – magnesium stearate
range of 5 to 70
2q. blend.
 Solid-State Properties and Magnesium Stearate
effective performer’’) for further characterization and to
evaluate the role of various solid-state levels on the
performance of magnesium stearate. Work of compres-
sion can be used as one of the parameters for estimating
lubrication efficacy. Net work done was determined either
by measuring the maximum achievable force or maxi-
mum distance traveled by upper punch. One parameter is
determined to keep the other constant. Method A provides
an assessment of intergranular friction, as the initial stage
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
of compression involves rearrangement and repacking of
particles, with the contribution of die wall friction being
negligible. Method B studies the later stages of compres-
sion, wherein contribution of both intergranular and die
wall friction need to be accounted for. Figure 2a and 2b
show the net work done to travel a distance of 4 mm and
the distance traveled by upper punch at constant force,
respectively, and the order of efficacy of different samples
of magnesium stearate. The less work needed for a blend,
the better is its lubrication efficacy, as seen in sample I,
which was found to be superior to all other samples.
For personal use only.
Figure 2. Measurement of lubrication efficacy of magnesium
stearate samples. (a) Method A: net work done by the upper
punch to travel a distance of 4 mm. (b) Method B: distance
traveled by upper punch at force of 475 N. Note: The error bars
represent the %RSD of n = 6.
427
The results of net work done were subjected to one-
way analysis of variance (ANOVA) to check whether
there was any significant difference in the net work done,
and hence, the lubrication efficacy of different blends.
The data were found to show normal distribution even at
99% confidence limits. Both Tukey (more conservative—
less likely to show statistically significant difference) and
Student-Newman-Keuls (less conservative—more likely
to show statistically significant difference) tests illustrated
the difference in the mean net work done among the
‘‘effective performers’’ (samples I and II) and the ‘‘less
effective performers’’ (samples III and IV) was more than
that expected by chance (P< 0.05), hence showing a
statistically significant difference between them.
The results obtained from method B, by measuring
the distance traveled by the upper punch to achieve a
specified force, were in close agreement to that of method
A (Figure 2b). Because the upper punch moves with a
constant speed, time to achieve the maximum force can be
used to judge the lubrication efficacy. The longer the
time, the more the distance traveled and the better the
lubrication efficacy. To circumvent the problem of
differences in the bulk densities of different samples, all
blends including dicalcium phosphate were precom-
pressed to attain a constant bed volume. This also helps
normalize the effects caused due to differences in
particulate arrangement after filling the die cavity. Two
best (samples I and II) and two worst (samples III and IV)
performers were selected for further characterization to
recognize the solid-state properties imperative for perfor-
mance as a lubricant.
Flowability Test
Flow characteristics of blends were characterized in
their static and dynamic stages by performing density and
flow rate experiments, respectively.
Density Measurements
The plot of tapped density against number of taps is
shown in Figure 3. Magnesium stearate enhanced the rate
of packing of blends on tapping. Therefore, it can be
concluded that magnesium stearate facilitates in reaching
an ultimate value of density in a shorter period. However,
no appreciable difference was found between different
samples of magnesium stearate.
Flow Rate Measurements
Results of the flow rate were found to be interesting
because ‘‘poor performers’’ in a lubrication efficacy test
 428 K. P. Rao et al.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12

Figure 3. Density measurements – comparative behavior of different blends on tapping.

improved the flow rate to a greater extent as compared with
the ‘‘better performers.’’ Samples I (33.5 s) and II (33.9 s)
took longer time to flow, in contrast to samples III (32.0 s)
and IV (31.8 s). The data were treated with one-way
ANOVA and were found to show normal distribution.
For personal use only.
reverse correlation between the lubrication efficacy and
tendency to improve flow of magnesium stearate samples
was seen (Figure 4). During flow of a powder, unlike as in
compression, much space is available around the particles,
facilitating rolling and sliding of particles over each other

Tukey and Student-Newman-Keuls tests exemplified that
there was a statistically significant difference in the flow
rate of samples I and II versus samples III and IV blends, at
99% confidence intervals with P<0.05. These differences
can be attributed to the intrinsic cohesiveness reflected as
agglomeration of the lubricants. Optical microscopy
(discussed later) revealed that sample III contained more
agglomerates, which might cause a ‘‘ball bearing’’-like
effect, facilitating rolling of particles over each other. A
with less resistance and friction. In contrast, during
compression, movement of particles is restricted, and the
relative movement of static and dynamic planes in opposite
direction generate resistance, restricting the rolling move-
ment. Therefore, the particle characteristics required to
improve flowability and to diminish the frictional forces
during compression are bound to be different.
Flow of powders involves the rolling or sliding of
particles over one another, with continuous change in

