Use of modified silica nanoparticle for fixation and elimination of colloidal and dissolved substances from white water Aurelien Guyard, Claude Daneault and Bruno Chabot, Université du Québec à Trois-Rivières, Canada KEYWORDS: Cationic polyelectrolyte, Silica, Adsorption, Grafting, Pectic acid SUMMARY: The objective of this study was to understand how modified cationic silica, obtained by adsorption of cationic polyelectrolyte or by grafting, traps dissolved and colloidal substances of the wood and eliminates them from white water systems. Polygalacturonic acid, considered as a representative model of dissolved and colloidal substances in bleached thermomechanical pulp was used in this study. The pre-adsorption of cationic polyelectrolyte as well as the adsorption of this anionic trash on the modified silica surface were studied at pH 4.5 and 6.5 and various NaCl concentrations by isotherm adsorption measurements. Zeta potential measurements were also used to estimate the potential of the modified particles. We first showed that the adsorption of poly-DADMAC on the silica surface increased with increasing ionic strength, while the zeta potential went through an optimum. Adsorption of the polygalacturonic acid on the silica surface covered by poly-DADMAC was treated as an adsorption on a soft surface, where the interaction is viewed as being purely electrostatic, varying according to the zeta potential of cationic silica but independent of the amount of cationic polyelectrolyte pre-adsorbed. The adsorption of polygalacturonic acid on grafted silica surface was treated as an adsorption on a hard charged surface. This last interaction, which is a case of screening-enhanced adsorption, did not vary according to the zeta potential of cationic silica and was dominated by a non-electrostatic interaction at high ionic strength. ADDRESS OF THE AUTHORS: Aurelien Guyard ([email protected]), Claude Daneault ([email protected]), Bruno Chabot ([email protected]): Centre Intégré en Pâtes et Papiers, Université du Québec à Trois-Rivières, 3351 Bld. des Forges, Trois-Rivières (Qc) G9A 5H7, Canada Corresponding author: Bruno Chabot

During pulp production and subsequent treatments, numerous dissolved and colloidal substances (DCS) are released in the process water. The origin of these substances, their characteristics, as well as their behavior varies tremendously according to the nature and the type of pulp used and also to the conditions of production. For virgin pulp, these substances are produced primarily from various wood components, while their composition is significantly different in recycled pulp. Because of environmental constraints, the paper industry is required to reduce their fresh water consumption. However, closing the water circulation system of a paper machine and possibly converting it into a totally effluent-free papermaking process results in a buildup of detrimental substances in the white water. The system then becomes more sensitive to disturbances and variations in the process. This implies that the white water has to be continuously cleaned 574

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

(Linhart et al. 1987). Obviously, not all the substances present in the white water are harmful, but some of them are responsible for the main problems occurring during the papermaking process (Wearing et al. 1985). For example, these substances can adsorb or precipitate on fibers and fines and change their properties, leading to a reduction of their capacity to form hydrogen bond, to a decrease of paper gloss, or to the decrease of the paper strength (Dunham et al. 2002; Thornton et al. 1993). Furthermore, many of these harmful substances are anionic and thus react with cationic polymer, often used as retention agents. This interaction leads to the formation of polymeric complexes being able to precipitate, thus rendering ineffective the chemical process using these cationic polymers (Pelton et al. 1980; Tay 1980). It thus becomes necessary to eliminate these harmful substances or at least to decrease their negative effects in order to obtain a totally closed water circuit. Actually, white water cleaning and DCS elimination are mainly performed using cationic polymers. The aim of this chemical treatment is to neutralize the anionic trash and then to fix them in the paper (Zippel 2001; Wågberg, Alsell 1995; Gill 1996). This technique clearly leads to a decrease in the concentration of “anionic trash” in the water circuits, but it also tends to decrease the quality of the paper. The purpose of the study is the development of a process allowing the selective adsorption of various constituents of the dissolved and colloidal substances (DCS) on materials such as silica, which favors their elimination of the paper machine white water. Since both silica and “anionic trash” are negatively charged, Thorton et al. (1994), it is essential to first produce cationic silica or at least screen the electrostatic repulsion (Kekkonen, Stenius 1999). We first describe two methods to cationize silica particles and study how each surface modification can influence and affect the adsorption of DCS. The first method consists in the adsorption of cationic polymers, at various NaCl concentrations and pH. For the second method, we directly graft cationic coupling agents on the surface of the silica, at various concentrations and pH. Polygalacturonic acid (PGA), which has been identified as one of the main components of “anionic trash” in bleached themomechanical pulps, Peltonen et al. (1980), was chosen as a representative model substance. We find that the removal efficiency of the methods depends tremendously on the experimental conditions, each method having its own advantages and disadvantages. Nevertheless, it is clear that both methods can efficiently clean the white water system, thereby limiting the increase of DCS without degrading the quality of the paper.

