Influence of colloidal interactions on pigment coating layer structure formation

Influence of colloidal interactions on pigment coating layer structure formation

Journal of Colloid and Interface Science 332 (2009) 394–401 Contents lists available at ScienceDirect Journal of Colloid and Interface Science www.e...

1MB Sizes 0 Downloads 8 Views

Journal of Colloid and Interface Science 332 (2009) 394–401

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Influence of colloidal interactions on pigment coating layer structure formation Anders Sand a,∗ , Martti Toivakka a , Tuomo Hjelt b a b

Laboratory of Paper Coating and Converting, and Center for Functional Materials, Åbo Akademi University, Porthaninkatu 3, 20500 Turku, Finland KCL, Tekniikantie 2, 02150 Espoo, Finland

a r t i c l e

i n f o

a b s t r a c t

Article history: Received 28 October 2008 Accepted 5 January 2009 Available online 10 January 2009

The consolidation of pigment coating layers was simulated using a three-dimensional particle dynamics model. The model included hydrodynamic interactions as well as colloidal force models and the Brownian motion. The impact of various colloidal model parameters on the z-direction solids profile development and coating layer thickness was investigated. Also, the influence of continuous (liquid) phase viscosity was tested. Particle systems resembling a polydisperse ground calcium carbonate (GCC) distribution were studied. Results show that a lower particle surface potential on pigments increased the thickness of the coating layer, while the electrostatic double layer thickness influenced the internal coating structure. An increased viscosity of the continuous phase slowed the consolidation down, but did not have a significant impact on the final microstructure. The work contributes to the understanding of the influence of colloidal system properties on the consolidation and structure development of coating layers. The results may aid in the understanding of the impact of chemical additives on coating layer structure formation. © 2009 Elsevier Inc. All rights reserved.

Keywords: Pigment coating Suspension Colloidal interactions DLVO Consolidation Dynamics Solids structure

1. Introduction Pigmented coatings are widely used for improving the properties of paper. The coating can be applied using several different methods of which the most common is the application of an aqueous slurry of pigment, binder and additives followed by blade or rod metering. After application, the coated paper is dried by IR-, air foil or cylinder dryers [1,2]. Coating layer microstructure, both in terms of z-direction solids content gradients and distribution of different coating components (e.g. binders), has been shown to have an impact on many coated paper properties. These include the optical and mechanical properties of the coating layer. Runnability on the printing press and the resulting print quality are also related to the microstructural arrangement of the coating layer [3–6]. Print quality relates to various coating layer structural properties, both at the surface and in the bulk coating. These properties can include permeability [7], weight variations [8], surface chemistry and the surface topography [9]. It is believed that colloidal interactions between particles in a coating slurry play a relevant role in the formation of the coating layer structure in coating processes. Interactions on microscopic scale influence macroscopic, observable properties both in free flowing and consolidated coatings. For example, interparticle

*

Corresponding author. Fax: +358 2 215 3226. E-mail addresses: [email protected]fi (A. Sand), [email protected]fi (M. Toivakka), [email protected]fi (T. Hjelt). 0021-9797/$ – see front matter doi:10.1016/j.jcis.2009.01.005

© 2009

Elsevier Inc. All rights reserved.

separations are made evident both in the energy dissipation of the flow (i.e. viscosity) as well as in the particle packing density of the dried coating (i.e. porosity). Deliberate or unintentional alteration of the colloidal properties of the coating slurry, e.g. by chemical additions, can be a potential cause for runnability problems during coating and also induce quality problems to the finished coated product [10–12]. The drying process of a pigment coating was divided by Watanabe and Lepoutre [13] into three phases, separated by a first and second critical concentration (FCC and SCC). A third point, called the inter-critical concentration (ICC) was introduced by Laudone et al. [14]. Consolidation of a coating layer is commonly divided into these phases [15]. The initiation of each phase is the starting point of different particle/liquid redistribution processes, which may impact the cross-structural solids concentration and particle distributions [16]. Filter cake formation at the coating/substrate interface [17–19] is known to take place as the continuous phase of the coating formulation is absorbed into the substrate. Also, an increased particle concentration at the suspension surface, called skinning, has been observed experimentally [20]. Continuum models for describing film formation during drying have been suggested and applied in several publications. These studies have mostly focused on the collective behaviour of the particle/liquid system, not taking individual particles into account. Some examples of such studies, reporting on the mechanical and structural properties of drying films, can be found in [21–23]. Modelling and simulation has grown in popularity over the years and has become complementary to experimental methods

