вход по аккаунту


Effects of ultrasound in coating nano-precipitated CaCO3 with stearic acid.

код для вставкиСкачать
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
Published online 20 July 2009 in Wiley InterScience
( DOI:10.1002/apj.342
Special Theme Research Article
Effects of ultrasound in coating nano-precipitated CaCO3
with stearic acid
K. W. Kow* , E. C. Abdullah and A. R. Aziz
Department of Chemical Engineering, University of Malaya, 50603, Kuala Lumpur, Malaysia
Received 22 October 2008; Revised 20 March 2009; Accepted 20 March 2009
ABSTRACT: Nano-Precipitated CaCO3 (NPCC) are coated with stearic acid to improve its dispersion in polymer as
well as to reduce agglomeration. In this work, coating was done by wet method using ethanol. Ultrasonication was
applied to NPCC to de-agglomerate micron size NPCC into smaller size. Effects of amplitude, temperature and energy
input of ultrasonication was investigated. The amplitude was varied from 60% to 100% whereas temperature was varied
from 5 ◦ C to 45 ◦ C. The energy input was stressed up from 3.6 kJ to 180 kJ. Mean diameter of NPCC were observed by
using Particle Size Analyzer and Transmission Electron Microscopy (TEM). It was found that mean diameter of NPCC
do not vary significantly with temperature. Mean diameter of NPCC, however, decreases exponentially with the energy
input. Comparisons were done on NPCC coated with others methods such as dry ball milling and aqueous coating.
TEM images show that coating NPCC with ultrasonication is more uniform than other methods. In addition, first
derivative mass loss with temperature (DTG) reveals that NPCC coated with ultrasonication do not contain excessive
free acid as exhibited by those coated in aqueous and dry ball milling.  2009 Curtin University of Technology and
John Wiley & Sons, Ltd.
KEYWORDS: coating NPCC; stearic acid; agglomeration; ultrasonication
Precipitated calcium carbonate (PCC) is commonly used
as a polymer filler to reduce material cost. Nano size
precipitated calcium carbonate (NPCC) has recently
emerged as a functional filler to enhance the properties
of polymers. According to Chen et al .,[1] notched
impact strength of PVC added with 15% of nano
calcium carbonate reached a maximum value of 39 J/M,
i.e. 30% higher than neat polymer.
NPCC tends to agglomerate due to strong cohesive
forces. Micron size agglomerates formed may initiate cracks in the polymer, thus acting as degradants
rather than reinforcing the polymer. To overcome this
problem, surface modification is applied where the surface of NPCC is coated with a layer of stearic acid.
The hydrophobic layer of stearic acid prevents NPCC
from agglomeration by lowering its surface energy.
Puka’nszky and Karger[2] have proved that calcium
carbonate coated with stearic acid reduced its surface
tension from 210 mJ/m2 to 40–50 mJ/m2 .
The coating of NPCC can be carried out in dry or
wet conditions. In the dry method, NPCC and stearic
*Correspondence to: K. W. Kow, Department of Chemical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala
Lumpur, Malaysia. E-mail:
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
acid are ball-milled whereas wet method requires the
presence of liquid. Most wet coatings use water due to
low cost and ease in handling. Stearic acid is insoluble
in water and thus the coating is precipitation controlled
rather than adsorption controlled. This leads to inferior
quality of coating such as multilayer coating, formation
of free acid, and irregularities in shape.
NPCC in this work was coated in ethanol to obtain
monolayer adsorption of stearic acid. Stearic acid forms
a perfect match with NPCC as its adsorption sites are
spaced at just the size of the stearic acid molecule.[3]
As a result, molecules of stearic acid are tightly packed
with one another, thus forming a stable layer that resists
penetration of small molecules.