Figure 4. Lubrication and flow performance of good performer and bad performer blends of magnesium stearate. The bars show
the average net work done by the upper punch to travel a distance of 4 mm, and the points depict the results of flowability test (the
values are the time in seconds taken for 100 g of blend to flow through the sample holder [orifice diameter = 6 mm]). Note: The error
bars represent the %RSD of n = 6 for net work done, and n = 3 for flowability test.
 Solid-State Properties and Magnesium Stearate
point of contact. Compared with this, compaction
involves relatively static particles with relative movement
of static and dynamic planes in the opposite direction
against a lot of resistance. The particle characteristics
required to improve flowability and to decrease frictional
forces during compression are expected to be different.
Magnesium stearate samples having particles in a
deagglomerated state will distribute quickly and perform
better in the lubrication efficacy test. However, a smaller
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
particle size will lead to poorer flow properties.
Solid-State Characterization
Lubrication efficacy of magnesium stearate has been
associated with factors such as chemical composition and
its physicochemical properties, such as the specific
surface area, particle size distribution, moisture content,
hydration state, crystal structure, and crystal habit.[13,14]
Earlier reports have shown that fatty acid content (stearic
and palmitic) do not influence magnesium stearate to an
extent to alter its lubrication efficacy.[9,14,15] Magnesium
stearate samples with different chemical compositions
have been reported to exhibit the same lubrication ef-
For personal use only.
ficacy, and the same batch of the lubricant with different
physicochemical properties has shown different perfor-
mance patterns, hence suggesting the role of properties
other than composition in the lubrication efficacy of mag-
nesium stearate.
Different reports claim the role of diverse properties
such as particle size and surface area,[9,14] or moisture
content and crystalline structure[15] or bulk density,[6] to
influence its lubrication properties. Hence, the study was
aimed at comprehensively studying the role of all solid-
state levels on the lubrication efficacy of magnesium
stearate. Characterization at different solid-state levels of
Table 1
Particle size and specific surface area of selected grades of magnesium stearate
Experiment I II
Particle size (mm) d1% > 24.12
d50% > 6.12
d95%> 1.12
BET one-point surface area (m2/g) 5.49
BET specific surface area (m2/g) 6.63
Correlation coefficient 0.9946
Monolayer volume,* Vm at STP# (cc/g) 1.5224
Note: From texture analyzer studies, it was found that samples I and II are high-efficiency lubricants in contrast to samples III and
IV that are less effective.
*The larger the surface area, the more volume of gas adsorbed.
#
Standard temperature and pressure.
429
the selected samples was done in two steps to determine
the critical contributors:
1. Characterization of particle- and bulk-level parameters
(i.e., state of agglomeration, surface area, particle size,
and crystal habit)
2. Characterization of molecular-level parameters (i.e.,
hydration state and crystal structure)
Studies on Particle Size, Shape, and Specific
Surface Area. Because lubrication mechanism is a
surface phenomenon, it can be correlated to properties
such as particle size and specific surface area. The particle
size and surface area of selected magnesium stearate
samples I to IV are reported in Table 1. The difference in
the particle size and area demonstrates the batch-to-batch
and manufacturer-to-manufacturer variability in magne-
sium stearate. The order of samples in terms of specific
surface area was sample I> III >II > IV, and in terms of
particle size was sample IV > Iffi II ffi III. Smaller particles
tend to accommodate themselves well on other particles
and form an effective coating, thereby reducing the
frictional forces. Sample III, despite having particle size
distribution equivalent to I and II, as well as surface area
better than II, did not perform well. This can be attributed
to presence of agglomerates as was evident in optical
microscopy. The sample preparation during determination
of particle size distribution (PSD) involves sonication that
breaks the agglomerates into individual particles, and
thus, does not account for agglomerates present in the
original samples. Effective coating, however, can signif-
icantly increase the disintegration and dissolution times,
as well as decrease the hardness of the tablets,[16] because
the film makes it difficult for interparticle bonding to
occur.[13] Milling, which leads to a change only in particle
Sample of magnesium stearate
III IV
d1% > 27.41 d1% > 20.37 d1% > 84.63
d50%> 7.21 d50%> 5.85 d50%> 24.42
d95%> 1.22 d95%> 1.11 d95%> 1.38
2.75 3.89 1.31
3.24 4.52 1.66
0.9965 0.9976 0.987
0.7445 1.0395 0.3814
 430 K. P. Rao et al.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
For personal use only.