Material and Methods Silica Silica used for the adsorption of cationic polymer was nanometric silica Ludox TM 50 purchased from Fluka [7631-86-9]. This silica has a specific surface area of 140 m2/g and a density of 1.4 g/cm3 at 25°C. These colloidal silica particles have similar characteristics as commercial silica particles generally used in papermaking in retention and drainage technology. This is the reason why they have been used in this study. Of course such silica could be used as substrate for the grafting of cationic coupling agent. However, the silica used for grafting were fumed silica with a diameter of 14 nm, purchased from Sigma Aldrich. These silica nanoparticles have a specific surface area of 200 m2/g, a density of 3.69 g/cm3 and have a number of silanol groups estimated between 3.5 and 4.5 OH/nm2. This pyrogenic silica, which is generally used as model sample for investigation of silica surface chemistry by IR spectroscopy, is thus employed for that purpose. Various diluted silica dispersions at 1% concentration by weight were prepared by diluting the commercial solution (50% by weight) with distilled water. The pH was then adjusted using solutions of HCl (0.1 M) and NaOH (0.1 M). The salt concentration was also adjusted with NaCl solutions. Cationic polymer and coupling agent The poly-diallyldimethylammonium chloride (p-DADMAC -C8H16NCl) was purchased from Aldrich [26062-79-3]. It has a density of 1.04 g/am3 at 25°C, a charge density of 5.1 meq/g, and a molecular weight (average) of 4.4x105 g/mol. The cationic coupling agent used for grafting was Ntrimethoxysilylpropyl-N,N,N-trimethylammonium chloride (C9H24ClNO3Si), purchased from United Chemicals, with a concentration varying between 47 and 58% by weight. Polygalacturonic acid The polygalacturonic acid (PGA) was purchased from Fluka [25990-10-7]. The molecular weight was approximately 25,000g/mol (White et al. 1999; Sundberg et al. 1994). Solutions at 1g/L (1000 ppm) were prepared by dilution in a 0.1M caustic solution during 24 hours. The solubilization of PGA is only possible at pH 12. Table 1 presents the effect of pH on the charge density. Table 1. Variation of the charge density of the Polygalacturonic acid according to the pH. Polygalacturonic acid pKa = 3.5 pH

charge density (meq/g)

Charge (c/g)

4.5 6.5

4.72 6.49

455 625

Grafting In a 250 mL three-neck flask equipped with continuous reflux, 1g fumed silica powder was dispersed in 50 mL of deionized water. The pH of the dispersed silica was adjusted to pH 4.5 or 9 according to the experimental

conditions, using HCl (0.1 M) and NaOH (0.1 M). 0.1 mL of the cationic silane was then added, and the mixture was continuously stirred in a water bath for 1 hour at a temperature of 25 or 80°C according to the experimental conditions. To estimate the effect of time on the grafting efficiency, 2.5 mL of silica sample was taken at various times and diluted in 50 mL of deionized water. The surface properties of the cationized silica obtained through adsorption of cationic polymers were then analyzed by measurement of zeta potentials and adsorption isotherms, while those of the grafted silica were determined using also measurements of charge density, zeta potential, total carbon content and IR spectroscopy. Fig 1 presents the grafting process of silica with the cationic silane.

Fig 1. Grafting process of the silica by a cationic silane.

Charge density The charge density of the polyelectrolytes used was determined by colloidal titration (Terayama 1952; Toei, Kohara 1976). To determine the end-point, a Mütek Charge Titrator (PCD03) was used (Kam, Gregory 1999). The charge density of the polygalacturonic acid was titrated with the standard poly-DADMAC (1 meq/L or 6.19 meq/g), while the charge density of the cationic polyelectrolyte was titrated with a sodium polyethylene sulphonate (PES-Na) with a charge density of 0.98 meq/L or 7.81 meq/g. However, if the salt concentration is too high, that is if the specific conductivity of the sample is higher than 1000 µS/cm, Chen et al. (2001), the measurement of the end-point will be distorted and overestimated (Chen et al. 2003). Under these conditions, the charge density measurement must be made for salt concentrations in the order of 10-3M with sodium chloride (NaCl). Adsorption isotherms Adsorption isotherms of polymer on the silica surface particles were determined at room temperature (25°C). The pH of polymer solutions and colloidal silica dispersions were adjusted to pH 4.5 and 6.5, respectively using 0.1M HCl or NaOH solutions. The ionic strength of all solutions was adjusted to the required level using NaCl solutions. For the adsorption of PGA on cationic modified silica, the silica concentration was 1.3 g/L and the PGA concentration was varied from 0.03 to 0.3 g/L. These mixtures were then homogenized under slow agitation for 1 hour to promote adsorption of PGA on cationic silica. The samples were then centrifuged during 30 minutes at 10000 rpm at 25°C. The equilibrium concentration of the polymer in the supernatant was measured by polyelectrolyte titration. The amount of adsorbed polymer, based on silica surface area was Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

575

determined from the concentration of excess polymer in the supernatant. Zeta potential and electrophoretic mobility The electrophoretic mobility of silica particle was measured initially, then after its surface modification, using a Zetasizer ZS (Malvern Instruments). The zeta potential (ζ) was then calculated from the electrophoretic mobility (µe) of the particles using Eq 1 (Henry’s law) (Henry 1931):

ζ=

3 ηµ e ⋅ 2 εε o f κ a

( )

[1]

where ε is the dielectric constant, εo the dielectric permittivity of vacuum, κ the Debye’s length, and a the particle radius. For f(κa), Oshima’s approximation was used (Oshima 1994):  2.5 1+ 2exp −κ a 1  f κ a = 1+ 1+ 2 κa 

(

( )

)

−3

 

[2]

