`
`www.elsevier.com/locate/jcis
`
`Temperature influence of nonionic polyethylene oxide and anionic
`polyacrylamide on flocculation and dewatering behavior
`of kaolinite dispersions
`Patience Mpofu, Jonas Addai-Mensah,∗ and John Ralston
`
`Ian Wark Research Institute, University of South Australia, Mawson Lakes, Adelaide, SA 5095, Australia
`Received 11 May 2003; accepted 23 September 2003
`
`Abstract
`Nonionic polyethylene oxide (PEO) and anionic polyacrylamide (PAM) flocculation of kaolinite dispersions has been investigated at pH 7.5
`in the temperature range 20–60 ◦C. The surface chemistry (zeta potential), particle interactions (shear yield stress), and dewatering behavior
`were also examined. An increase in the magnitude of zeta potential of kaolinite particles, in the absence of flocculant and at a fixed PEO and
`PAM concentration, with increasing temperature was observed. The zeta potential behavior of the flocculated particles indicated a decrease
`in the adsorbed polymer layer thickness, while at the same time, however, the adsorbed polymer density showed a significant increase with
`increasing temperature. These results suggest that polymer adsorption was accompanied by temperature-influenced conformation changes.
`The hydrodynamic diameter and supernatant solution viscosity of both polymers decreased with increasing temperature, consistent with a
`change in polymer–solvent interactions and conformation, prior to adsorption. The analysis of the free energy ( Gads) of adsorption showed
`a strong temperature dependence and the adsorption process to be more entropically than enthalpically driven. The polymer conformation
`change and increased negative charge at the kaolinite particle surface with increasing temperature resulted in decreased polymer bridging and
`flocculation performance. Consequently, the shear yield stress and the rate and the extent of dewatering (consolidation) of the pulp decreased
`significantly at higher temperatures (>40 ◦C). The temperature effect was more pronounced in the presence of PEO than PAM, with 40
`and 20 ◦C indicated as the optima for enhanced performance of the latter and former flocculants, respectively. The results demonstrate that
`a temperature-induced conformation change, together with polymer structure type, plays an important role in flocculation and dewatering
`behavior of kaolinite dispersions.
`© 2003 Elsevier Inc. All rights reserved.
`
`Keywords: Temperature; Flocculation; Dewatering; Adsorption; Polymer conformation
`
`1. Introduction
`
`Flocculant-assisted thickening processes are commonly
`used in the minerals industry for the dewatering of kaolin-
`ite and other clay mineral waste tailings, but these are far
`from being efficient. High settling rates (e.g., 1–10 m/h)
`are typically achieved for kaolinite and mixed mineral ox-
`ide tails with the aid of polyelectrolytes. However, this good
`settling behavior is invariably accompanied by space-filling
`flocs and thickener underflow tails of characteristically low
`solid loadings (e.g., 20–30 wt%) [1–3].
`
`* Corresponding author.
`E-mail address: jonas.addai-mensah@unisa.edu.au
`(J. Addai-Mensah).
`
`0021-9797/$ – see front matter © 2003 Elsevier Inc. All rights reserved.
`doi:10.1016/j.jcis.2003.09.042
`
`Several studies have shown that conventional polymeric
`flocculants such as high-molecular-weight nonionic and
`ionic polyacrylamides (PAM) and, to a limited extent, un-
`conventional, nonionic polyethylene oxide (PEO) may be
`used to readily dewater kaolinite and other clay mineral
`tailings [1,3–9]. Flocculation with PAM, however, leads to
`space-filling flocs with low-compaction behavior. The fail-
`ure to achieve a high solid density during thickening of
`kaolinite dispersions may be attributed to several physical
`and chemical factors that affect the conformation and ad-
`sorption of polymeric flocculants, hence the flocculation and
`dewatering behavior [10–15]. These factors include floccu-
`lant characteristics such as molecular weight, charge density,
`functionality, dosage, and dilution and slurry properties such
`as pH, ionic strength, temperature, the presence of simple
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`
`
`146
`
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`
`and hydrolyzable metal ions, particle surface area, size, zeta
`potential, and solid concentration [15–21].