Figure 5. Optical photomicrographs of different samples of magnesium stearate at a magnification of 630
. (a) Sample I,
(b) sample II, (c) sample III, and (d) sample IV.

size without causing change in any other property
(moisture content, hydration state, etc.), improved lubri-
cation efficacy confirming direct correlation.[13,14] It has
also been observed that the particle size is directly related
to radial transmission ratio (R) and tensile strength,[17]
which are a measure of lubrication efficacy and bond
formation, respectively.
Optical Microscopy. Optical microscopy indicat-
ed that samples I and II have smaller particles as
compared with samples III and IV. The shape of crystals
was flat and platelike in samples I and II, whereas
agglomerates were observed in sample III (Figure 5).
Optical microscopy provides a better reflection of the
actual particle size distribution, which gets ‘‘masked’’
during particle size distribution determinations using
automated processes involving sonication during sample
preparation, leading to deagglomeration.
Thermal Analysis. DSC and thermogravimetric
analysis (TGA) were performed on magnesium stearate
samples to characterize the type of hydrate and to study
the nature of water present because it is stated that
lamellar dihydrates are more efficient lubricants.[18] None
of the samples showed endotherm corresponding to the
adsorbed water (which would have been converted into
ice at subzero temperatures) when subjected to low
temperature DSC indicating nonhygroscopic nature of
magnesium stearate. More than one peak of dehydration
in DSC indicated the presence of water in different
thermodynamic states.
In sample I, desolvation in TGA was observed as an
endotherm in DSC. In TGA (Figure 6a), the weight loss
was observed up to 107
C, indicating that the first three
endotherms in DSC correspond to dehydration. The
number of water molecules calculated by TGA was found
to be 1.8. Both cleavage of interactions of water with the

Figure 6. DSC and TGA curves of various samples of magnesium stearate. (a) Sample I, (b) sample II, (c) sample III, and
(d) sample IV. The values in parenthesis represent the percentage weight loss, with a standard deviation of ± 0.5. Temperature range
was 10
C to 210
C for DSC and 35
C to 210
C for TGA. Sample analysis was performed in pin-holed aluminum pans in DSC and
alumina pans in TGA at a heating rate of 5
C/min. All samples contain varying proportions of water molecules, showing the
inconsistency in different samples from different manufacturers.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
For personal use only.

Solid-State Properties and Magnesium Stearate

431
 Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
For personal use only.

432

Figure 6.
K. P. Rao et al.

Continued.
 Solid-State Properties and Magnesium Stearate 433

Mg2 + and the evaporation of free water, thus formed,
contribute to the sum of enthalpies of peaks.
Sample II had no sharp thermal events indicating
loosely bound water and less crystallinity. The two peaks
in DSC can be attributed to water loss because they
coincide with the thermal events of TGA. Water loss was
observed till 115
C in TGA, and the corresponding
thermal events in DSC were found to merge with melting.
The prominent peak in sample III shows both dehydration
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
that samples I and IV have their water almost tightly
bound, and prominent water loss starts at 72
C. Sample II
with large bulk of loosely bound water, showed 82%
weight loss by the time it reached a temperature of 85
C.
Sample III behaved differently because the water loss was
observed until 130
C, most of it being lost between 94
C
to 116
C, indicating more tightly bound water compared
with other samples (I, II, IV) that showed weight loss only
up to 115
C.