In fact, the value of f(κa) is 1 when the electrolyte concentration is lower than 10-3 M, and 1.5 for all other cases. IR Spectroscopy Infrared spectra of modified and unmodified silica particles were carried out in carbon tetrachloride using a cell for solvent. Before analysis, the silica was dried at 100°C for 3 hours. The measurements were afterward made with a Perkin Elmer 2000 FT-IR, equipped with the software Spectrum v3.02. Carbon analysis Since the cationic coupling agent possesses carbon atoms, it is possible to determine the amount of carbon grafted on the silica surface. After reaction, the modified silica was washed with deionized water, to eliminate the excess of coupling agent. The silica was then dried at 100°C. The carbon content was estimated using a TOC analyser (Total Organic Carbon) equipped with a boat, which allow to make measurement on powders. Moreover, it is possible to calculate the ligand density of silanized silica; however the molecular weight of the ligand and the carbon content of the bonded silica need to be known. The ligand density (α) was calculated according to Eq 3 (Buszewski 1989):

α=

Pc ·10 6 1 · 1200nc − Pc M − nx as

(

)

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

Cationic polyelectrolyte adsorption The adsorption of cationic polyelectrolyte on silica surface of opposite charge is of the type strong affinity. The adsorption of p-DADMAC is considered as very fast and irreversible (Kekkonen, Stenius 1999; Hoogeveen et al. 1996). The amount of p-DADMAC adsorbed at silica surface saturation according to the pH and the ionic strength is shown in Fig 2. For the same ionic strength, the adsorbed amount at saturation increased with the pH of the solution. Indeed, the higher the pH of the solution, the more negative is the charge density of the silica surface because of the ionization of the silanol surface groups, and the stronger the interactions with a cationic polymer will be at pH 6.5 compared to trials at pH 4.5 (Böhmer et al. 1990). Under these conditions and because of the principle of electroneutrality, a more negative surface charge leads to higher cationic polymer adsorption because there is more charges to be neutralized on the surface. The adsorption of p-DADMAC on silica surface increased with increasing ionic strength. In fact the adsorption of polyelectrolyte on surface of opposite charge depends on the balance of opposing forces: the electrostatic attraction drives the adsorption process but at the same time limits it because the accumulation of charges on the surface is still unfavorable. At low ionic strength, the amount of cationic polymer adsorbed is low, because the mutual segmental repulsion of p-DADMAC prevents a strong coverage of the surface and the conformation of p-DADMAC on the silica surface is flat. With increasing salt concentration, the segment repulsion and the electrostatic attraction are screened by salt. The presence of salt promotes the charge accumulation because the range of electrostatic repulsion between the bound charges is reduced. At the same time, the presence of salt can screen the intramolecular repulsion, the polymer chain becomes more flexible and the polymer adsorbs on silica surface with loops and tails. In these conditions, p-DADMAC adsorption increases with increasing salt concentration. At very high ionic strength (0.1M), when the electrostatic contributions are totally screened, the adsorption is theoretically expected to decrease. However, the results showed higher adsorption of p-DADMAC on the silica surface when ionic strength increase. Of course, totally screened polymer charges can only adsorb on a surface if there are non-electrostatic

[3]

where α is expressed in µmol/m2, Pc is the measured carbon content, %C (w/w), nc is the number of carbon atoms of carbon in the bound silane, M is the molecular mass of the bound silane, nx the number of functional group substituents in the silane molecule and as the specific surface area. 576

Results and Discussion

Fig 2. Dependence of the amount of p-DADMAC adsorbed at silica surface saturation on the pH and the electrolyte concentration. (g pH 6.5, ● pH 4.5).

attractive interactions between the segments of the polymer and the surface (Vermöhlen et al. 2000). Indeed, several studies have clearly demonstrated that this is exactly the case observed here between the silica surface and the p-DADMAC, and more specifically with the quaternary amine group of the p-DADMAC (Hoogeveen et al. 1996; Rutland, Pashley 1989; Böhmer et al. 1995). The adsorption of p-DADMAC on the silica surface was typical of screening-enhanced adsorption (van de Steeg et al. 1992). At low ionic strength, the adsorption was low because the electrostatic contribution dominates. On the other hand, at very high ionic strength, the adsorption increased even if the electrostatic attraction was screened by electrolyte counterions because there was a non-electrostatic interaction which dominated. The zeta potential of silica covered by p-DADMAC at saturation is shown in Fig 3. It appears clearly that the adsorption of p-DADMAC on silica surface yielded positive zeta potential values. This is a good indication of the cationization of the silica surface through overcompensation. In fact, the overcompensation is due to the extension of a certain number of polymer segments in solution, represented especially by loops (Samoshima et al. 2003; Schönhoff 2003). The zeta potential at saturation varies according to the ionic strength of the solution. As shown in Fig 3 the zeta potential at saturation increased when the ionic strength increased from a low (0.001M) to a middle level (0.01M), due to the higher adsorbed amount, resulting from the formation of loops and tails of the polymer. However, at very high ionic strength (0.1M), the zeta potential at saturation decreased. This decrease of the zeta potential at high ionic strength (0.1M), compared to the value at 0.001M, is attributed to the screening of the charges of the adsorbed polymer by the electrolyte ions (Cl-) (Bauer et al. 1998; Rehmet, Killman 1999; Bauer et al. 1999). For the same ionic strength, the higher the amount of cationic polymer adsorbed, the higher the zeta potential of modified silica. Under these conditions, the zeta potential at saturation is more important at pH 6.5 rather than at pH 4.5, in spite of a more important surface charge density of the silica to be compensated. However, it is well established that the zeta potential of silica particle depends only on the charge density of the polyelectrolyte and do not depend on the surface charge densityoftheadsorbent (Böhmer et al. 1990; Rehmet, Killman 1999).