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

in the studying of pigment coating phenomena. This is partly a result of the increasing sophistication and reliability of the models used. Overviews on the progress of different computational methods have been presented by Vidal and Bertrand [24] as well as Pianet et al. [25]. The simulation methods used, have ranged from relatively simple deposition methods [26–29] to deterministic particle motion models including advanced hydrodynamics, colloidal and other interaction forces [30–33]. The current work is based on the Stokesian dynamics simulation technique, which is a general method for studying the interactions between colloidal-sized particles in arbitrary flow fields [34–36]. The technique was first introduced to the field of pigment coating by Bousfield [37], and has since then been applied to many different coating-related phenomena. Such studies have included 2D microstructure development and levelling of coating layers at different drying rates [38], as well as studies on pigment suspension rheology [39,40]. Eriksson et al. [41] studied experimentally, the influence of solids content and dispersant addition on the colloidal properties of the coating system. The present work develops on the research in [41] by reporting on possible influences the changes in colloidal parameters might have on the structure of the coating layer. Furthermore, it extends the work reported in other publications by Sand et al. [16,42–44], which focused on the influence of drying strategy and applied coating layer thickness. The aim of this paper is to understand the influence of colloidal interactions on the consolidation of coating layers. The dynamics of coating layer formation is studied numerically using different colloidal force model parameters. Also, in order to clarify the possible effects of added chemicals such as thickeners, the influence of continuous phase viscosity on consolidation, was investigated. The method and results of this work also find relevance in fields other than pigment coating. Such fields might include e.g. paints, printing and inks or processes where particulate suspensions interact with porous substrates, such as in filtration or evaporation processes. 2. Materials and methods The simulation results presented in this paper are based on a modified Stokesian dynamics approach as described by Toivakka [45], which was further developed by Nopola to allow the study of 3D systems [46]. The model comprises lubrication-type hydrodynamic interactions, a free surface model, DLVO-type colloidal interactions, the Brownian motion and a Born repulsion model for particles in close contact. Only the colloidal interaction models are described below, while details of the other models can be found elsewhere [45,46]. The colloidal interaction model is expressed by a combination of electrostatic repulsion and van der Waals attraction forces. Furthermore, a Born repulsion model is used to prevent particle overlapping at close contact. The electrostatic repulsion model is expressed as a pairwise short-range DLVO-type repulsive force [47,48] as F el = 4πκεr ε0 ψ1 ψ2

a1 a2

e −κ Δ

a1 + a2 1 + e −κ Δ

,

(1)

where κ is the reciprocal double layer thickness, εr is the dielectric constant of the continuous phase, ε0 the permittivity of vacuum, ψ1 and ψ2 the surface potentials of the interacting particles and Δ the surface separation distance. a1 and a2 are the radii of the particles. The van der Waals attraction component is approximated as reported by Dabros and van de Ven [48] as F vdw = − A H

a1 a2 a1 + a2



 λ(λ + 22.232Δ) , 6Δ2 (λ + 11.116Δ)2

(2)

395

Table 1 Colloidal interaction model parameters, as used in this work. Symbol

εr ε0 ψ1 , ψ2 a1 , a2

Δ 1/κ AH

λ a

Value

Description

Electrostatic repulsion parameters 80 Continuous phase dielectric constant (water)a − 12 2 8.85 × 10 C /N m2 Permittivity of vacuum −50 to +50 mV Surface potential of interacting particles Radii of interacting particles Surface separation distance 2.5 to 10.0 nm Double layer thickness Van der Waals attraction parameters 1 × 10−21 J Hamaker constant 100 × 10−9 m London characteristic wavelength

Dielectric constant of vacuum 1.0 per definition.