and melting. A close observation of this peak (Figure 6c)
revealed that it starts with a small hump, indicating the
dehydration of tightly bound water at 82
C, followed by
simultaneous dehydration and melting. Single broad
endotherm in sample IV (Figure 6d) indicated merging
of dehydration and melting events.
Calculation of water molecules from TGA (Figure 6)
indicated that none of the samples are dihydrate in nature,
although the loss of water was observed in two stages in
sample I, II, and III. The FTIR studies however, showed
that samples IV and I are dihydrates with respect to their
crystal structure (discussed later). This anomaly is
explained by considering the magnesium stearate samples
as mixtures of different hydrated forms with varying
For personal use only.
From the previous results, various possibilities can be
predicted for the arrangement of water molecules. They
can be present as either noncrystalline water (adsorbed
water) or as crystalline water (within the crystal). The
crystalline water can be in different thermodynamic states
as pure dihydrate or higher hydrates. Furthermore, a
mixture of different hydrates may have water at different
energy levels.
FTIR Studies. Table 2 summarizes the results of
FTIR. All magnesium stearate samples showed a carbonyl
peak at about 1570 cm 1, indicating the interaction of Mg
atom with the fatty acid chain changing the unionized
state of acid to an ionized state. This was confirmed in a

proportions of water molecules. This also explains the
reason for the number of water molecules getting
expressed as fractions. Sample I and IV are close to
dihydrate in terms of their moisture content. TGA showed
(Figure 6) that all samples except sample IV have water at
two or more different thermodynamic states because the
water loss was observed in more than one step. The
simultaneous differential analysis (SDTA) curves showed
previously reported work, by observing a similar peak
related to carbonyl group in the stearic acid sample at
1700 cm1.[19] Bands between 3600 cm1 and 3100 cm1
provide evidence for crystal water. Bands related to fatty
acid portion showed the presence of a doublet at
720 cm1 in samples I and IV, indicating orthorhombic
or monoclinic structure in them.[20] This particular band is
due to the CH2 rocking mode, and its appearance is

Table 2
Summary of Fourier transform infrared results

Sample
Peak (cm 1), threshold—5%,
normalized to 0 – 100 scale Vibration I II III IV

3600 – 3100 Water of 3447 (m)

crystallization 3525 (m)
3447 (m) 3525 (m)
3546 (m) 3567 (m)
35,967 (m) 3546 (m)
3567 (m)
3567 (m)
3300 – 2500 -OH stretching 2957 (s)
2957 (s) 2957 (s) 2957 (s)
2850 (s)
2900 C – H symmetric 2918 (s) 2917 (s) 2917 (s) 2918 (s)
1700 C = O asymmetric — — — —
1616 – 1540 COO asymmetric 1542 (s) 1574 (s) 1574 (s) 1543 (s)
720 CH2 rocking 723 (m) (d) 721 (m) (sig) 722 (m) (sig) 723 (m) (d)

Note: s, strong; m, moderate; sig, singlet; d, doublet. Standard values of bands were taken from Ertel and Carstensen (1988).[15]
 434 K. P. Rao et al.
related to the packing of the hydrocarbon chains within
the crystal.[19] Furthermore, it can be concluded that these
samples are dihydrates because the monoclinic and
orthorhombic packing was shown to be consistent with
this form in the metallic salts of stearic acid.[21] The TGA
results also showed that the number of water molecules is
close to 2. The other varieties (samples II and III) can be
concluded to have either hexagonal or triclinic packing
because a singlet is observed at 720 cm 1. Therefore,
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
they can be expected to have anhydrate or partially
hydrated form.
The FTIR analysis of samples dried at 90
C under
vacuum for 20 hr showed loss of water bands above
3100 cm1 in all samples except sample III. This sample
showed two bands at 3563 and 3468 cm1, indicating the
strong interaction of water with magnesium. Even in
TGA, the water loss was observed until 130
C, indicating
the strongly bound nature of water in this sample. Band
For personal use only.
Figure 7. XRD pattern of magnesium stearate samples from 2q 5
to 70
. (a) Sample I, (b) sample II, (c) sample III, and
(d) sample IV.
at 1542 cm1 of samples IV and I shifted to 1579 and
1594 cm1, respectively, after drying. This shift was not
observed in the other two samples. It indicates that water
loss from the higher hydrates (samples I and IV) and
transformation to lower hydrates is responsible for this
shift. Water loss leads to the absence of interaction with
water, which otherwise satisfies the positive charge on the
Mg2 + ion to some extent. Therefore, water loss leads to
strengthening of the bond between fatty acid chain and
Mg2 +. It resulted in the COO asymmetric stretch to shift
to higher wave numbers. Because other samples are
already in a lower hydration state, they did not show
much change in this band. It was also observed that the
720-cm1 band became singlet upon drying in samples I
and IV confirming the previous result. It can also be
concluded that the dehydration is accompanied by the
change in crystal structure, and thus, crystal water plays
an important role in crystal system formation. Thus, from
 Solid-State Properties and Magnesium Stearate 435