`An effective flocculant for a given dewatering application
`must be of the right molecular weight and charge density
`and contain functional groups that are predisposed to inter-
`act favorably with specific sites on the particle surface, for
`given dispersion conditions, including temperature. Further-
`more, it must also have an extended and flexible (elastic)
`configuration in the solution to achieve better particle bridg-
`ing and to produce flocs capable of withstanding moderate
`shear forces without rupturing [1,8,19,21–23]. The use of
`shear-sensitive polymers (e.g., high-molecular-weight PEO)
`has been of considerable interest in recent years due to
`their ability to produce thickened tails of solid loadings
`significantly greater than those achieved with conventional
`flocculants (e.g., PAM), under moderate shear conditions
`[11,12,23–26].
`PEO is perceived as a “flexible” polymer, with a structure
`described as a random coil and can change conformation
`dynamically in solution [27,28]. The size of the coil is de-
`pendent upon the solvent quality, which in turn is dependent
`on temperature, concentration, and molecular architecture of
`the polymer molecule. The initial contact between particles
`occurs through loops and tails which suggest that the con-
`figurations of the polymer prior to and after adsorption are
`important. The number and size of these polymer loops and
`tails will determine the flocculation efficiency and, hence,
`the settling rates and consolidation of kaolinite dispersions.
`While the influence of physicochemical factors such as poly-
`mer structure type, molecular weight, charge density, dilu-
`tion, solution pH, and mechanical effects on flocculation and
`dewatering are well researched and documented, the knowl-
`edge available in the literature on the influence of tempera-
`ture is very limited [29,30].
`Several of the reported studies [31–38] on the influence
`of temperature on flocculation and dewatering process indi-
`cate mixed or conflicting results. According to some studies
`[31–35], enhanced flocculation of mineral dispersions such
`as kaolinite, alumina, and TiO2 occurred upon increasing
`temperature using PEO and PAM flocculants. Binner and
`Murfin [36], who studied the flocculation and dewatering of
`alumina particles with polyacrylate, observed an increase in
`settling rates with increasing temperature up to 45 ◦C, which
`then decreased with a further increase of temperature. On
`the other hand, the studies by Mohtadi and Rao [37] showed
`that temperature did not have any noticeable effect on the
`flocculation of kaolinite and smectite dispersions using non-
`ionic polymers and polyelectrolytes in the temperature range
`1–20 ◦C. A noticeable decrease of particle zeta potential
`with decreasing temperature was, however, observed.
`Temperature is, thus, a process variable of central impor-
`tance in the dewatering of mineral dispersions, as it can have
`a profound impact on the interactions between both mineral
`particles and polymer in aqueous media. To date, it appears
`that no conclusive information on how temperature affects
`the conformation and adsorption of a nonionic PEO or an-
`
`ionic PAM, particle zeta potentials, and the flocculation and
`dewaterability of kaolinite particles has been reported. The
`main aim of this work, therefore, was to investigate the ef-
`fect of temperature on the conformation and adsorption of
`nonionic PEO and anionic PAM flocculants, flocculation of
`colloidally stable kaolinite dispersions and how this impacts
`on surface chemistry, shear yield stress, and dewatering be-
`havior. The temperature range studied was 20–60 ◦C as this
`is the range mostly encountered in clay mineral tailings
`treatment in industry. The investigations were performed at
`a constant pH of 7.5 and flocculant concentration range of
`0–1000 g/ton kaolinite solid.
`
`2. Materials and methods
`
`2.1. Materials
`
`Colloidal-size kaolinite particles (K15GM, 99% pure,
`quartz and mica 1%, Commercial Minerals, Australia) were
`used in this work. The particle density and BET [39] surface
`area were 2.60 kg dm−3 and 24.8 m2 g−1, respectively. The
`median particle diameter of kaolinite particles was found to
`be 2.80 μm by laser diffraction. A Na-acrylate, carboxyl-
`substituted polyacrylamide copolymer flocculant (30 mol%
`anionically charged, Nalco, Australia) with an average mole-
`cular weight of 2.7 × 106 was used. Nonionic polyethylene
`oxide (SNF, Australia) with an average molecular weight of
`2.5 × 106 was also used. The chemical structures of the two
`polymers are shown in Fig. 1. Average polymer particle size
`in solution determined at pH 7.5 and 20 ◦C by dynamic light
`scattering analysis as hydrodynamic diameter was found to
`be ∼ 326 and 449 nm for PEO and PAM, respectively. Fresh
`0.1 wt% polymer working solutions used were prepared
`daily by diluting a previously prepared 1 wt% stock solu-
`tion stored at room temperature. Analytical grade KNO3 was
`used to increase solution ionic strength, while high-purity
`potassium hydroxide and nitric acid (BDH, Australia) were
`used to control the dispersion pH. All solutions and disper-
`sions were prepared with Milli-Q water (surface tension =
`72.8 mN m−1, specific conductivity < 0.5 μS cm−1 at pH 5.6
`and 20 ◦C).