where A H is the Hamaker constant and λ the London characteristic wavelength [49]. The parameters used in this work are presented in Table 1. It should be noted that the surface potential of calcite in aqueous suspensions is traditionally described by the zeta potential, ζ , which can be measured by a number of different electrokinetic techniques [50]. As the zeta potential measures the potential at a certain distance from the particle surface, it does not correspond directly to the surface potential, ψ . The surface potential and zeta potential relationship can be derived using the Stern model of the double layer. This model for charge distribution is, however, somewhat crude and not necessarily suitable for calcite surfaces in water [41,50]. The continuous phase (water/electrolyte) dielectric constant was set to 80 (pure water at NTP), although it has been shown to vary slightly with electrolyte concentration (e.g. 84 at 1.0 M of CaCl2 ) [53]. However, as the solubility of carbonate in water is low, and the concentration of other ions is also expected to be relatively low (>0.5 M), the effect of changes in the dielectric permittivity on the electrostatic repulsion model can be considered negligible. The current model finds its best accuracy roughly between the application of coating to the substrate and the second critical concentration (SCC). Beyond that point, menisci will form between particles and air will start to penetrate into the porous structure, making forces not covered by the Stokesian dynamics technique come into play. For instance, the large negative capillary pressures generated at the menisci will give rise to strong attractive forces between particles. The model utilised in this work does not take such considerations into account. However, a free surface model was introduced to enable some form of free surface interactions and to partly offset the constraints discussed above. The free surface model describes the interactions between the free surface and particles by a hydrodynamic/surface tension force, which has been discussed in detail in other work [38,59]. Although the model is not able to grasp the full transition from an aqueous coating slurry to a granular-type system and finally a dry coating layer, this paper gives an indication to particle redistribution and structure build-up at the point when the coating particles are most mobile. This has been shown in other publications to take place roughly between 0 and 0.4 s. Consequently, results beyond this timescale are not reported in this work [16,42,44]. 2.1. Theoretical considerations The Peclet number, Pe, is used to compare the convective effect with the effect of diffusion of particles. Thus, it can be utilised in estimating the relative influence of the hydrodynamic forces compared to diffusion. The convection in this work is driven by liquid absorption into the substrate and evaporation from the free sur-

396

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

face, while the diffusion results from the Brownian motion. For the current problem, the Peclet number can be expressed as Pe =

6πμaH U kT

(3)

,

where μ is the viscosity of the continuous phase, a the characteristic particle size, H the characteristic length scale, U the characteristic velocity, k the Boltzmann constant and T temperature. Peclet numbers larger than one indicate that effect of diffusion is negligible [51,52]. The calculation of the Peclet number is not straight-forward in these simulations, since many of the conditions change during the consolidation process. However, considering values relevant to the simulations of the present work, it can be concluded that the Peclet number varies from over 10,000 at the early stages of the consolidation for large particles experiencing rapid dewatering/evaporation (e.g. a = 2 μm, H = 5 μm, U = 30 μm/s, T = 298 K) to values towards one and below at the end for the smallest particles and minimal dewatering/evaporation (e.g., a = 0.2 μm, H = 1 μm, U = 1 μm/s, T = 353 K). The characteristic length scale above is chosen such that movement of particles over this length scale would have a significant effect on the microstructure of the coating layer and its properties. The consideration above indicates that the effect of Brownian motion in simulations is mostly insignificant, while possibly having an effect for the smallest particle size fraction. Pigment coating formulations include polymeric additives and soluble binders, for which the Peclet number would be even smaller. It should also be noted that the Brownian motion might also have an effect on the structure formation by helping particles to rearrange and move past another. Similarly, a dimensionless expression can be formulated to assess the relative importance of the viscous (hydrodynamic) forces and the colloidal interaction forces. This expression, R, can be written as R=

aμU F el + F vdw

(4)

,

where values below 1 indicate a higher relative influence of colloidal forces. Estimating from the parameters of the current simulations, we can conclude that this ratio can vary in the range of 0.15 to infinity. Typically, the relative influence of colloidal forces can be expected to dominate at the later stages of the consolidation while the hydrodynamics dominate in the beginning. Although these dimensionless numbers aid in the understanding of the relative influences of the three dominating force models, hydrodynamic, Brownian and colloidal forces, it need to be stressed that their values should only be considered as indicative. As a simulation is composed of 1700 interacting particles in a complex system with constantly changing conditions, many of the variables of the equations above will be particle pair specific and also change with time. Considering that the particle system is generated with a size distribution, the ratios will also be different for different particle size fractions. 2.2. Simulation setup Double layer thicknesses, 1/κ , in the range of 2.5 to 10 nm were tested along with particle surface potentials, ψ , between −50 and +50 mV. The values were selected according to what has been experimentally determined as realistic for coating suspensions [41,54,55]. A consolidating particle suspension of fine-grade GCC-corresponding size distribution was simulated. The distribution satisfied the log-normal distribution described by the equation p (d) =