Table 3
Correlation between solid-state properties at different levels and lubrication efficacy of magnesium stearate samples

Bulk level Particle level Molecular level

Sample Agglomeration a
PS (mm) a 2
SSA (m /g) Crystal habit Water molecules 720 cm 1 band d spacing

I XX No XX 6.12 XX 6.63 XX Plate XX 1.8 XX Doublet XX 11.7

II XX No XX 7.21 X 3.24 XX Plate  0.6 X Singlet X 10.5
III  Yes XX 5.85 X 4.52  Irregular  1.26 X Singlet X 10.8
IV XX No  24.42  1.66  Irregular XX 1.86 XX Doublet  9.9
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12

Note: The unshaded cells with XX show favorable properties of magnesium stearate as a lubricant, the dark-shaded cells with 
show the different properties at various levels of solid state that hamper the lubrication efficacy of magnesium stearate, and the light-
shaded cells with X show the intermediate behavior of magnesium stearate.
a
PS (d50%), particle size; SSA, specific surface area.

TGA and FTIR studies, it is evident that samples I and IV
have dihydrate structure, whereas samples II and III have
mono or still lower hydrate structures.
X-Ray Diffraction Studies. XRD patterns of the
four samples were studied to know the difference in
crystal structure and its influence on the lubrication
efficacy. Variation in the composition and proportion of
For personal use only.
particle-level characteristics, which fundamentally influ-
ence the layer-forming tendency of the lubricant. As
mentioned earlier, deagglomerated particles, smaller
particle size, greater specific surface area, and platelike
crystal habit are the optimum characteristics that can
attribute good lubrication efficacy. The microscopic
studies revealed that samples I and II have platelike
crystal habit. Therefore, they can delaminate during