The theoretical mass of stearic acid required to
coat 1 g of NPCC (mc ) can be calculated by using
Eqn (1) where S is maximum BET surface area of
the NPCC (approximately 40 m2 /g), NA is Avogadro
constant (6.02 × 1023 mol−1 ), Aa is the surface area
occupied by one stearic acid molecule in the vertical
direction (0.205 nm2 ) and MR is molecular weight of
stearic acid (284.48 g/mol). The theoretical mass of
stearic acid needed is 0.09 g per gram of NPCC.
mc =
Asia-Pacific Journal of Chemical Engineering
As NPCCs are agglomerated prior to the coating
process, with surface area less than 40 m2 /g, stearic acid
is actually coated on micron size agglomerates but not
nanoparticles. The surface area will increase to 40 m2 /g
only if sufficient energy is supplied to de-agglomerate
the uncoated NPCC. In other words, stearic acid used
in the coating is always in excess. There is very limited
information published regarding the determination of
optimum amount of stearic acid to provide monolayer
coating. Coated NPCCs are normally contaminated with
free stearic acid and multilayer coating. To overcome
this, ultrasonication may be use to de-agglomerate
NPCC into nano size which will increase the surface
area of NPCC prior to the coating process. Though deagglomeration may be accomplished by other method
such as ball milling, the high impact force of milling
causes same problems as in aqueous coating. Ultrasonication with more uniform energy distribution is able to
eliminate those problems. Other researchers had studied
the effect of amplitude in ultrasonic de-agglomeration
by varying amplitude at constant ultrasonic duration.[4,5]
Results give x α E −b where x is particle size, E is
energy input, and b is a constant. However, they varied both amplitude and energy input simultaneously.
Due to this energy input also increased when higher
amplitude was applied for the same duration. Thus, deagglomeration may be affected by either one of these
variables, or even both at the same time. In this work,
the effects of amplitude and energy input were studied
separately. For the same amplitude, energy input was
varied by adjusting the duration of ultrasonication (t)
which can be estimated by using Eqn (2)[6]
2π f · ξmax · A
where E is energy input (J), ρ is density of ethanol
(kg·m−3 ), c is the speed of ultrasound travelled in
ethanol (m·s−1 ), a is effective area of sonotrode (m2 ),
f is the ultrasound frequency (24 kHz), ξmax is the
maximum amplitude (100 µm), and A is the percentage
of maximum amplitude applied. This shows there may
be optimum amplitude that may de-agglomerate NPCC
to a targeted size within the shortest duration.
Ultrasonication released enormous heat during implosion of cavities and therefore cooling is required to prevent ethanol from boiling due to high temperature. Studies on the effect of temperature on de-agglomeration are
very limited[7] . The possibility of applying ultrasonication at room temperature or even higher, but below
boiling point of ethanol, was explored in this work.
Besides that, coating process in this work, using ethanol,
is expected to remove excess free stearic acid as stearic
acid is soluble in ethanol.
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Materials and equipments
NPCC-111 was obtained from NanoMaterials Technology Pte. Ltd., Singapore. The maximum BET surface
area was 40 m2 /g and the size ranged from 40–100 nm.
It is in the form of cubic calcite crystals. The stearic acid
used was a blend of stearic acid and palmitic acid of
98% purity.
Ultrasonic processor (Hielscher, UP 400S) with
22 mm diameter sonotrode was used in this experiment. The amplitude of the equipment can be adjusted
from 20 to 100% of the maximum amplitude (100 µm)
with a corresponding intensity of 85 W/cm. Samples
were ultrasonicated in a flow vessel equipped with
a cooling jacket to control the temperature in the
vessel. The energy input was indicated by a power
Coating process of NPCC
Uncoated NPCC (2 g) was first ultrasonicated in ethanol
(30 ml) at various amplitudes (60–100%) and energy
inputs (3.6–180 kJ) under constant temperature to form
suspension. The corresponding ultrasonication time was
recorded. Fresh ethanol was used to dissolve 9 wt%
of stearic acid in a mixing vessel. Ethanol dissolved
with stearic acid was then mixed with ultrasonicated
suspension. Experiments were repeated for temperatures ranges of 5–45 ◦ C. Suspension of coated NPCC
obtained was then centrifuged and dried. NPCC was
also coated with other methods including dry ball
milling (Fristch, Pulverisette 6), aqueous coating, and
direct coating without de-agglomeration for comparison.
Samples characterization
Particle size distributions of uncoated and coated particles were analyzed using Zetasizer (Malvern, ZS90).
