Wettability Assessment of Submicrometer Alumina Powder Using by liaoqinmei


									                                                                                                                              J. Am. Ceram. Soc., 93 [9] 2515–2518 (2010)
                                                                                                                                   DOI: 10.1111/j.1551-2916.2010.03854.x
                                                                                                                                  r 2010 The American Ceramic Society

                             Wettability Assessment of Submicrometer Alumina Powder Using a
                                                                 Modified Washburn Method
                                                Bram Neirinck, Johannes van Deursen, Omer Van der Biest,* and Jef Vleugelsw
                                          Department of Metallurgy and Materials Engineering (MTM), K. U. Leuven, B-3001 Heverlee, Belgium

The possibility to accurately measure the liquid–solid contact                                   on capillary wetting of powder packings. Among these tech-
angle on submicrometer powders by means of a modified cap-                                        niques, the Washburn or capillary rise method is the most widely
illary rise setup using cold isostatically pressed powder compacts                               used.8 Yet, it has been demonstrated that the values obtained
is reported. The geometric factor needed for the determination                                   through these measurements strongly depend on system param-
of the contact angle from liquid penetration data was calculated                                 eters such as particle size and morphology.12
from the pore properties of the powder compacts. During the                                         The reproducibility of contact angles measured via capillary
capillary rise experiments, the pellets retained their structural                                rise depends on the repeatability of the powder packing used
integrity and concomitant porosity ensuring the reproducibility                                  during these measurements.13,14 This is typically achieved by
of the contact angle values obtained.                                                            preparing columns of tapped or stamped powder according to
                                                                                                 standards such as ASTM B 527 and DIN ISO 787. Yet, even
                                                                                                 when a reproducible and sufficiently dense packing is obtained,
                                  I. Introduction                                                the results can vary slightly with column diameter due to edge
                                                                                                 effects.14 Nevertheless, most of the efforts on particle–liquid in-
T    HE interaction between solids and solvent can determine
     whether particles are stable in suspension or flocculate. Fur-
thermore, this parameter determines the interaction of particles
                                                                                                 teraction quantification thus far have been directed toward par-
                                                                                                 ticles with sizes above 1 mm, while colloidal processing is
                                                                                                 evolving toward submicrometer and nanoparticles. Such small
with the fluid–fluid interfaces generated in particle-stabilized                                   particles exhibit an inherent lower packing density when stacked
foams and emulsions, which are currently being explored for the                                  in the tubes used for capillary rise experiments, resulting in a
production of complex porous structures.1–7 Nonetheless, this                                    lower structural integrity of the packing. Combined with the
parameter is often overlooked when studying suspensions in                                       large capillary forces exerted on the powder during measure-
colloidal processing routes. This is partly not only because of the                              ment, this can cause geometrical changes in the packing during
underestimated importance of these interactions in suspensions                                   the experiment, yielding inaccurate and irreproducible results.
but also because the wettability assessment of finely divided sol-                                Hence, stronger powder compacts with reproducible properties
ids is not as straightforward as the direct measurement of con-                                  are needed.
tact angles on dense flat substrates.8                                                               Furthermore, for the calculation of the contact angle from
   Direct optical measurements on the particle surface is only                                   liquid penetration data, a geometrical factor that represents the
feasible on large, and because of the effect of surface texture,                                 pore structure in the powder compact is needed. This value is
smooth particles.8 However, such particles are an exception in                                   typically obtained by performing a measurement using a refer-
commercial applications. When visually measuring a contact                                       ence liquid, which is assumed to be perfectly wetting (contact
angle on a powder pellet, a portion of the liquid penetrates the                                 angle 5 01). The contact angles of other tested solvents are thus
porous powder compact, causing a systematic error. Hence, in-                                    calculated relative to the wetting of the reference and therefore
stead of direct optical measurements, alternative methods have                                   can only be used for a qualitative assessment of the wetting
been developed for determining particle–solvent interactions.                                    behavior.8
Therefore, other techniques for the determination of the wetting                                    In this work, we suggest the use of reproducibly precompact-
behavior have been developed. However, most of those tech-                                       ed powder samples with a force larger than the maximal capil-
niques either require spherical monomodal particles or are of a                                  lary force expected during measurement in order to prevent
qualitative nature. When the particles are sufficiently large, i.e.                               structural changes during capillary rise experiments. Character-
 ! 10 mm and spherical, either the interaction force can be mea-                                 ization of the porous properties by means of commercially avail-
sured directly using an AFM or the curvature around a particle                                   able methods is used to determine the geometrical factor,
floating on the liquid surface can be quantified using interfe-                                    independent of a reference liquid.
rometry or confocal microscopy.9,10 For ordered layers of
smaller spherical monomodal particles, the contact angle can
be determined by measuring the degree of submersion of the
                                                                                                                  II. Experimental Procedure
particles at a liquid interface by ellipsometry.11 For smaller or
irregularly shaped particles, only indirect measuring techniques                                 All measurements were carried out using pellets obtained from a
can deliver useable results. A number of these methods are based                                 high-purity a-Al2O3 powder (SM8, Baikowski, Annecy, France)
                                                                                                 with a particle size of approximately 390 nm, as measured by
   A. Krell—contributing editor                                                                  light scattering (Mastersizer micro1, Malvern, Worcestershire,
                                                                                                 U.K.). For each pellet, approximately 10 g of powder was uni-
                                                                                                 axially pressed in a 15-mm-diameter graphite die using 5 kN of
                                                                                                 force. The pellets obtained were consequently treated using a
    Manuscript No. 27507. Received February 3, 2010; approved April 8, 2010.                     cold isostatic press (CIP) (Engineered Pressure Systems Inter-
    This work was supported by the Flemish Institute for the Promotion of Scientific and
Technological Research in Industry (IWT) under grant SB/53092 and the Research Fund              national, Temse, Belgium) at 200 MPa.
K. U. Leuven under project GOA/08/007.                                                              The pore size distribution of the pellets obtained was char-
    *Member, The American Ceramics Society.
     Author to whom correspondence should be addressed. e-mail: jozef.vleugels@mtm.              acterized by mercury intrusion porosimetry (Autopore IV 9500,
kuleuven.be                                                                                      Micrometrics, Norcross, GA). The overall porosity was deter-
2516                                     Rapid Communications of the American Ceramic Society                                 Vol. 93, No. 9