Fig 3. Zeta potential of silica covered by p-DADMAC at saturation according to pH and electrolyte concentration. ( g pH 6.5, ● pH 4.5).

Quaternization of silica nanoparticles Qualitative analysis by IR Spectroscopy From a qualitative point of view, the IR spectroscopy allows to determine the presence of the coupling agent on the silica surface, by the appearance of new peaks during the analysis of the modified silica. Fig 4 presents the IR spectra of the silica nanoparticle before and after treatment. The infrared spectrum of the silica is characterized by a sharp absorption band at 3747 cm-1 with a broad tail to low wavelength having a maximum near 3550 cm-1. The sharp peak is due to the OH stretching vibration of isolated non-hydrogen-bonded SiOH or Si(OH)2 groups and the broad tail having a peak near 3550 cm-1 is due to these groups when they are H-bonded. The weak broad features between 2000 and 1300 cm-1 are due to overtones and combination modes of bulk SiO2 vibrations and the region of total absorption between 1300-1000 cm-1 and from 850-750 cm-1 are due to absorption of IR radiation by bulk modes of SiO2 (Morrow, Gay 2000). Much of the earlier IR work which was related to the adsorption of molecules on silica was concerned with studying the disappearance of features associated with SiOH stretching vibrations, and if applicable, looking at the appearance of new features due to adsorbed species. This is relatively straightforward if the adsorbed moiety contains a functional group such as one containing a hydrogen atom. Therefore, adsorbed species which contain CHx or NHx functionalities are easily detected because their CH or NH stretching and deformation vibrations lie above 1300 cm-1. To decrease scattering by the pressed silica gel plate in potassium bromide (KBr), the spectra of samples were determined in carbon tetrachloride (CCl4). In carbon tetrachloride, the band of free silanol groups at 3750 cm-1 of untreated silica shifts by 58 cm-1 to a lower frequency and in the spectrum there is a broad band at 3700-3000 cm-1

Fig 4. IR Spectra of silica before and after grafting reaction.

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

577

consisting of overlapping bands of free silanol groups perturbed by CCl4 and of surface bonded silanol groups (Davydov 2000). The various vibrations which can be observed, in the case of the modified silica are presented in Table 2. Table 2. Representative wavelength and bonding type of the grafted silica. Frequency cm-1

Vibration mode

Bonding

Si – O - C Si – CH2R Si – OH C–N CH2 – CH2

1100 - 1000 1250 – 1175 3700 – 3200 1250 – 1020 2950 – 2830

stretching rocking stretching stretching stretching

N+ – CH3

1500 2200- 1800

deformation stretching

Si – O -C Si – C O-H C-N C–H (alkane) C–H amine)

Groups

Thus, at the level of the spectrum of the modified silica, all the frequencies lower than 1300 cm-1 could not be discovered because of the total absorption of the IR radiations by the silica below this value. Only the vibrations due to the connections C-H are thus observed in the modified silica spectrum. Quantitative measurements of grafting The variations of the zeta potential and the ligand density grafted for various amount of silane coupling agent added during the synthesis are presented in Fig 5. The figure clearly shows that the addition of a cationic coupling agent yielded positive silica. The zeta potential and the ligand density increased with increasing amount of added silane but eventually leveled off at around 10% by weight (based on SiO2). Further addition of cationic coupling agent produced little effect on the zeta potential and on the ligand density. This change of properties of the surface can be explained only by the formation of chemical or covalent bond between the silica surface and the cationic silane coupling agent (IR spectroscopy). This significant change can be attributed only to the effect of the quaternization of silica nanoparticle. In fact, it is well accepted that the silica surface modification results from the condensation of the silanol groups of the silica with the hydroxyl groups, resulting from the hydrolysis of the alkoxy groups of the coupling agent (Plueddeman 1992). However, it is necessary to consider that all the silanol

groups are not condensed during the synthesis. We consider moreover that only half of the original silanol groups have reacted under otherwise optimum conditions (Buszewski 1989). With the silica used in this experiment, the silanol surface concentration αOH amounted to approximately 4 µmol/m2, and we obtained at best a ligand density (α) of 1.1 µmol/m2. It means that 27.5% of condensed silanol groups had reacted with the silane coupling agent. Fig 6 shows the variation of the zeta potential of the silica particles as a function of the reaction time at 10% by weight of silane coupling agent at pH 4.5. The variation of the zeta potential of modified silica with time appeared to be immediate in aqueous solution, with a maximum reached after 1 hour. The effects of reaction pH and temperature are presented in Table 3. At the same reaction temperature, the reaction at higher pH seemed to improve the quaternization of silica nanoparticles. Indeed, under alkaline pH conditions, the silanol groups lose more easily their proton, which favors the nucleophilic attack of the methoxy groups of the silane coupling agent (El Shafei 2000), thus allowing a better silanisation of silica and increasing its cationicity. In the same way, under the same pH, it seems that the grafting of the silane coupling agent on silica surface increased with increasing the reaction temperature due to a better condensation of the silane on the silica surface. The isoelectric point of silica before and after modification in the absence of salt, can be determined from the zeta potentials in a broad range of pH presented in Fig 7. Clearly, the isoelectric point of the modified silica was reached at around pH 8.9, while that of the unmodified silica was achieved between pH 3 and 4. This difference is naturally explained by the fact that the modified silica possesses a strong and positive zeta potential. Indeed, under acidic conditions, the ions H+ surround the quaternized groups on the silica surface and the zeta potential is high. On the other hand, under more alkaline conditions,