Z -average mean diameter from zetasizer was used to
compare relative change in particle size. BET surface area of samples was measured (Micromeritics,
ASAP 2010 Chemisorption) to compare the degree of
agglomeration. TEM images (Zeiss, Libra 120) were
taken to observe the shape of coated particles. Thermogravimetry analysis (Mettler Toledo, TGA/SDTA 851e)
was carried out to detect, if any, the formation of
free acid and multilayer adsorption. In TGA, samples were heated up continuously in air atmosphere
from room temperature to 1000 ◦ C at a heating rate of
15 ◦ C/ min.
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
Shape of NPCC
Figure 1(a) shows uncoated NPCC which is in cubic
shape. In Fig. 1(b), NPCC under the coated layer
remains cubic but the uneven coated layer causes
particles change to irregular shape. This is attributed
to the rapid precipitation of stearic acid from aqueous
form during coating process. In Fig. 1(c), high impact
force of dry ball milling causes not only the coated layer
but also the shape of NPCC inside exhibit high degree
of irregularities. For ultrasonication in ethanol, coated
NPCC remained in cubic shape and a uniform layer of
coating was observed in Fig. 1(d). Multilayer adsorption
of stearic acid on ultrasonicated NPCC is suspected
since the coated layer is relatively thick compared to
the molecular length of stearic acid.
Effects of ultrasonication to particle size
In Fig. 2 and 3, mean diameter of NPCC does not
vary significantly with amplitude and temperature but
rather energy input. The mean size of NPCC decreased
nonlinearly when higher energy input was applied.
Combining results of other researchers,[4,5,7] Model 1
was proposed in this work,
ln x = b0 + b1 · ln E + b2 · ln A + b3 · T
(Model 1)
where x is the mean size of NPCC (nm), E is ultrasound
energy input (kJ), A is ultrasound amplitude (%), T
is temperature of ultrasonication ( ◦ C) and b0 , b1 , b2
and b3 are regression coefficients. Experimental data
were fitted into Model 1 and tested using ANOVA
method and the result is shown in Table 1. A power
relation was obtained where x α E −0.117 . This agreed
well with Sauter’s findings where the power was in
the order of −0.1.[5] This also indicates that particles
size is strongly dependent on energy input but not on
ultrasound amplitude. Similar results had been reported
by Vinay and Ali [7] where particles breakage was
uncorrelated with amplitude, but rather energy input. As
higher amplitude utilized higher power input, shorter
duration is required to transfer the same amount of
energy. Therefore, de-agglomeration should be carried
Figure 1. TEM images of (a) uncoated NPCC, (b) NPCC coated in aqueous, (c)
NPCC coated with dry ball milling, and (d) NPCC coated with ultrasonication.
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
Figure 3. Effects of ultrasound amplitude and temperature
on mean size of NPCC. This figure is available in colour
online at
Figure 2. Effects of ultrasound amplitude and energy input
on mean size of NPCC. This figure is available in colour
online at
out with maximum amplitude to reduce the processing
Though temperature has no effect on particle size,
it affects the duration to transfer energy as in Fig. 4.
Based on Fig. 4 and Eqn (2), Model 2 was proposed to
regress the data of ultrasonication duration:
ln t = c0 + c1 · ln E + c2 · ln A + c3 · T
BET surface area
(Model 2)
where t is ultrasonication duration (s) and c0 , c1 , c2
and c3 are regression coefficients. Table 2 shows that
all the three parameters have significant effects on
ultrasonication duration. The power of E i.e. in the order
approximate to 1 agrees with the linear relationship
as shown in Eqn (2). Unlike Eqn (2), ultrasonication
duration (t) is in fact inversely proportional to A rather
than to the square (A2 ).
Similar results were reported by Vinay and Ali,[7]
and Kuijpers, Kemmere, and Keurentjes[8] . The results
obtained in this work once again supported that it is
more realistic to take ultrasonication duration to be
inversely proportional to amplitude.
Energy input, E (kJ)
Figure 4. Effect of temperature on ultrasonic duration
(A = 100%). This figure is available in colour online at
By comparing the results fitted into Model 2 and
Eqn (2), a semiempirical relation (3) is obtained.