Table I. Physical Properties of the Used Solvents Measured
                         at 201C
                         Density   Surface tension     Viscosity
Solvent                  (g/mL)     (10À3 N/m)       (10À3 Pa Á s)

Demineralized water      0.998     71.5070.53 1.019470.0064
Absolute ethanol         0.786     21.9770.17 1.201470.0004
Cyclohexane              0.779     26.2170.12 0.998270.0006
n-Heptane                0.683     19.3870.07 0.413870.0005
Paraffin oil              0.851     26.6270.53 31.920070.0740
Ethane diol              1.114     47.6770.08 21.19370.007

mined using the Archimedes method in ethanol using lacquer
encapsulation (Red lacquer, Enthone, Langenfeld, Germany).
   For the assessment of the proposed modified Washburn
method, five different liquids were selected: demineralized wa-
ter, freshly prepared using a commercial ion exchange setup
(Sation Aqualab 50, Elga, Madrid, Spain), ethanol (Pro analy-
sis, Prolabo, Fontenay sous Bois, France), cyclohexane (Pro
analysis, Prolabo), n-heptane (analytical reagent grade, Fisher
Scientific, Pittsburgh, PA), paraffin oil (Nujol mull spectro-
scopic grade, Acros, Geel, Belgium), and ethane diol (Pro anal-
ysis, Acros). Before the capillary rise measurements, the physical
properties of these solvent needed for the contact angle calcu-
lation are quantified. The surface tension was characterized by
means of the pendant drop method (CAM 200, KSV, Helsinki,
Finland), whereas the viscosity was determined using capillary
viscometers (Ubbelohde, Schott Instruments, Mainz, Germany).
During viscosity assessment, the viscometers and test liquids              Fig. 1. Detailed view of the three-point clamping mechanism used to
were kept at 201C using a thermostatic bath (RTE-300, Neslab               suspend the pellets above the liquid during measurement.
Instruments, Newington, NH). The density of the liquids
was measured with the Archimedes method using a calibrated                 where m is the weight of the penetrated solvent, rl the solvent
glass volume of 10 mL (Sartorius, Goettingen, Germany). The                density, e the powder packing, R the powder packing radius,
values obtained are summarized in Table I.                                 and h the height of the liquid front. Because it is not always
   The actual capillary rise experiments were carried out using a          possible to clearly distinguish the liquid front in powder pack-
digital analytical balance (BP 221S, Sartorius). For each exper-           ing, weight gain measurement as function of time is often pre-
iment, one pellet was suspended above a beaker in this balance             ferred over recording the height. Combining Eqs. (1) and (2)
using a homemade three-point clamping mechanism so that the                yields a modified Washburn Eq. (3) used for such experiments
flat sides of the pellet are parallel to the level plane. The setup
and detail of the clamping mechanism are shown in Fig. 1. Be-                  m2 r2 f gl cosy
fore measurement, the dimensions of the pellets were measured                     ¼ l                                                      (3)
                                                                                t      2Z
using a digital caliper. Liquid was slowly supplied to this beaker
using a peristaltic pump (323Du, Watson-Marlow, Wilmington,                  In this equation, f represents a geometric factor that depends
DE), causing a gradually rising liquid level. Upon contact be-             on the powder samples radius, porosity, and equivalent pore
tween liquid and pellet, the pump was stopped and the weight               radius:
increase of the pellet was recorded as a function of time using the
digital readout of the balance and digitally logged.                           f ¼ rc e2 ðpR2 Þ2                                           (4)