Fig 6. Zeta potential of cationic silica as a function of the reaction time. Table 3. Temperature and pH effects for silane-modified silica particles for a reaction time of 1h and a silane content of 20% by weight.

Fig. 5. Effect of reactants weight ratio on zeta potential ( g ) and grafted ligands density (o) on the cationic silica at pH and after a reaction time of 1 hour.

578

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

pH

Temperature· (°C)

α (µmol/m2)

Zeta potential (mV)

4.5 4.5 9 9

25 80 25 80

0.743 0.767 1.053 1.095

40.16 42.83 42.55 43.55

Fig 7. Variation of zeta potential of silica before ( ) and after surface modification ( g ) with pH, without salt.

the presence of OH- produces a decrease of the zeta potential because of the neutralization of the quaternary groups. Naturally, since cationic silica has a higher charge density after quaternization, a larger amount of OH- must be attracted toward the cationic surface groups of silica to reach the isoelectric point at a high pH. The zeta potential of silica before and after modification according to the pH and the ionic strength is presented in Fig 8. The zeta potential of the unmodified silica was more negative at pH 6.5 than at pH 4.5. This is due to a more extensive ionization of the silanol surface groups at alkaline pH. Whatever was the pH, the zeta potential decreased towards zero when the ionic strength of the solution increased because of the adsorption of the electrolyte counterions (Na+). In fact, after grafting of a certain number of cationic groups, the pH modification influences the zeta potential of modified silica nanoparticle. Indeed, the more alkaline the pH, the higher the number of quaternary ammonium groups neutralized. Under these conditions the zeta potential of modified silica was more important at pH 4.5 than at pH 6.5. In the same way as the unmodified silica, the zeta potential of grafted silica decreased with increasing the ionic strength, because of the screening of the cationic surface charges of modified silica by electrolyte counterions (Cl ).

negatively charged species at low electrolyte concentration. At the same time, adsorption can only occur if the electrostatic repulsions are totally screened and if there is a chemical surface affinity that is a non-electrostatic interaction between the silica surface and the PGA. Therefore, it clearly appears that for the salt concentration range in this study, the electrostatic interaction between the silica particle and the PGA dominated. Even if there was a non-electrostatic interaction between these two species, it did not have a major effect in this study. These results are in good agreements with other studies, where no adsorption of PGA onto negative substrate was observed (Kekkonen, Stenius 1999; Hoogeveen et al. 1996). The adsorption of the polygalacturonic acid on the cationic silica surface, at saturation is shown in Fig 9 as a function of the ionic strength and the pH of the solution. The zeta potential of modified silica is also shown. The adsorption of polygalacturonic acid on the cationic silica surface ws relatively different according to the nature of the modification carried out on the silica surface. For the silica modified by grafting, the fixation of the polygalacturonic acid was more important at pH 4.5 than at pH 6.5. It increased with increasing ionic strength of the solution and did not depend on the zeta potential of cationic silica. On the other hand, the fixation of polygalacturonic acid on the silica surface covered by p-DADMAC was more important at alkaline pH. It went through an optimum according to the electrolyte concentration, and varied especially according to the zeta potential of the silica particles covered with p-DADMAC. It clearly seems that the fixation of the polygalacturonic acid on the cationic silica surface varies strongly according to the nature of the modification. Indeed, the fixation of polygalacturonic acid will be considered as an adsorption on a hard

Fixation of polygalacturonic acid on particles Of course, before using cationic silica particles to promote the fixation and the elimination of polygalacturonic acid from white water, some experiments were directly made with unmodified silica particle. In fact, no deposition of PGA on silica particle was observed at either pH 4.5 or pH 6.5 for NaCl concentrations lower than 1M. According to the DLVO theory, one expects strong repulsion between

Fig 8. The dependence of the zeta potential of silica before and after modification on pH and salt concentration. Parameters: g unmodified pH 4.5, c unmodified pH 6.5, modified pH 4.5, ● modified pH 6.5.

°

Fig 9. Variation of the adsorption of PGA on cationic silica at pH 4.5 (A) and pH 6.5 (B) according to the zeta potential and the ionic strength. Parameters: g Amount of PGA adsorbed by grafted silica, c Amount of PGA adsorbed by silica pretreated by p-DADMAC, ● Zeta potential of grafted silica, Zeta potential of silica pretreated by p-DADMAC.