E =γ
ρ cω2 a ξmax
· A · t · e−0.0078
Table 1. ANOVA result of regression based on Model 1.
Adjusted R 2
ln E
ln A
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
Table 2. ANOVA result of regression based on Model 2.
ln E
ln A
where is coefficient with the confidence interval of
2091.05 to 2100.26 at 95% level of confidence within
the temperature range of 5–45 ◦ C and V is the volume
of ethanol used in ultrasonication (cm3 ). Based on
Eqn (4), to achieve a certain particle size, ultrasonic
de-agglomeration should be carried out at a maximum
amplitude and low temperature to reduce the processing
Adjusted R 2
Duration of ultrasonication (s)
where γ is the coefficient with the confidence interval of 4.18 × 10−12 to 4.34 × 10−12 at 95% level of
confidence within the temperature range of 5–45 ◦ C.
Confidence interval of γ is used because the density
of ethanol (ρ) and the speed of ultrasound travelled in
ethanol (c) changed appreciably from 1209.9 kg/m3 to
1069.7 kg/m3 and 802 m/s to 768 m/s respectively as
temperature increased from 5 to 45 ◦ C.
Mean particle size is a strong function of energy input
and the energy input in this work was only manipulated
directly by using a power meter. If the ultrasonic
energy input is controlled explicitly by parameters such
as temperature, amplitude, and ultrasonication duration
as in Eqn (3), a targeted mean particle size becomes
dependent on such parameters. This is illustrated in
the semiempirical relation (4) obtained by combining
Eqn (3) and the results fitted into Model 1:
ρcω2 a t ξmax −0.0078 T
· A−0.0352 (4)
x =·
3.6 kJ
18 kJ
36 kJ
108 kJ
180 kJ
Temperature (°C)
Figure 5. BET surface areas of coated and uncoated
NPCC. This figure is available in colour online at
is more energy efficient than ultrasonication in deagglomerating NPCC. One of the explanations is ultrasonication converted relatively high amount of energy
input into heat during implosion and thus less energy
was transferred to particles. Compared with ultrasonication, temperature during dry ball milling was only
slightly increased, i.e. less energy was transformed into
heat and particles are de-agglomerated efficiently. If
particle shape is not the main consideration, NPCC
coated with dry ball milling is a better method than
ultrasonication. BET surface area of all NPCC in Fig. 5
increased with increase in energy input, i.e. agreed with
the result in Fig. 2, which again shows that energy input
is the major factor in controlling particle size.
Effect of coating
Thermagravimetry analysis (TGA)
Figure 5 shows the BET surface area of NPCC. Measurements were carried out 20 days after ultrasonication
or dry milling. It can be seen that uncoated NPCC
experienced higher degree of reagglomeration compared with coated NPCC. By comparing coated NPCC
with uncoated NPCC that ultrasonicated with the same
energy input, surface areas of coated NPCC are 4–17%
higher than uncoated NPCC. Based on this, stearic acid
coating at least slowed down, if not completely desisted,
the reagglomeration of NPCC.
NPCC coated with dry ball milling gives much
higher (12–39%) BET surface area than those coated
with ultrasonication. This shows that dry ball milling
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
According to other researchers,[9,10] peak for monolayer
coating was observed ca 310 ◦ C in Derivative Thermogravimetry (DTG), whereas free acid and local bilayer
show peaks at ca 270 and 200 ◦ C respectively. Since all
stearic acid decomposed at temperature below 350 ◦ C,
the mass of stearic acid that coated on NPCC is given by
the corresponding weight loss in TGA (Table 3). Since
all weight loss is less than stearic acid that dissolved (9
wt%), none of the samples actually reaches the maximum surface area of 40 m2 /g. Figure 6 shows that all
samples give the highest peak within 320–340 ◦ C, i.e.
peak corresponding to the monolayer coverage.