                                                                               Because measuring the pore radius and porosity on loose
                  III. Results and Discussion                              powder packing is not feasible, this factor is typically deter-
                                                                           mined by means of a reference liquid, such as octane, which is
The Washburn method is based on powder that is reproducibly                assumed to be perfectly wetting. As discussed in Section I, con-
packed in a cylindrical vessel with porous bottom through which            tact angles obtained from capillary rise measurements only hold
a wetting liquid can rise. The powder packing is typically ob-             value for qualitative wetting assessment. Comparing with the
tained by tapping the powder-filled cylinder. The original Wash-            literature data can lead to erroneous conclusions when a differ-
burn Eq. (1), which itself is derived from Poiseuille’s law for            ent reference is used. Because these parameters cannot be quan-
liquid flow through a capillary, describes the height of the liquid         tified, it is difficult to assess the reproducibility of the powder
front as a function time:                                                  packing method used to perform an experiment. As a result, the
                                                                           Washburn method is often considered to be inaccurate.8 Yet, it
     h2 rc gl cosy                                                         is still widely used as it is one of the few methods that can be
        ¼                                                            (1)
      t     2Z                                                             applied on a wide variety of powders and solvents at low cost
                                                                           per measurement.
with h the penetration height, t the elapsed time, rc the equiv-               On the other hand, measurement of both porosity and pore
alent pore radius, gl the solvent surface tension, y the contact           sizes has been used to characterize and assess the quality of
angle, and Z the dynamic viscosity of the solvent. If the packing          green ceramic objects. Furthermore, industrial applications
properties are known, the weight of the solvent in the pores can           show that compacts with a homogeneous and reproducible
be calculated from the height of the liquid front:                         green density can be obtained using CIP. Because the surface
                                                                           properties of a powder do not change during pressing, a com-
                                                                           pact produced using such means should be ideal for measuring
     m ¼ rl epR2 h                                                   (2)   the wetting behavior.
September 2010                           Rapid Communications of the American Ceramic Society                                           2517

                                                                             Table II. Experimentally Determined Contact Angles
                                                                                             Measured at 201C
                                                                         Solvent                                           Contact angle (deg.)

                                                                         Demineralized water                                 68.4170.32
                                                                         Absolute ethanol                                    64.2171.60
                                                                         Cyclohexane                                         62.1873.05
                                                                         n-Heptane                                           56.9471.12
                                                                         Paraffin oil                                         37.5970.41
                                                                         Ethane diol                                         57.5570.22