1



σd 2π



exp

−(d − μd )2 2σd2



,

(5)

Fig. 1. Simulated log-normal cumulative weight distribution compared to experimental fine grade GCC distribution with and without normalisation for the size fraction cut-off.

with the mean particle diameter, μd , 1.5 μm and a standard deviation, σd , of 10.0 μm. This compares to an experimental GCC distribution (CoverCarb,CC75, Omya Ag, Switzerland) determined by a Malvern Instruments Zetasizer 3000. In order to limit the number of particles in the systems simulated, a size fraction cut-offs was applied for particles below 0.2 and above 2 μm in diameter. A comparison between the simulated particle size distribution and the experimentally determined distribution is shown in Fig. 1. The number of particles used in each simulation was about 1700. The initial positions of the particles were set by a random placement procedure and the initial dry solids content of the suspension was 65 wt%. Two different coating layer post-metering thicknesses were set at 20 and 30 μm. These thicknesses denote the situation at metering and as the filtercake already starts to form before this point, the actual coating thickness at t = 0 s will be somewhat higher. The target post-metering thickness should be achieved at 0.1 s, following the initial filtercake formation, but may show some deviation due to the different colloidal parameters utilised in simulation. The thicknesses are somewhat higher than the most common commercial LWC (Light Weight Coated) and MWC (Medium Weight Coated) paper grades [56]. However, based on earlier experiences [16,42–44], these thicknesses give good accentuation of coating layer structural properties, which is the interest of this study. Note that these are the approximate thicknesses just after metering. Thus, before metering, the coating layers will be thicker, while shrinking below the post-metering value as consolidation proceeds. The consolidation of a coating layer is driven by simultaneous absorption of liquid into the substrate and evaporation from the coating free surface. Since the influence of various drying strategies was not the focus of this study, all the simulations performed in this work utilised the same dewatering strategy. The dewatering profile was based on macroscopic drying simulations and pilot trials [57], as described by Sand et al. in [44] and [58]. The most likely drying method to be applied in industrial coating was utilised. This involved an initial intense drying step (usually by IR and air foil dryers) followed by mild drying (cylinder drying), which corresponds to the high–low–low (HLL) drying case in [57, 58]. The drying strategy is presented in Fig. 2. The influence of colloidal parameters on the interaction forces between particles are shown in Fig. 3. For the illustration, only a few cases of particle surface potentials and double layer thicknesses for two equally sized particles (D = 1 μm) are shown. Note that the model calculates the electrostatic repulsion component by multiplying the particle surface potential of two equally charged particles. Thus, the model does not distinguish between

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

397

Table 2 Calculated interparticle separation distances and energies in the energy curve at a few different particle surface potentials and double layer thicknesses. 1/κ [nm]

Primary maximum

Secondary minimum

[mV]

[nm]

Energy/kT

[nm]

Energy/kT

−50 −50

10 2 .5

0 .4 0.2

900.5 848.1

147.8 29.0

−0.003 0.071

−25 −25

10 2 .5

7 .9 4 .0

212.8 187.1

130.2 24.3

0.004 0.094

−5 −5

10 2 .5

4 .1 2 .6

5.3 1.1

86.1 11.6

0.009 0.255

0 0

10 2 .5

a

a

a

a

a

a

a

a

ψ

a

No electrostatic barrier.

Fig. 2. Drying strategy utilised in simulations. The drying method is composed of evaporation from the top of the coating and absorption by the base substrate, and corresponds to the HLL case in [57].

Fig. 3. DLVO interaction forces between two D = 1 μm particles at a few cases of different double layer thicknesses and surface potentials. The zero interparticle force corresponds to the secondary minimum of the particle interaction energy curve calculated by the DLVO theory.

the −50 mV and +50 mV potential. When referring to maximum and minimum particle surface potentials, it is the absolute amounts that are considered, and 0 mV is consequently the minimum surface potential. In Table 2, it is shown how some key DLVO-related interparticle separation distances and interaction forces change with a few different particle surface potentials and double layer thicknesses. All cases were calculated for two equally sized particles of D = 1 μm and the zero interparticle force in the table indicates the surface separation distance at the secondary minimum of the DLVO model. The weaker electrostatic repulsive force in the low particle surface potential cases is reflected by the lower electrostatic barrier. Furthermore, as can also be observed from Table 2, a reduction in double layer thickness will also weaken the electrostatic barrier. 3. Results and discussion The solid structures in this work are calculated as z-direction solids profiles. Solid structures in the lateral directions are for this application considered as being of minor interest and therefore omitted. For converting the solids content to wt%, all solid material is assumed to be calcite (ρ = 2700 kg/m3 ), and non-solid material is pure water (ρ = 1000 kg/m3 ). Given the time scales reported in this work, before the second critical concentration (SCC), air penetration into the structure has a relatively small influence on the