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
DOI: 10.1002/apj
bilayer Free
Abs. Derivatives, |dm/dT|
100 150 200 250 300 350 400 450 500
Temperature (°C)
Figure 6. DTG traces of NPCC coated
with various methods; (A) coated
without de-agglomeration, (B) coated
with ultrasonication (180 kJ 100%
amplitude), (C) coated with dry milling
(3.6 kJ) and (D) coated in aqueous.
This figure is available in colour online
Table 3. Weight loss of coated NPCC in TGA up to
350 ◦ C.
NPCC samples
Coated without
Coated with dry
milling (3.6 kJ)
Coated in aqueous
Coated with
(180 kJ, 100%
Weight loss up to 350 ◦ C (%)
For NPCC that coated directly without de-agglomeration, no free acid or local bilayer was observed. This
is due to excessive stearic acid dissolved in ethanol
and no free acid precipitated out as free acid. The
corresponding weight loss was only 3.54% (Table 3).
As no de-agglomeration was applied, stearic acid was
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
Asia-Pacific Journal of Chemical Engineering
actually coated on large NPCC agglomerates. The
surface area of these agglomerates is very low and only
required a small amount of stearic acid in the coating.
The residue was still dissolved in ethanol.
For NPCC coated in aqueous (trace D), a large
amount of free acid was observed and also local bilayer
presented at peak ca 220 ◦ C. The existence of both
free acid and local bilayer resulted from the rapid
precipitation of stearic acid in water. The detected local
bilayer generally agreed with the thick coated layer
observed under TEM in Fig. 1(b).
DTG trace for dry milling is very different from
others where no local bilayer is observed in the sample.
This is contradicting with TEM images in Fig. 1(c),
which showd a thick and irregular coated layer. The
phenomenon is attributed to the mechanism of dry
milling where stearic acid molecules are not adsorbed
on NPCC surface as in ultrasonication. Stearic acid
molecules were actually ‘fused’ on the first layer rather
than adsorbed physically on it. A higher amount of
energy is required to decompose such a ‘fused’ layer.
Stearic molecules are ‘fused’ to form a firm layer due
to high impact force by ball milling as well as the high
temperature caused by friction in milling process.
For NPCC coated with ultrasonication, there was
no free acid observed. Since ethanol was used in
ultrasonication, excessive stearic acid remains dissolved
in ethanol without precipitating out from the solvent.
This proves that coating in ethanol can prevent the
sample from contamination by free acid. A similar
result cannot be achieved by dry milling or aqueous
coating. Local bilayer is still detected in sample at
peak ca 250 ◦ C. This is agreed with the TEM images
in Fig. 1(d) where the coated layer is thicker than
monolayer adsorption. The existence of local bilayer
shows overcoating of ultrasonicated NPCC as a result
of multilayer adsorption. Based on the BET surface area
for the sample (24.1022 m2 /g), the mass required for
monolayer coverage was estimated by using Eqn (1),
which is 5.56%. However, the weight loss of the sample
is slightly higher (5.62%) than the estimated amount.
This again confirmed the existence of local bilayer in
the sample. Though NPCCs coated with ultrasonication
contain local bilayer, the amount of excess stearic acid
is still relatively small compared with others coated with
dry milling or aqueous form.
For both coated with ultrasonication and coated
without de-agglomeration, peaks are observed at ca
150 ◦ C. The corresponding peak was not observed for
NPCC coated in aqueous and dry milling. Since both
NPCC coated with ultrasonication and without deagglomeration use ethanol as solvent, it is suspected
that ethanol is the source of these peaks. The peaks may
represent the following; (1) ethanol, (2) ester formed in
by ethanol and stearic acid, and (3) dimer of stearic
acid. Ethanol is the least possible cause since the sample
is dried in an oven before any analysis. Even if the
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering
drying is insufficient to remove ethanol from the sample
completely, ethanol will desorb from the surface at low
temperature due to its low boiling point (∼79 ◦ C). Ester
ethyl stearate may form from the reaction of alcohol and
carboxylic acid. However, such esterification requires
high temperature and catalyst. The stearic acid and
ethanol were only mixed at a temperature below 50 ◦ C
and in the absence of any catalyst, esterification is
unlikely to occur. The boiling point of ethyl stearate
(∼215 ◦ C) is also much higher than the detected peak.