                                                                         aration of the precompacted samples, but sufficient to influence
                                                                         the properties of loosely packed powder samples.
Fig. 2. Pore size distribution of three cold isostatic pressed alumina      As described in the literature, the contact angle can be cal-
pellets, as measured by mercury porosimetry.                             culated from the slope of the linear part of the square weight
                                                                         versus time plots. For all liquids, except the highly viscous par-
   Because the value of capillary rise measurements depends on           affin oil, the linearity between 500 and 2000 s yielded a corre-
the reproducibility of the used powder packing, the porosity and         lation factor of at least 0.999, as shown in Fig. 3, which could be
pore size distribution of a number of CIP samples were mea-              used for the contact angle calculation. For paraffin oil, the val-
sured using the Archimedes method and respectively mercury               ues measured between 2000 and 10 000 s were used. The exper-
intrusion porosimetry. The overall porosity of the isostatically         imentally determined contact angles are summarized in Table II.
pressed Al2O3 compacts measured on three samples was                     The contact angle measured for water on the specific as-received
42.10%70.03%. Mercury intrusion porosimetry measurements                 Al2O3 powder intuitively appears to be too large and should not
on four samples show that the pore size distribution in these            be taken as a general value for alumina powder. However, due
samples is equally reproducible, as shown in Fig. 2. When the            to the high surface tension of water, the surface energy of the
average pore size is calculated as a function of volume, an av-          system increases when wetting with this solvent. Hence, rela-
erage value of 69.171.2 nm is obtained. Although the volume              tively large contact angles are obtained even on ideally flat sur-
average is 4 nm higher than the peak value found in the pore size        faces even though the material is generally considered to be
distribution curves (Fig. 2), this value is used during the remain-      hydrophilic.15 Experiments with ethane diol confirm that a hy-
der of this work because the mass increase measured during the           drophilic solvent with a lower surface energy yields lower con-
capillary rise experiments is directly related to the volume of          tact angles (Table II).
liquid penetrating the pores. The diameter of the samples after             The reproducibility of the results decreased with the increas-
CIP varied from 14.28 to 14.37 mm. Because of the wider dis-             ing volatility of the solvent. The evaporation of the solvent
tribution of the sample diameter, this value was measured for            causes a gradual drop of the liquid level in the measuring vessel.
each new experiment.                                                     For volatile solvents, this drop is fast enough to cause a devi-
   Similarly, in order to assess the reproducibility of the capil-       ation of the weight2 versus time curve from the linear case, in-
lary rise itself, three runs were performed for each solvent, as         ducing a measurement error. As the measurement is started at
shown in Fig. 3 for water. The results revealed an exceptional           the moment of initial contact between liquid and powder com-
reproducibility for water and mineral oil, indicating that not           pact, a high evaporation rate can even lead to a premature end
only the powder packing properties are indeed highly reproduc-           of the experiment as the contact between powder and fluid is
ible but also that they remain stable during the measurement. In         lost. Both effects result, despite performing the experiments in a
other words, the capillary pressure, Pc, exerted on the pores in         closed and vapor-saturated environment, in a wider scatter of
the compact during capillary rise measurements on precompact-            the calculated contact angle (Table II). Nonetheless, the scatter
ed powders is not strong enough to induce a change in their              on the results reported in the literature obtained through tradi-
structural properties                                                    tional capillary rise measurements on micrometer-sized powders
                                                                         is typically of the same level or even larger.16
           2glv cosy                                                        The results show that the powder pellets are poorly wetted by
    Pc ¼                                                          (5)    the short-chain organic solvents that due to their low surface
                                                                         energy, are frequently used as reference fluids in traditional
   The capillary pressure exerted by water in the used powder            Washburn experiments. Therefore, using an equivalent capillary
pellets is, as calculated using Eq. (5), approximately 0.68 MPa,         size derived from these experiments leads to a systematic under-
which is only a fraction of the pressure applied during the prep-        estimation of contact angles and can even cause impossible re-
                                                                         sults such as negative contact angles, as would be the case for
                                                                         paraffin oil.