Fig. 4. Example of consolidating particle system with plotted trajectories (20 μm thick coating layer with 1/κ = 10 nm and ψ1 and ψ2 = 25 mV). The colour coding signifies the velocity of particles (red = fastest, blue = slowest). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

results. Consequently, for the particle size distribution used in this work, the coating layer can be considered consolidated at around 85 wt%, which normally occurred within 0.3–0.5 s [16]. An example of a consolidating system with plotted (non-wrapped) particle trajectories, is shown in Fig. 4. 3.1. The influence of viscosity on coating structure Thickener addition increases the viscosity of a coating slurry. The increase can be caused by a number of mechanisms, e.g. by increased interaction between the pigment particles (e.g. bridging from adsorbed polymers), by depletion flocculation (increased osmotic pressure differences due to free polymer in the continuous phase) or by simple increase of the continuous phase viscosity leading to an increased energy dissipation from the particle movement relative to each other. The influence of the last mechanism was investigated for a few selected cases. Fig. 5 shows the difference in solids structure of a consolidating coating layer after 0.2 s

398

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

Fig. 5. Coating layer solids structure at different continuous (liquid) phase viscosities (20 μm layer, 1/κ = 10 nm, ψ = −50 mV at 0.2 s).

Fig. 7. Coating layer thickness (30 μm layer, 1/κ = 10 nm) when applying different particle surface potentials to the coating system. The different curves illustrate how the coating thicknesses change with time.

Fig. 6. z-Direction solids structure of the 30 μm coating layer (at 0.2 s, 1/κ = 10 nm) applying different particle surface potential. Positive surface potential cases were omitted.

when using 3 different continuous phase viscosities. In the example, the double layer thickness, 1/κ , was set to 10 nm and the particle surface potential, ψ , was −50 mV. It can be observed that an increased viscosity slows down the consolidation, with implications such as immobilisation at an earlier stage and possibly reduced shrinkage of the coating layer. This would also result in a less dense coating structure, possibly with persisting solids concentration gradients towards the substrate and the top side of the coating. The locally increased solids concentration near the coating/air interface results as the hydrodynamics near the free surface [45,58,59] works to push particles downwards into the suspension. This effect is commonly termed skinning. However, due to the fact that consolidation takes place at different rates, it is difficult to show whether the high-viscosity cases result in denser skinning. 3.2. The influence of particle surface potential The particle surface potential had a substantial impact on the shrinkage of the consolidating coating layer. As can be seen in Fig. 6, there could be up to a 20% difference in coating layer thickness. Furthermore, it could be concluded that the surface potential did not give rise to any systematic changes in internal solids structure in the z-direction. As shown in Fig. 7, particle surface potentials close to 0 mV contributed to significantly thicker coating layers while having much less influence at the higher surface potentials. Consequently, following a relatively sharp drop at 0– 10 mV, the thickness evened out at higher particle surface potentials.

Fig. 8. Coating layer shrinkage (30 μm layer, 1/κ = 10 nm) as function of time at different particle surface potentials.

Thus, a reduced electrostatic repulsion between particles did not yield more compact structures. As the attractive van der Waals forces dominated in these cases, the result might seem somewhat counterintuitive. However, the effect can be explained by the attractive force generating more loosely packed aggregates, which due to the lack of repulsive force hindered particles from arranging into a more dense packing. The issue of particle clustering will be more thoroughly addressed later in this paper. Regardless of particle surface potential, the shrinkage of the coating layers proceeded at roughly the same rate, Fig. 8. An indication of particle arrangement can be obtained by calculating particle surface separation distributions. The radial distribution is generated, for each particle, by calculating surface separation distances to its nearby particles. This will give a probability distribution, at given colloidal settings, of where particles can be found relative to each other. Thus, strong particle flocculation can be seen as sharp peaks at smaller interparticle distances. The polydispersity of the particle size distribution will induce an error to the probability at large interparticle distances. To reduce the error, the calculation is only extended to the radius of the smallest particle in the system. The surface separation distribution of some combinations of particle surface potential and double layer thickness are shown in Fig. 9. The results support the previous theory of strongly aggregated clusters of particles at low surface potentials and double layer thicknesses. Consequently, particles are not allowed to reach a high packing degree, resulting in reduced coating layer shrink-

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

399

Fig. 10. Solids structure in 30 μm coating layer (ψ = 25 mV) at time 0.2 s. Difference in filter cake and skin density.