the fatty acids makes it difficult to get consistent XRD
pattern. In general, the strong and fine harmonics are
typical of single well-crystallized form, whereas broad
reflections are indicative of poorly crystallized form.
Therefore, sample II with only one prominent peak at
2q =21
can be said to be ‘‘poor in crystalline content.’’
Figure 7 shows the XRD pattern of magnesium stearate
samples. Samples I and IV indicated a crystal structure as
a mixture of dihydrate and other hydrates, whereas sample
III indicated a partially dehydrated dihydrate.[7,15,17] The
amount of moisture content (crystalline water) dictates the
crystal structure, and it was found to depend on the drying
rate and the pH of the precipitation medium during
preparation of magnesium stearate.[22] Only a small
difference (0.6 moles of water) in the number of
molecules of water per molecule of magnesium stearate
is sufficient to convert a poorly crystalline form into a
well-crystallized form. Furthermore, the crystal long
spacing is a key property because it attributes shearability
to the lubricant.[23] Highest d spacing (Table 3) shown by
sample I and lowest by sample IV correlate well to their
good and poor lubrication efficacy, respectively.
Solid-State Levels and Performance of Magnesium
Stearate: A Reminisce
Magnesium stearate acts by coating the particles and
thereby reducing the interactions between moving surfac-
es. Particle size, surface area, and the crystal habit are the
mixing and compression, which ultimately leads to
reduced friction. Second, lubricants act in the later stages
of compression when the contact area between the
particles and the die wall increases. At this stage,
lubricants having sufficient shearability only can perform
the function. Shearing involves elongation of one of the
largest lattice spaces of the crystal. So, crystals having
higher lattice space perform better. The order (sample
I >sample III> sample II > sample IV) of highest lattice
spacing (d spacing) values showed the same array. Hence,
higher hydrates were found to improve the lubrication
efficacy by delaminating to a greater extent. Sample I was
found to be a dihydrate with greater long spacing
compared with other samples and showed good lubrica-
tion efficacy. However, the lubrication efficacy of sample
IV was not in agreement with this principle. The reason
can be attributed to the other solid-state properties of
different levels that overshadow the molecular level. The
difference in their d spacing can be owed to the different
proportions of fatty acids, which ultimately dictates the
crystal structure. These molecular-level characteristics
depend on the source of stearic acid, manufacturing, and
storage conditions. Each property has different extent of
contribution toward lubrication efficacy, and it is difficult
to compare the samples, which vary in more than one of
the previously mentioned variables.
The extreme performers, samples I and IV, have
similar crystal structure. The difference in lubrication ef-
ficacy is purely attributed to the difference in specific
surface area, indicating that it plays an important role in
 436 K. P. Rao et al.
lubrication. The d spacing is more in sample I compared
with all other samples, indicating its importance as
mentioned earlier. Samples II and III are almost similar
in FTIR studies, particle size, and specific surface area.
Interestingly, although sample III has slightly higher spe-
cific surface area and smaller particle size, the lubrication
efficacy is less compared with sample II, having less
specific surface area. This can be explained on the basis of
the bulk- and particle-level characteristics. At the bulk
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
level, sample II had no agglomerates, allowing it to be
adsorbed onto the other excipient particles in the blend. At
the particle level, the platelike crystal habit favored it to
spread uniformly in comparison to sample III, which had
agglomerates and irregular crystal habit. Because not
much difference was observed in the particle size and
specific surface area of sample III over sample II, the
previously mentioned affects could not be overshadowed,
leading to better performance of sample II over sample III.
Although sample II is a monohydrate with a specific
surface area half the value of sample I, it performed well.
The reason can be explained in terms of crystal habit and
lack of agglomerates and considerable d spacing.
It can be said that the inferior performance of sample
For personal use only.
IV is due to the presence of irregular crystal habit, large
particle size (less specific surface area), and smaller
lattice spacing.
Therefore, the criterion for good lubrication can be
given in terms of different levels of solid state:
Bulk level: Absence of agglomerates
Particle level: Smaller particle size, larger specific surface
area, platelike crystal habit
Molecular level: Dihydrate in nature, especially water
should present in two thermodynamic states, larger
d spacing
CONCLUSION
The commercially available magnesium stearate,
used as a lubricant in pharmaceutical industry, is subject
to quality variations imposed by source and manufactur-
ing process. The type of hydrate formed due to different
manufacturing conditions dictates many solid-state –
related properties, which have momentous contributions
to its lubrication performance. In this investigation, an
attempt was made to correlate the solid-state properties