                                                                                                 IV. Conclusions
                                                                         Precompacted submicrometer powders with reproducible prop-
                                                                         erties were generated by means of uniaxial pressing followed by
                                                                         CIP. The pore size and porosity were, respectively, measured by
                                                                         mercury intrusion porosimetry and the Archimedes method and
                                                                         used to calculate the geometrical factor in the Washburn equa-
                                                                         tion, which in turn was used to obtain the contact angle from
                                                                         capillary rise experiments performed using powder compacts.
                                                                         This modified procedure allows determining the wetting behav-
                                                                         ior of finely divided solids with a higher accuracy and repro-
                                                                         ducibility than reported in the literature for micrometer-sized
Fig. 3. Capillary rise data obtained from three measurements using       powders using a traditional capillary rise procedure and without
freshly prepared demineralized water.                                    the need for a perfectly wetting reference liquid.
2518                                             Rapid Communications of the American Ceramic Society                                              Vol. 93, No. 9
                              Acknowledgment                                              M. Lazghab, K. Saleh, I. Pezron, P. Guigon, and L. Komunjer, ‘‘Wettability
                                                                                      Assessment of Finely Divided Solids,’’ Powder Technol., 157 [1–3] 79–91 (2005).
   Gratitude is extended to the Chemical and Civil Engineering Departments of             R. Mohammadi and A. Amirfazli, ‘‘Contact Angle Measurements for Dis-
K. U. Leuven for surface tension and mercury intrusion porosimetry measure-           persed Microspheres Using Scanning Congocal Microscopy,’’ J. Dispersion Sci.
ments.                                                                                Technol., 25 [5] 567–74 (2004).
                                                                                           M. Preuss and H.-J. Butt, ‘‘Measuring the Contact Angle of Individual Col-
                                                                                      loidal Particles,’’ J. Colloid Interface Sci., 208 [2] 468–77 (1998).
                                  References                                               T. N. Hunter, G. J. Jameson, E. J. Wanless, D. Dupin, and S. P. Armes,
                                                                                      ‘‘Adsorption of Submicrometer-Sized Cationic Sterically Stabilized Polystyrene
    S. U. Pickering, ‘‘Emulsions,’’ J. Chem. Soc., 91, 2001–21 (1907).                Latex at the Air–Water Interface: Contact Angle Determination by Ellipsometry,’’
    R. Aveyard, B. P. Binks, and J. H. Clint, ‘‘Emulsions Stabilised Solely by Col-   Langmuir, 25 [6] 3440–9 (2009).
loidal Particles,’’ Adv. Colloid Interface Sci., 100–102, 503–46 (2003).                   T. T. Chau, W. J. Bruckard, P. T. L. Koh, and A. V. Nguyen, ‘‘A Review of
    B. Neirinck, J. Fransaer, O. Van der Biest, and J. Vleugels, ‘‘Production of      Factors that Affect Contact Angle and Implications for Flotation Practice,’’ Adv.
Porous Materials Through Consolidation of Pickering Emulsions,’’ Adv. Eng.            Colloid Interface Sci., 150 [2] 106–15 (2009).
Mater., 9 [1–2] 57–9 (2007).                                                               A. Siebold, M. Nardin, J. Schultz, A. Walliser, and M. Oppliger, ‘‘Effect of
    B. Neirinck, J. Fransaer, O. Van der Biest, and J. Vleugels, ‘‘Electrophoretic    Dynamic Contact Angle on Capillary Rise Phenomena,’’ Colloid Surf. A, 161 [1]
Deposition of Liquid Templates,’’ Key Eng. Mater., 412, 299–305 (2009).               81–7 (2000).
   5                                                                                     14
    B. Neirinck, J. Vleugels, J. Fransaer, and O. Van der Biest, ‘‘Electrophoretic         U. Teipel and I. Mikonsaari, ‘‘Determining Contact Angles of Powders by
Deposition of Particle-Stabilized Emulsions’’; pp. 13–25 in Developments in Po-       Liquid Penetration,’’ Part. Part. Syst. Char., 21 [4] 255–60 (2004).
rous, Biological and Geopolymer Ceramics, Edited by J. Salem, and D. Zhu. Wiley,           H. Kita, H. Hyuga, K. Yoshida, K. Osumi, and T. Iizuka, ‘‘Enhancement of
London, 2007.                                                                         Hydrophilic Properties of Alumina-Based Ceramics,’’ J. Ceram. Soc. Jpn., 114
    B. Neirinck, J. Fransaer, O. Van der Biest, and J. Vleugels, ‘‘A Novel Route to   [1328] 347–50 (2006).
Produce Porous Ceramics,’’ J. Eur. Ceram. Soc., 29 [5] 833–6 (2009).                       A. Siebold, A. Walliser, M. Nardin, M. Oppliger, and J. Schultz, ‘‘Capillary
    U. T. Gonzenbach, A. R. Studart, E. Tervoort, and L. J. Gauckler, ‘‘Ultra-        Rise for Thermodynamic Characterization of Solid Particle Surface,’’ J. Colloid.
stable Particle-Stabilized Foams,’’ Angew. Chem. Int. Ed., 45 [21] 3526–30 (2006).    Interface Sci., 186 [1] 60–70 (1997).                                          &

To top