Fig. 9. Surface separation distributions at different colloidal parameters.

age. The broad distribution at large interparticle distances, as is the case with large double layer thicknesses and surface potentials, indicates particle arrangement into loose structures. These particles are able to move relative to each other and form denser structures than what is the case with strongly aggregated particles. 3.3. The influence of double layer thickness The minimum double layer thickness used in simulations was 2.5 nm. Due to the extreme gradient of the interaction energy curve at low double layer thicknesses, it was not possible to simulate values below this. The reason for this is that high force gradients lead to numerical difficulties in the integration of interparticle forces and excessive computational times. However, the double layer thickness of calcite particle pigment slurries can typically be expected within the 5–10 nm range [51]. In contrast to the particle surface potential, the double layer thickness appeared to have a relatively small influence on the thickness development of the coating layer. However, the internal solids structure of the coating layer varied quite notably with different double layer thicknesses. This was the case both while observing filter cake formation at the coating/substrate interface, as well as skinning at the surface of the coating. Fig. 10 shows the solids structure of the 30 μm coating layer at 0.2 s, with 4 different double layer thicknesses. As can be seen in the figure, the layer is at such a consolidation stage, that both filter cake formation (at 0–10 μm) and skinning (at roughly 25 μm) are taking place. When comparing the most extreme cases, 2.5 and 10 nm, there is a trend of more distinct regions of filter cake, bulk coating and skin with the 2.5 nm double layer. In the 10 nm case, the solids gradients are smaller and the regions less apparent. The two intermediate cases, 5 and 7.5 nm, fit well into the trend. All the coating layers have roughly the same thickness, whereby it is concluded that the double layer thickness will mostly have an influence on the internal coating structure. The corresponding situation can also be seen in Fig. 11, which illustrates the solids structure at an earlier degree of consolidation. Drying of the layer, with skinning as a result, has not yet started. Thus, at this point, there is only filter cake formation. Also in this example, the same tendency can be observed. The filter cake appears much more distinctly in the high double layer thickness case, as a result of the higher relative solids concentration in the filter cake. The bulk coating structure, however, appears almost identical in the non-filter cake region. The overall thickness of the coating layer is also the same in all cases.

Fig. 11. Solids structure in 30 μm coating layer (ψ = 25 mV) at time 0.1 s. Difference in filter cake density.

Another option to illustrate the structural differences in the coating layer, is to use cumulative dry solids curves. In Fig. 12, the cumulative solids structures are shown for t = 0.2 s case above, calculated from the substrate/filter cake interface towards the top of the coating layer. The figure also shows an enlargement of the filter cake. Not only does this method of visualisation accentuate the differences in structure, but also shows how far the structures extend in the z-direction. It can be seen that it is primarily the filtercake, but to some extent also the skin that is affected when altering the thickness of the double layer. The trend of increased dry solids content in the filter cake can be clearly seen as result of reduced double layer thickness. The maximum change in the simulations was about 5 wt%. The influence of the reduced double layer thickness in the skin formation was even less, only some 2 wt%. One should note that the current model excludes the influence of increasing local electrolyte concentration due to water evaporation. The evaporation would further decrease the thickness of the electrostatic double layer, but also possibly coagulate the system into the primary minimum of the particle interaction energy curve. 4. Summary Colloidal interaction model parameters were shown to have an influence both on the microstructure of the coating layer and on macroscopic properties such as the thickness of the coating layer. Altering the continuous phase viscosity of the coating system, which in practical coating processes can be achieved by the addition of thickener, showed that coating layer structure build-up was significantly slower at higher continuous phase viscosities. This could also possibly result in higher immobilisation thicknesses of the coating layers.

400

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

Fig. 12. Cumulative solids in the full coating layer (30 μm, ψ = 25 mV) and filter cake at t = 0.2 s.