`
`(a)
`
`(b)
`
`(a) and
`Fig. 1. Structure of a nonionic polyethylene oxide (PEO)
`a Na-acrylate carboxyl-substituted anionic polyacrylamide copolymer
`(PAM) (b).
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`
`
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`
`147
`
`2.2. Suspension preparation
`
`2.5. Zeta potential measurements
`
`Particle zeta potential (ζ ) was determined from dynamic
`mobility measurements made with an acoustosizer (Col-
`loidal Dynamics Inc., Australia). The measurements were
`carried out on agitated suspensions (8% w/w) of kaolinite
`particles dispersed in 10−2 M KNO3. The 10−2 M KNO3
`was used to avoid the effect of anomalous surface conduc-
`tance on the zeta potential, which can be present in kaolin-
`ite clay systems at very low electrolyte concentrations [42].
`A thermostatically controlled water bath connected to the
`conditioning vessel’s water jacket was used to maintain
`a constant temperature. Calibrations using K4(SiW12O40)
`were performed once a day, prior to zeta potential analysis.
`Depending on the state or flow behavior (e.g., low or high
`yield stress) of the flocculated dispersion, agitation rates be-
`tween 500 and 2000 rpm were used. The conductivity, tem-
`perature, and pH were continuously monitored in situ using
`probes attached to the instrument.
`
`2.6. Estimation of adsorbed polymer layer thickness
`
`Kaolinite dispersions of 8 wt% solid were prepared at
`pH 7.5 by thoroughly mixing a known mass of kaolinite par-
`ticles with a known mass of 10−3 M KNO3 solution at an
`agitation speed of 250 rpm for 1 h. For each test, 500 cm3 of
`fresh well-mixed slurry, equilibrated at 20–60 ◦C for 30 min,
`was transferred to a 500 cm3 graduated glass cylinder for
`flocculant addition.
`
`2.3. Dewatering tests
`
`To perform flocculation and settling tests, a known vol-
`ume of 0.1 wt% PEO or PAM polymer solution was added
`in a single step to an 8 wt% solid slurry using a syringe and
`mixed by moving a perforated disc plunger up and down
`eight times to ensure that the suspension was well homoge-
`nized and the dispersion level was visually checked. All tests
`were carried out at pH 7.5 and isothermally at temperatures
`in the range 20.0–60.0 ± 0.1 ◦C, using a thermostatically
`controlled water bath. The initial settling rates of the floccu-
`lated suspension was determined by recording the time taken
`for the “mud line” (solid–liquid interface) in the 500 cm3
`cylinder to pass between the 450 cm3 and 350 cm3 marks
`(over a distance of ∼ 5.5 cm). The flocculated suspension
`was allowed to stand quiescently for a period of 24 h and
`the consolidated bed volume recorded. After settling, the
`presedimented slurries were agitated using a four-blade im-
`peller for 5 min at 250 rpm for further consolidation. The
`bed height was measured again after 1 h to investigate the
`effect of shear on final sediment bed consolidation.
`
`2.4. Polymer adsorption experiments
`
`A batch depletion method was used for determining poly-
`mer adsorption isotherms at different temperatures. The ad-
`sorption measurements were conducted in 20 cm3 vials by
`adding a known concentration of the flocculant to 20 cm3
`of kaolinite dispersion of 8 wt% solid equilibrated either
`at 20, 40, or 60 ◦C. The suspension was agitated by a ther-
`mostatically controlled reciprocating water bath (Paton In-
`dustries, Australia, Model RW 1812) for 24 h for equilib-
`rium to be reached. A 10 cm3 aliquot of the dispersion was
`then removed and centrifuged at 5000 rpm for 5 min. The
`PAM solution concentration was determined by complex-
`ation using the starch–triiodide method [40]. The method
`involves bromine oxidation of the amide functional groups
`with the removal of excess bromine with sodium formate.