of magnesium stearate with the lubrication efficacy and
to come up with the critical parameters, which affect its
properties as a lubricant. The variations in solid-state
properties (state of agglomeration, surface area, particle
size, crystal habit, hydration state, and crystal structure) of
magnesium stearate provide an appealing explanation, as
to how the critical balance between the three levels affects
its ultimate lubrication efficacy. It is this interplay be-
tween the three levels that decides the exact behavior of
the sample. With the use of an effective lubricant, it is
possible to decrease the concentration of hydrophobic
lubricants in the formulation, which would otherwise be
required in higher quantities, and therefore affect the
biopharmaceutical performance of the dosage form,
especially for BCS class II and IV drugs.
Our study revealed that all the levels of solid state
contribute to, but to a different extent individually, the
performance of the lubricant. Particle-level characteristics
seemed to be the most important and overshadow the
molecular-level characteristics in the experimental con-
ditions. Moreover, total work of compression can be used
to screen the lubricants for their efficacy and texture
analyser is a sensitive and suitable instrument to evaluate
lubrication efficacy by this method.
Such studies can be incalculably useful in scale-up of
both tablets and capsules, where discrepancy at any of the
molecular, particle, or bulk level may multiply manyfold,
making the processing unpredictable and affect the
performance of the dosage form.
ACKNOWLEDGMENTS
We are grateful to Prof. R. K. Ray, Head, Advanced
Center for Material Sciences (ACMS), IIT Kanpur, for his
kind permission to work in ACMS.
REFERENCES
1. Vitkova, M.; Chalabala, M. The use of some
hydrophobic substances in tablet technology. Acta
Pharm. Hung. 1998, 68 (6), 336– 344.
2. Marshall, K. Compression and consolidation of
powdered solids. In The Theory and Practice of
Industrial Pharmacy; Lachman, L., Liberman, A.,
Kanig, J.L., Eds.; Varghese: Bombay, India, 1976;
66– 99.
3. Billany, M.R.; Richards, J.H. Batch variation of
magnesium stearate and its effect on the dissolution
rate of salicylic acid from solid dosage forms. Drug
Dev. Ind. Pharm. 1982, 8 (4), 497– 511.
4. Banker, G.S. Tablets. In The Theory and Practice of
Industrial Pharmacy; Lachman, L., Liberman, A.,
Kanig, J.L., Eds.; Varghese: Bombay, India, 1976;
293– 345.
5. Jones, B.E. The filling of powders into two-piece
hard capsules. Int. J. Pharm. 2001, 227 (1), 5 –26.
6. Baichwal, A.R.; Augsburger, L.L. Development
 Solid-State Properties and Magnesium Stearate
validation of modified annular shear cell (MASC) to
study frictional properties of lubricants. Int. J.
Pharm. 1985, 26 (1), 191 –196.
7. Butcher, A.E.; Jones, T.M. Some physical character-
istics of magnesium stearate. J. Pharm. Pharmacol.
1972, 24 (Suppl.), 1P –9P.
8. Bolhuis, G.K.; Lerk, C.F.; Broersma, P. Mixing
action and evaluation of tablet lubricants in direct
compression. Drug Dev. Ind. Pharm. 1980, 6, 15 –33.
Pharmaceutical Development and Technology Downloaded from informahealthcare.com by University of Guelph on 05/04/12
9. Dansereau, R.; Peck, G.E. Effect of variability in the
physical and chemical properties of magnesium
stearate on the properties of compressed tablets.
Drug Dev. Ind. Pharm. 1987, 13 (6), 975 – 999.
10. Miller, T.A.; York, P. Physical and chemical
characterization of some high pure magnesium
stearate and palmitate powders. Int. J. Pharm.
1985, 23 (1), 55 –67.
11. Wada, Y.; Matsubara, T. Pseudopolymorphism and
lubricating properties of magnesium stearate. Pow-
der Technol. 1994, 78 (2), 109 – 114.
12. Buckley, D.H.; Johnson, R.L. Lubrication with
solids. Chem. Technol. 1972, 2, 302 –310.
13. Barra, J.; Somma, R. Influence of physicochemical
For personal use only.
variability of magnesium stearate on its lubricant
properties. Drug Dev. Ind. Pharm. 1996, 22 (11),
1105– 1120.
14. Leinonen, U.I.; Vihervaara, P.A.; Laine, E.S.U.
Physical and lubrication properties of magnesium
stearate. J. Pharm. Sci. 1992, 81 (12), 1194 –1198.
15. Ertel, K.D.; Carstensen, J.T. Chemical, physical,
437
and lubricant properties of magnesium stearate.
J. Pharm. Sci. 1988, 77 (7), 625– 629.
16. Frattini, C.; Simioni, L. Should magnesium stearate
be assessed in the formulation of solid dosage forms
by weight or by surface area? Drug Dev. Ind. Pharm.
1984, 10 (7), 1117 –1130.
17. Bracconi, C.; Andres, C.; Ndiaye, A. Structural
properties of magnesium stearate pseudopoly-
morphs: effect of temperature. Int. J. Pharm. 2003,
262 (1), 109 –124.
18. Miller, T.A.; York, P. Pharmaceutical tablet lubri-
cation. Int. J. Pharm. 1988, 41 (1), 1– 19.
19. Ertel, K.D.; Carstensen, J.T. An examination of
physical properties of pure magnesium stearate. Int.
J. Pharm. 1988, 42 (1), 171 – 180.
20. Chapman, D. The 720 cm1 band in the infrared
spectra of crystalline long chain compounds. J.
Chem. Soc. 1957, 4489 – 4491.
21. Muller, B.W. Die pseudopolymorphie von magne-
siumsalzen hoherer fettsauren. Arch. Pharm. 1977,
310, 693 –704.
22. Rubino, J.T. Electrostatic and non-electrostatic free
energy contributions to acid dissociation constants
in cosolvent-water mixtures. Int. J. Pharm. 1988, 42
(1), 181– 191.
23. Shah, A.C.; Mlodozeniec, M.L. Mechanism of
surface lubrication: influence of duration of lubri-
cant-excipient mixing on processing characteristics
of powders and properties of compressed tablets.
J. Pharm. Sci. 1977, 66, 1377 –1378.