It was shown that lower particle surface potentials, resulting in reduced electrostatic repulsion, generated loose particle clusters. This agglomeration of particles disallowed the system from reaching denser packing degrees and consequently resulted in a lower tendency for shrinking. The particle surface potential did not appear to have any systematic impact on the internal structure formation of coating layers. Changes in the electrostatic double layer thickness, however, were shown to influence the internal structure of the coating. A lower double layer thickness increased the energy barrier in the DLVO-model but simultaneously decreased its range, which allowed for sharper internal solids concentration gradients to form within the coating layer. This could be observed as faster solids concentration built up in the filter cake, while having a lesser effect on skinning. As the particle surface potential of pigments is influenced by adsorbed chemicals on their surfaces and the double layer thickness is reflected by ion concentration in the continuous phase, these results may aid in the dosing of additives to obtain coating layers of desirable porosity and structural properties. Acknowledgments The work is part of the KCLCONS consortium project “Coating Mechanisms” coordinated by KCL (Oy Keskuslaboratorio-Centrallaboratorium Ab). We thank Mr. Jani Kniivilä for developing visualisation tools and Dr. Parvez Alam for proofreading the text. The National Technology Agency of Finland (TEKES) and the international Ph.D. Programme in Pulp and Paper Science and Technology in Finland (PaPSaT) are acknowledged for financial support. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11]

P. Rajala, H. Häkkänen, C.-G. Berg, in: 13th Int. Drying Symp., 2002, p. 1800. P. Rajala, Ph.D. thesis, Helsinki Univ. Tech., Finland, 2004. N. Santos, J. Velho, Pulp Pap. Can. 105 (9) (2004) 43. M. Parpaillon, G. Engström, I. Pettersson, I. Fineman, S. Svanson, B. Dellenfalk, M. Rigdahl, J. Appl. Polym. Sci. 30 (1985) 581. P. Lepoutre, Prog. Org. Coat. 17 (1989) 89. K. Prall, S. Shaler, P. Lepoutre, Nord. Pulp Paper Res. J. 15 (5) (2000) 564. Y. Shen, D. Bousfield, A. van Heiningen, D. Donigian, J. Pulp Pap. Sci. 31 (3) (2005) 105. T. Inoue, H. Matsubayashi, J. Kline, R. Janes, in: Proceedings of TAPPI International Printing and Graphic Arts Conference, 1992, p. 33. G. Ström, J. Pulp Pap. Sci. 19 (2) (1993) 79. A. Vrij, R. Tuinier, in: J. Lyklema (Ed.), Structure of Concentrated Colloidal Dispersions, Fundamentals of Interface and Colloid Science, vol. 4, Particulate Colloids, Elsevier, 2005. P. Dahlvik, P. Salminen, G. Ström, in: Proceedings of TAPPI Coating Fundamentals Symposium, 1995, p. 63.

[12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]