°

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

579

surface in the case of grafting, while it will be considered as an adsorption on a soft surface in the case of the pretreatment of silica particles with cationic polymers. The adsorption of PGA, which is an anionic polymer, on the modified silica surface was similar to the adsorption of p-DADMAC on the unmodified silica surface. In general, the adsorption of a polyelectrolyte on a surface of opposite charge depends strongly on the surface charge density of the particle, on the charge density of the polymer, on its molecular properties, such as the conformation and the molecular weight, and on the ionic strength of the solution (Fleer et al 1993). The surface charge density of the particles and the ionic strength of the solution were the only parameters taken into account in this study. At the same ionic strength, the surface charge density and the zeta potential of cationic silica were higher at pH 4.5 than at pH 6.5, as presented in Fig 8. The modified silica being more positive at pH 4.5, the interactions with the negative groups of the PGA were thus more important. Due to the principle of electroneutrality, the adsorption of PGA on cationic grafted silica was stronger at lower pH since there was more charge of the cationic silica to be neutralized by the negative charge of the PGA. Concerning the fixation of polygalacturonic acid on grafted silica surface according to the ionic strength, we consider that the adsorption is typical of screening-enhanced adsorption. In fact, the adsorption of PGA increased with increasing salt concentration. This was due to the dominating effect of the non-electrostatic attraction between the PGA and the modified silica surface at high ionic strength. Of course, at low ionic strength, since the adsorption is charge limited, the amount of PGA adsorbed was low because the accumulation of negative charge of the PGA on the surface was still unfavourable. On the other hand, at high ionic strength, not only the segment-segment repulsions of PGA in the adsorbed layers, but also the attractive interaction between the PGA segments and the cationic site of the modified silica are screened by salt counterions. In fact, the presence of salt allows especially to screen and decrease the electrostatic repulsion between polymer segments of similar charges. These rather behave as uncharged polymers, where they can have a conformation with loops and tails, resulting in an increase of the adsorbed amount. It becomes evident that the fixation of PGA on the cationic silica surface was not exclusively governed by an electrostatic attraction, since the adsorption increased while the zeta potential of cationic silica decreased with increasing the ionic strength (Fig 9). Moreover, since non-electrostatic interaction is not involved between the PGA and the unmodified silica particle, we can suppose that the non-electrostatic contribution between the PGA and the grafted silica particle was only due to the presence of the grafting coupling agent on the silica surface. The fixation of polygalacturonic acid on the silica particles pretreated with p-DADMAC is considered as an adsorption on a soft surface of opposite charge. This adsorption differs significantly from the adsorption on a hard surface of the same surface charge density. This is not only due to the surface charge being distributed in the 580

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

z-direction (i.e. normal direction to the surface), but also due to the flexibility of the underlying layer (Schönhoff 2003). The adsorption is also favored by the fact that the segments of adsorbing chains are entangled into the preadsorbed layers. As shown in other studies, this adsorption is governed by electrostatic interactions (Kekkonen et al. 2001). Moreover, it clearly seems, by observing Fig 9 that the adsorption of PGA on silica surface pretreated by p-DADMAC was directly dependent on the zeta potential of the modified silica particle (Fig 3) but did not depend on the amount of p-DADMAC pre-adsorbed (Fig 2). Indeed, when the ionic strength increased, the quantity of p-DADMAC adsorbed increased too, whereas the quantity of PGA fixed to the cationic silica surface vary according to the same profile as the zeta potential of cationic silica; i.e. they both went through a maximum at 0.001M and then decreased at higher electrolyte concentrations. In fact, the process of PGA fixation requires the preliminary presence of loops of p-DADMAC, which are anchored to the silica surface and dangling into the solution. These loops create a positive layer which allows the fixation of PGA by the complexation of its negative charges with the positive charges of p-DADMAC (Joanny, Castelnovo 2002). Thus, the fixation of polygalacturonic acid by complexation on cationic silica surface pretreated by p-DADMAC depends on the amount of cationic charge available on cationic particles. This number of cationic charges present on silica surface and available for the complexation of anionic charge of PGA varies according to the total amount of cationic charge of p-DADMAC adsorbed and the concentration of the electrolyte counterions, which can screen some of these cationic charges. Under these conditions, the adsorption of PGA increased with increasing the ionic strength from 0.001M to 0.01M because in this range the amount of cationic charges available on the surface was higher. Indeed, the amount of pDADMAC adsorbed increased while their cationic charges were not screened by the low amount of electrolyte counterions. This promoted the fixation of PGA since the zeta potential of cationic silica was greater. On the other hand beyond an ionic strength of 0.01M, the increase of the salt concentration leads to the screening of the cationic charge adsorbed on silica surface by the electrolyte counterions. While the total amount of adsorbed cationic charge on silica surface increased with increasing the amount of p-DADMAC adsorbed, the quantity of available cationic charge on silica surface decreased. This lead to the decrease of zeta potential and thus to the decrease of the amount of PGA fixed on cationic silica. However, the decrease of the amount of PGA adsorbed could also be explained by the fact that in the presence of salt, the degree of interdigitation of PGA segments appears to be somewhat lower. This can be explained by the adsorption of a coil-shaped structure in contrast to rod-like chains in the absence of salt, which implies a reduced accessibility of the outer layer segments of p-DADMAC to the next adsorbing layer chains of PGA (Schönhoff 2003). Finally, for the same ionic strength, the adsorption of PGA on the cationic silica surface was more important at pH 6.5 than at pH 4.5 because the zeta potential of silica

covered by p-DADMAC was more important. Since the quantity of cationic charge adsorbed in surface was higher at alkaline pH, the complexation of anionic charge of PGA was greater.