`The amide oxidation product converts iodide ion to iodine
`and is measured as a starch–triiodide complex, determined
`using a Cary 5 UV–vis spectrophotometer (Varian Instru-
`ments) operating at a wavelength of 600 nm. The residual
`PEO concentration in the supernatant was determined using
`the tannic acid method [41].
`
`When a nonionic polymer is adsorbed on a particle sur-
`face, a displacement of the shear plane takes place with
`respect to its position in the absence of the adsorbed poly-
`mer. This displacement is dependent upon the thickness of
`the adsorbed polymer layer [43]. The decrease in zeta po-
`tential may be related to the shift in the shear plane, which
`corresponds to the hydrodynamic thickness of the adsorbed
`polymer layer (δ) and is calculated for nonionic polymers
`(e.g., PEO) using the expression [43]
`= tanh
`
`tanh
`
`,
`
`zeζ2
`zeζ1
`4kT eκ( −δ)
`4kT
`where ζ1 and ζ2 are the particle zeta potentials in the pres-
`ence and absence of polymer respectively, z is valency of
`ions in the double layer, δ is the adsorbed layer thickness,
`k is the Boltzmann constant, T is the absolute tempera-
`ture, 1/κ is the Debye length, e is the electron charge, and
` is the thickness of the stern plane. Fleer [44] and Fleer
`et al. [45] estimated a value of 0.4 nm for at an ionic
`strength of 0.001 M and a similar approach has been adopted
`in the present calculation, as the ionic strengths are compa-
`rable. This method is satisfactory when measurements are
`performed in indifferent electrolyte solutions. Several stud-
`ies [44–52] have used this approach to determine the layer
`thickness of nonionic homo- and copolymers adsorbed at
`solid–liquid and liquid–liquid interfaces. It is assumed that
`the shear plane displacement is the only effect the adsorbed
`polymer has on the double layer properties and that the ion
`distribution in the double layer does not alter upon polymer
`adsorption. As this approach is unsuitable for adsorbed poly-
`electrolytes, only the adsorbed layer thickness of PEO was
`determined in the present work using Eq. (1).
`
`(1)
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`
`
`148
`
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`
`2.7. Rheological measurements
`
`The rheological (shear yield stress) measurements, which
`provided an indication of the particle interactions in kaoli-
`nite dispersions, were performed with respect to slurry in-
`terfacial chemistry. The yield stress was directly measured
`using the vane technique [53,54]. The advantage of the vane
`technique is that it eliminates slip, usually encountered with
`the cuvette concentric cylinder technique when dealing with
`highly structured dispersions. The 8 wt% solid dispersions
`used in the zeta potential and settling tests were too dilute
`to give meaningful rheological data; hence only the floc-
`culated kaolinite dispersion of 28 wt% solid content was
`used in yield stress analysis. The vane was immersed in the
`flocculated kaolinite dispersions and rotated at a low and
`constant rate of 0.2 rpm. A thermostatted water-jacketed ves-
`sel connected to a temperature-controlled water bath main-
`tained a constant temperature. The volume of the sample
`used was 100 cm3. The length, radius, and blade thickness
`of the vane used were 30.17, 25.02, and 0.55 mm, respec-
`tively.
`
`To vitrify them, microgram size slurry samples were placed
`in copper rivets (7 mm, length of two rivets joined together,
`and 1 mm internal diameter) using a small microsyringe.