P. Dahlvik, G. Ström, D. Eklund, J. Nord. Pulp Pap. Res. 12 (1) (1997) 61. J. Watanabe, P. Lepoutre, J. Appl. Polym. Sci. 27 (1982) 4207. G. Laudone, G. Matthews, P. Gane, J. Colloid Interface Sci. 304 (2006) 180. H. Al-Turaif, P. Lepoutre, Prog. Org. Coat. 38 (1) (2000) 43. A. Sand, M. Toivakka, T. Hjelt, Nord. Pulp Pap. Res. J. 23 (1) (2008) 46. G. Engström, Wochenbl. Papierfabr. 114 (6) (1986) 195. U. Eriksson, M. Rigdahl, J. Pulp Pap. Sci. 20 (11) (1994) J333. S. Lohmander, M. Martinez, L. Lason, M. Rigdahl, T.-Q. Li, J. Pulp Pap. Sci. 27 (6), p. 183. Y. Ma, H. Davis, L. Scriven, Prog. Org. Coat. 52 (1) (2004) 46. W. Russel, N. Wu, W. Man, Langmuir 24 (5) (2008) 1721. M. Tirumkudulu, W. Russel, Langmuir 20 (7) (2004) 2947. H. Lei, J. Payne, A. McCormick, L. Francis, W. Gerberich, L. Scriven, J. Appl. Polym. Sci. 81 (4) (2001) 1000. D. Vidal, F. Bertrand, in: Proceedings of Tappi Advanced Coating Fundamentals Symposium, 2006, p. 241. G. Pianet, F. Bertrand, D. Vidal, B. Mallet, in: Proceedings of Tappi Advanced Coating Fundamentals Symposium, 2008, p. 278. G. Eksi, D. Bousfield, Tappi J. 80 (2) (1997) 127. T. Hiorns, T. Nesbitt, in: Proceedings of Tappi Advanced Coating Fundamentals Symposium, p. 252. D. Vidal, X. Zou, T. Uesaka, Tappi J. 86 (4), p. 3. D. Vidal, X. Zou, T. Uesaka, Tappi J. 86 (5), p. 16. M. Toivakka, P. Salminen, Y. Chonde, D. Bousfield, in: Proceedings of Tappi Coating Fundamentals Symposium, 1997, p. 89. T. Phung, J. Brady, G. Bossis, J. Fluid Mech. 313 (1996) 181. R. Jones, J. Chem. Phys. 115 (11) (2001) 5319. F. Bertrand, T. Gange, E. Desaulniers, D. Vidal, R. Hayes, Comput. Chem. Eng. 28 (2004) 2595. J. Brady, G. Bossis, Annu. Rev. Fluid. Mech. 20 (1988) 111. J. Brady, R. Phillips, J. Lester, G. Bossis, J. Fluid Mech. 195 (1988) 257. L. Durlofsky, J. Brady, J. Fluid Mech. 200 (1989) 39. D. Bousfield, in: Proceedings of Tappi Coating Conference, 1990, p. 325. M. Toivakka, D. Eklund, D. Bousfield, in: Proceedings of Tappi Coating Conference, 1992, p. 403. M. Toivakka, D. Eklund, Tappi J. 79 (1) (1996) 211. M. Toivakka, D. Eklund, D. Bousfield, J. Non-Newtonian Fluid Mech. 56 (1995) 49. R. Eriksson, J. Merta, J. Rosenholm, J. Colloid Interface Sci. 313 (1) (2007) 184. A. Sand, M. Toivakka, T. Hjelt, Nord. Pulp Pap. Res. J. 23 (1) (2008) 52. A. Sand, M. Toivakka, T. Hjelt, in: Proceedings of Progress in Paper Physics Seminar, 2008, p. 65. A. Sand, M. Toivakka, T. Hjelt, in: Proceedings of Tappi Advanced Coating Fundamentals Symposium, 2008, p. 289. M. Toivakka, Simulation of Particle Motion in Pigment Coating Colours, Ph.D. thesis, Åbo Akademi, 1997. T. Nopola, Simulation of Particle Motion in Concentrated Colloidal Suspensions, Lic. thesis, University of Turku, 2004. A. Suzuki, N. Ho, W. Higuchi, J. Colloid Interface Sci. 29 (3) (1969) 552. T. Dabros, T. van de Ven, J. Colloid Interface Sci. 149 (2) (1992) 493. R. Hogg, T. Healy, D. Fuerstenau, Trans. Faraday Soc. 62 (6) (1966) 1638. L. Madsen, in: A. Hubbard (Ed.), Encyclopedia of Surface and Colloid Science, vol. 4, Surface Charge of Calcite, Dekker, 2002. P. Hiemenz, Principles of Colloid and Surface Chemistry, Dekker, 1977. P. Ekanayake, P. McDonald, J. Keddie, Eur. Phys. J. Spec. Top. 166 (2009) 21. V. Navarkhele, R. Agrawal, M. Kurtadikar, Pramana J. Phys. 51 (3&4) (1998) 511.

A. Sand et al. / Journal of Colloid and Interface Science 332 (2009) 394–401

[54] M. Vanni, G. Baldi, Adv. Colloid Interface Sci. 97 (2002) 151. [55] H. Ackler, R. French, Y.-M. Chiang, J. Colloid Interface Sci. 179 (1996) 460. [56] J. Linnonmaa, M. Trefz, in: E. Lehtinen (Ed.), Pigment Coating and Surface Sizing of Paper, in: Papermaking Science and Technology, vol. 11, Fapet, Helsinki, 2000.

401

[57] O. Timofeev, H. Pajari, H. Koskela, J. Paaso, H. Kiiskinen, in: Proceedings of PAPTAC Annual Meeting, 2007, p. A221. [58] A. Sand, M. Toivakka, T. Hjelt, in: Proceedings of International Conference on Computational Methods, 2007, p. 97. [59] A. Sand, T. Nopola, T. Hjelt, M. Toivakka, Comput. Chem. Eng., submitted for publication.