Conclusion The purpose of this study was to understand how modified cationic silica can affect the fixation of a model substance of “anionic trash” on their surface, according to the type of modification realized on the particle. Pre-treatment of p-DADMAC on silica surface depends on the surface charge density of the particle and on the ionic strength of the solution. Adsorption increases with increasing pH conditions, since the silica is more negatively charged at alkaline pH due to the ionization of its silanol surface groups. The adsorption of p-DADMAC on the silica surface is dominated by a nonelectrostatic interaction at high salt concentration. Indeed the adsorption increases with increasing salt concentration of the solution. This adsorption is typical of screening-enhanced adsorption, and the presence of salt allows especially to screen the segment-segment repulsions of PGA in the adsorbed layer, thus enhancing the adsorption. The fixation of polygalacturonic acid on cationic silica surface depends strongly on the type of modification. The fixation of PGA on silica particles pretreated by p-DADMAC is similar to adsorption on a soft charged surface. This adsorption is only governed by electrostatic interactions. It varies in the same way as the profile of the zeta potential of cationic silica covered by p-DADMAC. In fact this adsorption depends on the number of adsorbed positives charges available on the particle surface for the complexation of the anionic segments of PGA. The amount of cationic charge available depends on the total number of adsorbed p-DADMAC cationic charge and on the proportion of these charges which are screened by the electrolyte counterions. On the other hand, the fixation of polygalacturonic acid on grafted cationic silica is similar to an adsorption on a hard charged surface. Thus, the adsorption of PGA on grafted cationic silica particle is above all, governed by a non-electrostatic attraction which dominates at high ionic strength. It is also a screening-enhanced adsorption, because the final amount of PGA fixed increases with increasing ionic strength. In spite of the screening of the electrostatic attraction between the cationic grafted silica and the negative PGA, the amount of PGA fixed on grafted silica particle increases. Acknowledgement The authors gratefully acknowledge the Canada Research Chair in Value-added Paper for their financial support.

Literature Bauer, B., Killmann, E. and Jaeger, W. (1998): Flocculation and Stabilization of Colloidal Silica by the Adsorption of poly-diallyl-dimethyl-ammoniumchloride

(PDADMAC) and Copolymers of DADMAC with N-methyl-N-vinyl-acetamide (NMVA), Colloid Polym. Sci. 276, 698. Bauer, D., Buchhammer, H., Fuchs, A., Jaeger, W., Killmann, E., Lunkwitz, K., Rehmet, R. and Scwharz, S. (1999): Stability of Colloidal Silica, Sikron and Polystyrene Latex Influenced by the Adsorption of Polycations of Different Charge Density, Colloids Surf. A 156, 291. Buszewski, B. (1989): Effect of Destruction of Chemically Bonded Reversed Phases by Hexamethyldisilazane during End-Caping, Chromatographia 28(1112), 574. Böhmer, M.R., Evers, O.A. and Scheutjens, J.M.H.M. (1990): Weak Polyelectrolytes between Two Surfaces: Adsorption and Stabilization, Macromolecules 23, 2288. Böhmer, M.R., Hesterbeek, W.H.A., Deratani, A. and Renard, E. (1995): Adsorption of Partially Quaternised Poly (Vinyl Imidazoles) onto SiO2 and Y2Os, Colloids Surf. A 99, 53. Chen, J., Hubbe, M.A., and Heitmann, J.A. (2001), Proceedings of the TAPPI Papermakers Conference, March 2001, CD doc. Chen, J., Heitmann, J.A. and Hubbe, M.A., (2003): Dependency of Polyelectrolyte Complex Stoichiometry on the Order of Addition. 1. Effect of Salt Concentration During Streaming Current Titrations with Strong Poly-Acid and Poly-Base , Colloids Surf. A 223, 215. Davydov, V.Ya. (2000): Chp.4 Adsorption on Silica Surfaces, Ed. by Eugène Papirer, Adsorption on Silica Surfaces, Marcel Dekker Inc., pp 63-118. Dunham, A.J., Sherman, L.M. and Alfano, J.C. (2002): Effect of Dissolved and Colloidal Substances on Drainage Properties of Mechanical Pulp Suspensions, J. Pulp Paper Sci. 28(9), 298. El Shafei, G.M.S. (2000): Chp.3 Silica Surface Chemical Properties, Ed. By Eugène Papirer, Adsorption on Silica Surfaces, Marcel Dekker Inc., pp 35-62. Fleer, G.J., Cohen Stuart, M.A., Scheutjens, J.M.H.M., Cosgrove, T. and Vincent, B. (1993): Polymers at Interfaces, Chapman & Hall, London. Gill, R.I.S. (1996): Chemical Control of Deposits-Scopes and Limitations, Paper Technology & Industry, 37(6), 23. Henry, D.C. (1931): Proc. Roy. Soc. Ser. A 133, 106. Hoogeveen, N.G., Cohen Stuart, M.A. and Fleer, G.J. (1996): Polyelectrolyte Adsorption on Oxides I: Kinetics and Adsorbed Amount, J. Colloid Interface Sci. 182, 146. Joanny, J.-F. and Castelnovo, M. (2002): Polyelectrolyte Adsorption and Multilayer Formation, Ed. by G. Decher & J.B. Schlenoff, Multilayer Thin Films, Wiley. Kam, S. and Gregory, J. (1999): Charge Determination of Synthetic Cationic Polyelectrolytes by Colloid Titration, Colloids Surf. A 159, 165. Kekkonen, J. and Stenius, P. (1999): The Effect of Short-Chain Cationic Polymers on the Deposition of Wood Resin Emulsion Droplets on Silica Surfaces, Colloids Surf. A 156, 357. Kekkonen, J., and Stenius, P. (1999): Interactions of Oxyde Surfaces with Dissolved and Colloidal Substances Released in Mechanical Pulping, Nord. Pulp Pap. Res. J. 14(4), 300. Kekkonen, J., Laukkanen, A., Stenius, P. and Tenhu, H. (2001): Adsorption of Polymeric Additives and their Effect on the Deposition of Wood Materials in Paper Production, Colloids Surf. A 190, 305. Linhart, F., Auhorn, W.J., Degen, H.J. and Lorz, R. (1987): Anionic Trash: Controlling Detrimental Substances, Tappi J. 70(9), 79. Morrow, B.A. and Gay, I.D. (2000): Chp.2 Infrared and NMR Characterization of the Silica Surface, Ed. by Eugène Papirer, Adsorption on Silica Surfaces, Marcel Dekker Inc., pp 9-33. Oshima, H. (1994): A Simple Expression for Henry’s Function for the Retardation Effect in Electrophoresis of Spherical Colloidal Particles, J. Colloid Interface Sci. 168, 269. Pelton, R.H., Allen, L.H. and Nugent, H.M. (1980): A Survey of Potential Retention Aids for Newsprint Manufacture, Pulp Paper Can. 81(1), 54. Plueddemann, E.P. (1992): Silane Coupling Agents, Plenum Press, New York. Rehmet, R. and Killmann, E. (1999): Adsorption of Cationic Poly(diallyl-dimethyl-ammoniumchloride), poly(diallyl-dimethyl-ammoniumchloride-co-N-methylNvinylactamide) and poly(N-methyl-N-vinyl-acétamide) on Polystyrene Latex, Colloids Surf. A 149, 323. Rutland, M.W. and Pashley, R.M. (1989): The Charging Properties of Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