`The rivets were then mounted onto a cryogenic transfer rod
`and very rapidly plunged (∼ 2 m s−1) into liquid propane
`that was cooled by submersion in liquid nitrogen to “solidi-
`fy” and arrest all supramolecular motion. The vitrified sam-
`ples were transferred to and kept in a liquid-nitrogen-cooled
`specimen stage of the FESEM. The samples were fractured
`under vacuum by breaking off the top rivet. The vitrified
`water was then removed by sublimation at a temperature
`of −90 ◦C, avoiding local melting by excessive temperature
`rises. After 5 min at this temperature, the stage temperature
`was lowered again to approximately −180 ◦C, to stop any
`further sublimation. The samples were then coated with plat-
`inum to a thickness of 3 nm using a high-resolution sputter
`coater. They were subsequently analyzed at an accelerating
`voltage in the range 5–10 kV.
`
`3. Results and discussion
`
`2.8. Viscosity and hydrodynamic size measurements
`
`3.1. Zeta potential
`
`The flocculant solution flow properties and polymer
`molecular conformation were probed by investigating the
`absolute viscosity as a function of temperature. The viscosity
`was measured using an Ubbelohde-type capillary viscome-
`ter (Canon Instrument, USA), type OC, with a calibration
`constant of 0.002749 mm2 s−2. PEO and PAM concentra-
`tions of 0.1 wt% were used, as this corresponded to the
`initial polymer solution concentration used in flocculation.
`Refractive indices (RI) of the solutions were measured in
`triplicate using an Abbe High Accuracy 60/ED refractome-
`ter (Bellingham and Stanley, England) with temperature
`control to within ± 0.1 ◦C. The precision of the refractive
`index determination was ± 0.0001. The RI data were used
`for polymer hydrodynamic size estimation carried out by
`dynamic light scattering (DLS) analysis (Lexel model 95,
`Argon ion laser, Brookhaven photomultiplier) at 20–60 ◦C.
`All DLS measurements were performed at a concentration
`of 0.05 wt% polymer. This dilute concentration was used be-
`cause the resulting interparticle spacing is so large that the
`polymer molecule interactions are minimized [55]. Further-
`more, this concentration is close to polymer concentration
`observed upon addition of a known volume of 0.1 wt% to a
`500 cm3 pulp to give 500 g ton/solid.
`
`2.9. Scanning electron photomicrographs
`
`Fig. 2 shows the zeta potential of kaolinite particles at
`8 wt% solid content and 10−2 M KNO3 electrolyte and
`as a function of PEO (a) and PAM (b) concentration and
`at temperatures 20, 40, and 60 ◦C at pH 7.5. In the ab-
`sence of flocculant the zeta potential became more negative
`(from −30 to −34 to −46 mV) with increasing temperature,
`a finding which is in good agreement of previous studies
`[37,38].
`In the presence of both PEO and PAM, the magnitude of
`the zeta potential decreased with increasing polymer con-
`centration. However, the decrease was more dramatic with
`PEO than PAM. The reduction in zeta potential is attributed
`to the effect of the shift in the plane of shear away from the
`particle surface [43]. For PAM, which is 30% anionically
`charged, the small decrease in zeta potential with increasing
`flocculant concentration may be due to a shift in the posi-
`tion of plane shear due to the adsorbed layer being offset by
`an increase in charge around the particles of the negatively
`charged carboxyl groups of the polymer. PEO appears to ad-
`sorb more onto kaolinite particles than PAM at all tempera-
`tures, a finding which is strongly supported by the adsorption
`data below.
`
`3.2. Polymer adsorbed layer thickness
`and adsorption isotherm
`
`A high-resolution field emission scanning electron mi-
`croscope (FESEM, Philips XL30) was used to examine the
`kaolinite floc structures formed at 500 g flocculant/ton solid
`before and after shear at different temperatures. Samples
`were analyzed after rapid vitrification to avoid drying, which
`can cause internal structure rearrangement or modification.
`
`The adsorbed layer thickness as a function of tempera-
`ture and PEO concentrations calculated according to Eq. (1)
`is shown in Fig. 3. The layer thickness decreased with in-
`creasing temperature, consistent with the zeta potential data.
`The adsorption density of PEO (a) and PAM (b) at pH 7.5 as
`a function of equilibrium polymer concentration in solution
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`
`
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`
`149
`
`(a)
`
`(b)
`Fig. 2. Electrokinetic zeta potential of kaolinite in 10−2 M KNO3 at 8 wt%
`solids as a function of PEO (a) and PAM (b) concentration and temperature
`20 ("), 40 (!), and 60 ◦C (a) at pH 7.5.