It is well known that carboxylic such as acetic acid will
form a dimer. It is possible for stearic acid to form a
dimer in the sample after all the ethanol is evaporated.
However, dimerization of carboxylic acid is unlikely to
occur in polar solvent like ethanol. Literature related to
this topic is still very limited and no proper conclusion
can be drawn from other researches. Further studies
must be carried out to identify another possible source
of contamination indicated by TGA.
Results from this work proved that stearic acid coating successfully reduced agglomeration of NPCC. Both
ultrasonication and dry milling offered different advantages and drawbacks in de-agglomerating nanoparticles.
Dry milling is more energy efficient but the particle
shape exhibits high degrees of irregularities. This does
not favour certain mechanisms, especially in toughening
polymers. In case regular shape is desired, ultrasonication offers a better means of de-agglomeration where
stearic acid can be coated evenly on the surface of
NPCC and can preserve the particle shape. Multilayer
coating on NPCC coated with ultrasonication can be
reduced if the optimum amount of stearic acid needed
can be determined. Liquid medium used in ultrasonication also affected the quality of coating. Since coating is precipitation controlled in aqueous form, stearic
acid layer coated is highly nonuniform. It also poses
another serious problem where the coated NPCCs contain high amount of excessive coating agent in the
form of free acid. For NPCC coated using ethanol, a
 2009 Curtin University of Technology and John Wiley & Sons, Ltd.
small amount of byproduct was detected in TGA. The
formation of such a byproduct is possibly caused by
reactions among raw materials used. Further investigations should be carried out to identify the byproduct. Studies on the effects of energy input, amplitude,
and temperature on the particle size of ultrasonicated
NPCC revealed that particle size is a strong function
of energy input as reported by Sauter.[4,5] Particle size
only increased slightly with ultrasound amplitude. Temperature has no effect on the particle size as long as the
same amount of energy input is used in ultrasonication.
When energy input is not controlled by energy measuring device, energy input is heavily dependent on factors
such as amplitude, ultrasonication duration, temperature
as well as properties of medium where ultrasound travelled. The study also revealed that the ultrasound energy
input is proportional to amplitude but not to its square
as stated in literatures.[6] Based on the semiempirical
relation obtained, one should carry out the ultrasonication at maximum amplitude and lower temperature to
reduce processing time. Though lowering the temperature reduced the processing time, utility cost involved
for such cooling requirement should also be considered.
A balance must be achieved between the ultrasonication
duration and temperature to optimize cost.
[1] N. Chen, C. Wan, Y. Zhang, Y.X. Zhang. Polym. Test., 2004;
23(2), 169–174.
[2] B. Puka’nszky, J. Karger-Kocsis. Polypropylene: Structure,
Blends and Composites, Chapman & Hall: London, 1995; p.46.
[3] R.N. Rothon, Rapra Technology, 2001; 157.
[4] C. Sauter, M.A. Emin, H.P. Schuchmann, S. Tavman. Ultrason. Sonochem., 2008; 15(4), 517–523.
[5] C. Sauter, M. Pohl, H.P. Schuchmann, Ultrasound for dispersing nanoparticles. In Proceedings of the 12th European Conference on Mixing, Bologna, 2006; pp.103–110.
[6] J.R. Frederick. Ultrasonic Engineering, John Wiley: New
York, 1965; p.29.
[7] R. Vinay, A. Ali. Ultrason. Sonochem., 2008; 15(1), 55–64.
[8] M. Kuijpers, M. Kemmere, J. Keurentjes. Ultrasonics, 2002;
40, 675–678.
[9] A.O. Maged, A. Ayman, W.S. Ulrich. Polymer, 2004; 45(4),
[10] K.S. Jaw, C.K. Hsu, J.S. Lee. Thermochim. Acta, 2001;
367–368, 165–168.
Asia-Pac. J. Chem. Eng. 2009; 4: 807–813
DOI: 10.1002/apj
Без категории
Размер файла
409 Кб
acid, effect, coatings, nano, stearic, cacoo, precipitated, ultrasound
Пожаловаться на содержимое документа