581

Monodisperse Colloidal Silica in Symmetrical Quaternary Ammonium Ions Solutions, J. Colloid Interface Sci. 130(2), 448. Samoshima, Y., Diaz, A., Becker, Y., Nylander, T. and Lindman, B. (2003): Adsorption of Cationic, Anionic and Hydrophobically Modified Polyacrylamides on Silica Surfaces, Colloids Surf. A 231, 195. Schönhoff, M. (2003): Layered Polyelectrolyte Complex: Physics of Formation and Molecular Properties, J. Phys. Condens. Matter 15, 1781. Sundberg, K., Thornton, J., Pettersson, C., Holmbom, B. and Ekerman, R. (1994): Calcium-Induced Aggregation of Dissolved and Colloidal Substances in Mechanical Pulp Suspensions, J. Pulp Paper Sci. 20(11), 317. Tay, C.H. (1980): Application of Polymeric Flocculant in Newsprint Stock Systems for Fines Retention Improvment, Tappi J. 63(6), 64. Terayama, H. (1952): Method of Colloid Titration (A New Titration Method between Polymer Ions), J. Polym. Sci. 8(2), 243. Thornton, J., Ekman, R., Holmbom, B. and Eckerman, C. (1993): Release of Potential “Anionic Trash” in Peroxide Bleaching of Mechanical Pulp, Paperi Puu 75(6), 426. Thornton, J., Ekman, R., Holmbom, B. and Örså, F. (1994): Polysaccharides Dissolved from Norway Spruce in Thermomechanical Pulping and Peroxyde Bleaching, J. Wood Chem. Technol. 14, 159.

Toei, K. and Kohara, T. (1976): A Condumetric Method for Colloid Titrations, Anal. Chim. Acta 83, 59. van de Steeg, H.G.M., Cohen Stuart, M.A., de Keizer, A. and Bijsterbosch, B.H. (1992): Polyelectrolyte Adsorption : a Subtle Balance of Forces, Langmuir 8, 2538. Wearing, J.T., Barbe, M.C. and Ouchi M.D. (1985): The Effect of White Water Contaminate on Newsprint Properties, J. Pulp Paper Sci. 11(4), 113. Vermöhlen, K., Lewandowski, H., Narres, H.-D. and Schwuger, M.J. (2000): Adsorption of Polyelectrolytes onto Oxydes-The Influence of Ionic Strength, Molar Mass and Ca2+ Ions, Colloids Surf. A 163, 45. White, G.W., Katona, T. and Zodda, J.P. (1999): The Use of High-Performance Size Exclusion Chromatography (HPSEC) as a Molecular Weight Screening Technique for Polygalacturonic Acid for Use in Pharmaceutical Applications, Journal of Pharmaceutical and Biomedical Analysis 20, 905. Wågberg, L. and Asell, I. (1995): The Action of Cationic Polymers in the Fixation of Dissolved and Colloidal Substances, Colloids Surf. A. 104,109. Zippel, F. (2001): Water Management in Paper Mills, Ed. by Zippel F., Heidenheim

Manuscript received May 19, 2006 Accepted june 21, 2006

Nordic Pulp and Paper Research Journal Vol 21 no. 5/2006

582