`
`Fig. 3. Adsorbed layer thickness as a function of PEO concentration and
`temperature: 20 ("), 40 (!), and 60 ◦C (a) at pH 7.5.
`
`and at temperatures 20, 40, and 60 ◦C is shown in Fig. 4.
`The polymer adsorption increased steadily with increasing
`equilibrium concentration until a plateau was reached for
`each temperature. The maximum adsorption density of PAM
`was lower than PEO at similar temperatures. It appears
`that the negatively charged COO− pendant of the PAM re-
`duced its adsorption onto the negatively charged particles.
`The observed adsorption densities are significantly lower
`
`(a)
`
`(b)
`
`Fig. 4. Adsorption isotherm of PEO (a) and PAM (b) as a function of tem-
`perature 20 (!), 40 ("), and 60 ◦C (a) at pH 7.5 and 10−3 M KNO3.
`
`than those reported for other flocculated systems such as
`PEO and SiO2, MoO3, and V2O5 particles [26], ethyl(hydro-
`xyethyl)cellulose and SiO2 [52], and polysaccharide and
`talc [50].
`According to polymer adsorption theory [29], an in-
`crease in adsorption density with increasing temperature
`is expected since the Flory–Huggins interaction parame-
`ter (χ ), which relates to the enthalpy of mixing for both PEO
`and PAM and solvency, increases with increasing tempera-
`ture [49]. A decrease in solvency as temperature increases
`is indicated, hence the adsorption density increased. These
`results are, however, inconsistent with the adsorbed layer
`thickness or zeta potential data.
`
`3.3. Langmuir plots
`
`All adsorption isotherms for PEO and PAM onto kaoli-
`nite particles at all different temperatures were fitted to the
`Langmuir adsorption isotherm model [58] given by
`=
`+
`1
`Ceq
`Ceq
`(2)
`x/m
`K(x/m)max
`(x/m)max
`where Ceq (mg dm−3) is the equilibrium solution concen-
`tration of polymer, x is the amount of polymer adsorbed
`
`,
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`
`
`150
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`(mg dm−3), m is the total solid particle surface area per
`unit volume (m2 dm−3), K is the Langmuir adsorption con-
`stant (dm3 mg−1), and (x/m)max is the maximum amount
`of polymer adsorbed per unit solid surface area. A linear
`relationship should be obtained when Ceq/(x/m) is plot-
`ted against Ceq. The assumptions made for the derivation
`of Eq. (2) are that the solute (flocculant) and solvent (wa-
`ter) have equal molecular cross-sectional surface areas, there
`is no net solute–solvent interaction in the surface or bulk
`phases, and the adsorption process is reversible. It is ac-
`knowledged that these assumptions may not be fully justified
`in the case of polymer adsorption. The Langmuirian treat-
`ment of the low affinity adsorption process of the present
`work allows a direct and consistent semiquantitative com-
`parison to be made between PEO and PAM adsorption ther-
`modynamics, in the manner of recent flocculation studies of
`talc [50], silica [52], and hematite [62] dispersions. Conse-
`quently, useful mechanistic information relating to the exact
`contributions of entropic and enthalpic effects has emerged.
`It is pertinent to note that more sophisticated models [29,58]
`exist for quantitative treatment of the polymer adsorption
`process but they are not readily intractable in the terms of
`the present experimentally measured variables for deconvo-
`luting the adsorption thermodynamics.
`The values of K and (x/m)max can be determined respec-
`tively from the intercept and gradient of such a plot. A plot
`of Ceq/(x/m) as a function of Ceq for PEO (a) and PAM (b)
`at three different temperatures is shown in Figs. 5A and 5B,
`respectively, and clearly conforms to a Langmuir type of ad-
`sorption as depicted by the linearity of the plots. Values of
`K, (x/m)max, and δmax are given in Table 1 for comparison.
`A significant variation is seen in the K and (x/m)max values
`which are observed to increase, while δmax decreased with
`increasing temperature. It is clearly evident that more poly-
`mer was adsorbed at 60 than at 40 or 20 ◦C. K represents
`the affinity of the polymers for a particular particle surface
`and may be related to the Gibbs free energy of adsorption
`( Gads) (J mol−1) as
` Gads = −RT ln K,
`(3)
` Gads = Hads − T Sads,
`(4)
`where R is the general gas constant (8.314 J K−1 mol−1)
`and T is the absolute temperature (K). Fig. 6 shows the
`plot of Gads versus T according to Eq. (4) from which
`the estimate of the enthalpy ( Hads) (J mol−1) and entropy
`( Sads) (J K−1 mol−1) changes for PEO and PAM adsorp-
`tion was made, determined from the intercept and gradient
`of the straight line, respectively. It is assumed that both para-
`meters are independent of temperature over the range stud-
`ied [45].
`The estimated parameters, Hads, −T Sads, and Gads
`for PEO (a) and PAM (b) are shown in Table 2. The Gads
`values are found to be negative, increasing in magnitude with
`increasing temperature and of similar order of magnitude to
`the free energy of formation of 1 mole of hydrogen bonds
`
`20
`40
`60
`
`20
`40
`60
`
`(−20.9 kJ mol−1) [55]. The data not only indicate that the
`adsorption of PEO and PAM onto kaolinite is not only more
`favorable at elevated temperatures, but also suggest that the
`adsorption process is dominated by entropic rather than en-
`thalpic effects since | Hads| < |−T Sads|.
`3.4. Viscosity and hydrodynamic size
`
`The viscosity and polymer hydrodynamic size (diame-
`ter) of the PEO and PAM solutions measured as a function
`of temperature relate to solvency and polymer conforma-
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`(a)
`
`(b)
`
`Fig. 5. Langmuir plot of the adsorption of PEO (a) and PAM (b) at temper-
`ature 20 ("), 40 (!) and 60 ◦C (a) at pH 7.5.
`
`Table 1
`The calculated values of (x/m)max, K, and δmax for the adsorption of
`PEO (a) and PAM (b) onto kaolinite at pH 7.5 at various temperatures
`Temp
`K, equilibrium constant for
`(x/m)max
`(◦C)
`(mg m−2)
`adsorption (dm3 mg−1)
`(a)
`0.118± 0.02
`228± 5
`0.145± 0.05
`1042± 9
`0.184± 0.01
`1572± 8
`0.042± 0.005
`164± 7
`0.044± 0.002
`569± 5
`0.074± 0.003
`707± 7
`
`δmax
`(nm)
`15± 0.1
`9± 0.075
`4.5± 0.05
`4.0± 0.01
`3.1± 0.05
`2.1± 0.01
`
`(b)
`
`
`
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`
`151
`
`Fig. 6. Plot of ( Gads) for PEO (") and PAM (!) as a function of temper-
`ature.
`
`Fig. 7. Absolute viscosity of 0.1 wt% PEO (") and PAM (!) concentration
`as a function of temperature.
`
`Table 2
`Thermodynamic parameters for PEO (a) and PAM (b) adsorption onto
`kaolinite at pH 7.5
`−T Sads
`Temp
` Gads
`(kJ mol−1)
`(◦C)
`(kJ mol−1)
`−14.2± 0.3
`−42± 1.4
`−17.1± 0.4
`−45± 1.8
`−20.4± 0.5
`−48± 2
`−12.5± 0.5
`−26± 2
`−16.5± 0.3
`−28± 2.8
`−18.2± 0.45
`−29± 3
`
`20
`40
`60
`
`20
`40
`60
`
`(a)
`
`(b)
`
` Hads
`(kJ mol−1)
`28± 5
`28± 5
`28± 5
`14± 4
`14± 4
`14± 4
`
`tion, prior to adsorption onto kaolinite particle surface. It
`is evident from Fig. 7 for the viscosity data of a 0.1 wt%
`polymer solution that the solvent quality, as determined by
`water–polymer and polymer–polymer interactions, was sig-
`nificantly affected by temperature. The viscosity decreased
`markedly with increasing temperature, and this is believed
`to be due to a breakdown of polymer–water hydrogen bond
`interactions and concomitant changes of polymer molecular
`conformation. The polymer chains may be considered more
`expanded in space due to stronger water–polymer interac-
`tions via hydrogen bonding at the lower temperature range
`(20–40 ◦C); hence a higher viscosity resulted. With the dis-
`ruption of the water–polymer hydrogen bonds at higher tem-
`peratures, the polymer–polymer interactions became more
`favored and led to the formation of a more coiled or con-
`tracted conformation. This is clearly shown by the average
`hydrodynamic size of the polymer units in Fig. 8 below.
`In the temperature range 20–40 ◦C, the hydrodynamic di-
`ameter of both PEO and PAM were substantially constant;
`thereafter it decreased sharply as the temperature was fur-
`ther increased (Fig. 8). This observation is consistent with a
`good solvent prevailing at 20–40 ◦C and the polymer chains
`are more solvated to a larger size than at a higher temperature
`where a more tightly coiled chain results from a poor sol-
`vent. At similar molecular weights and concentrations, PAM
`has a higher hydrodynamic size than PEO. This may be due
`to the fact that the anionic charge of the PAM molecules sub-
`
`Fig. 8. Hydrodynamic size of polymer units in solution at 0.05 wt%
`PEO (") and PAM (!) as a function of temperature.
`
`stantially facilitated polymer chain expansion and enhanced
`the effective hydrodynamic volume in solution.
`
`3.5. Particle interactions by rheology
`
`,
`
`The shear yield stress, which is diagnostic of the floc
`strength, may be considered as a measure of the maximum
`force per unit area that the floc can withstand before rup-
`turing. It is related to the total energy (Esep) required to
`separate the flocs into single units by the following expres-
`sion [59–61],
`τB = 3φs n
`Esep
`8π r3
`where φs is the volume fraction of the primary particles in
`the dispersion, n is the average number of contacts per parti-
`cle in the floc (i.e., the coordination number, assumed to be
`12 for a face-centered cubic arrangement), and r is the ra-
`dius of the primary particles. The effect of temperature on
`shear yield stress and energy of separation (floc strength)
`at 500 g PEO or PAM/ton solid and 28 wt% solid kaolin-
`ite dispersions is shown in Fig. 9. A decrease in shear yield
`stress and energy of separation with increasing temperature
`
`(5)
`
`KASHIV1043
`IPR of Patent No. 9,492,392
`
`
`
`152
`
`P. Mpofu et al. / Journal of Colloid and Interface Science 271 (2004) 145–156
`
`mance and settling rates of coal particles using PEO. Upon
`flocculation with PAM, the initial settling rates decreased
`monotonically with temperature in the range 20–60 ◦C. Ad-
`ditionally, PEO produced higher settling rates than did PAM
`and this may be attributed to differences in the polymer
`structure-related characteristics (e.g., charge and conforma-
`tion). The settling behavior of flocculated kaolinite disper-
`sions suggests that an optimum temperature exists for en-
`hanced flocculation with PEO, shown to be around 40 ◦C.
`This observation is in good agreement with other studies of
`polyacrylate-flocculated alumina dispersions [36]. It was re-
`ported that a further increase in temperature beyond 45 ◦C
`hindered the settling behavior considerably. The decrease in
`settling rates with increasing temperature for both floccu-
`lants may be explained in terms of conformation changes,
`as well as weakened hydrogen bonding resulting from in-
`creased particle charge. Improved dewatering behavior was
`achieved at 500 g flocculant/ton kaolinite, which gives ad-
`sorption densities ∼ 50% of saturation coverage values.
`
`3.7. Consolidation
`
`The polymer conformational changes also had a strong
`impact on the sediment solid content as indicated by the data
`in Fig. 11, before and after shear at 250 rpm for 5 min and
`a polymer concentration of 500 g/ton solid. A much better
`consolidation is observed at a temperature of 20–40 ◦C for
`PEO and 20 ◦C for PAM in comparison with 60 ◦C without
`shear. Upon agitation, the PEO-flocculated slurries showed
`a dramatic improvement in consolidation at 20–40 ◦C, which
`decreased thereafter as the temperature was further increased
`to 60 ◦C. However, for PAM, no significant improvement in
`consolidation was observed before and after shear. The flocs
`formed were mostly destroyed upon moderate shearing.
`For PEO, large flocs which coalesced and densified into a
`single mass upon shearing resulted. The dense flocs could
`be hand-squeezed like a sponge in