pubs.acs.org/Langmuir © 2010 American Chemical Society
Rheological Characterization of the Influence of PVOH on Calcite Suspensions Rasmus Eriksson,*,†,‡ Annaleena Kokko,§ and Jarl B. Rosenholm†,‡ †
Physical Chemistry, A˚bo Akademi University, Turku, Finland, ‡Center of Excellence for Functional Materials, Turku, Finland, and §Technical Research Center of Finland, Espoo, Finland Received December 21, 2009. Revised Manuscript Received March 17, 2010
Flow properties of the calcite/poly(vinyl alcohol) (PVOH) system were studied and related to the microstructure of the suspension. Adsorption of PVOH on calcite was confirmed, and it results in a shift of the slipping plane out from the surface. The charge density at the surface is assumed to remain unchanged. Since the PVOH used is only partially hydrolyzed, the most likely adsorption conformation consists of residual acetate groups adsorbed to the surface and vinylalcohol groups extending outward from the surface as loops and tails. The microstructure and flow properties of the calcite/PVOH system was found to go through several different stages as a function of PVOH concentration. At low PVOH concentrations a gradual weakening of the initially formed floc network is observed as a function of PVOH concentration. Further addition of PVOH eventually leads to breakdown of the flocs which results in a sterically stabilized suspension with a very low viscosity. This state persists for a narrow concentration range of PVOH, and increasing the PVOH concentration over a certain limit leads to a second gradual increase in viscosity. The system is believed not to undergo reflocculation at high PVOH concentrations as judged from the nonelastic nature of the suspensions. Instead, the polymers form a viscous matrix in the solution while the particles remain well-dispersed. At high enough PVOH concentration, the free volume available for the particles is greatly reduced, and the viscosity increases sharply.
1. Introduction Understanding the complex mechanisms behind different types of suspension flow is the subject of intensive research.1-7 The vast number of different types of colloids existing makes it somewhat cumbersome to develop common theories describing flow at higher than a rudimentary level. As a rule, every system behaves a little bit different from others, which has led to the development of mathematical models describing flow (e.g., power laws).8-10 Using these fundamental models, simple colloidal systems are fairly well characterized from a mathematical point of view. The physical reality, however, is still not well-characterized, and needs to be studied further. Any particle system can be studied at rest, but the situation changes significantly during flow. The suspension structure during flow is always dependent on shear rate, and it is strongly affected by particle interactions. There is no easy way to study these systems directly and qualified guesses have to be made. Key questions include to what degree flocs (if present) are broken and how the particles rearrange within a shear field? The most common type of flow for suspensions is shear thinning,11 where the viscosity can decrease quite dramatically as a function of (increasing) shear rate. This type of flow is usually attributed to
flocs breaking and particles rearranging into flow layers with intermediary liquid layers.11-13 This is also the basis for interpreting the results of this study. The system in the present study consists of calcite (GCC) including different concentrations of poly(vinyl alcohol) (PVOH). The introduction of a polymer confers an extra dimension to the system which complicates the analysis. Polymers can influence suspension properties in various ways,14 depending on whether they adsorb to the particle surface or not. An adsorbing polymer can lead to bridging flocculation or steric stabilization. Often the outcome is concentration dependent. Rheology is a method wellsuited for studying colloidal structures in suspensions. The equilibrium structures of the suspensions in this investigation were studied by means of low amplitude oscillatory shear while standard viscometry was used to study flow properties. Only indirect assumptions can be made from viscometry, but by combining with oscillatory shear it is possible to make a more comprehensive analysis of the different suspension states. In addition, the adsorption of PVOH and its effect on the electrokinetic potential of calcite was determined. The evolution of the suspension structure of the calcite/PVOH system is presented as a function of PVOH concentration.
*Corresponding author. E-mail:
[email protected]. Telephone: þ358 (0)2 215 4252.
2. Experimental Section 2.1. Materials. The calcite particles (Omya Ltd., HC-90) were
(1) Papenhuijzen, J. M. P. Rheol. Acta 1972, 11, 73–88. (2) Firth, B. A.; Hunter, R. J. J. Colloid Interface Sci. 1976, 57, 248–256. (3) Firth, B. A. J. Colloid Interface Sci. 1976, 57, 257–265. (4) Firth, B. A.; Hunter, R. J. J. Colloid Interface Sci. 1976, 57, 266–275. (5) Tadros, Th. F. Langmuir 1990, 6, 28–35. (6) Otsubo, Y. Langmuir 1995, 11, 1893–1898. (7) Wallstr€om, A.; J€arnstr€om, L.; Peltonen, J. Nordic Pulp Paper Res. J. 2007, 22, 102–110. (8) Krieger, I. M.; Dougherty, T. J. Trans. Soc. Rheol. 1959, 3, 137–152. (9) Cross, M. M. J. Colloid Sci. 1965, 20, 417–437. (10) Carreau, P. J. Trans. Soc. Rheol. 1972, 16, 99–127. (11) Barnes, H. A.; Hutton, J. F.; Walters, K. An Introduction to Rheology, Third Impression; Elsevier Science Publishers B.V.: Amsterdam, 1993; Chapter 7.
7946 DOI: 10.1021/la9048117
used as received. The particles have an average size of 1.1 μm, which was determined by sedimentation velocity (Quantachrome Microscan). PVOH (Clariant, Mowiol 40-88) was received as a 10% w/w solution. A 5% stock solution was prepared, which was used in the Acoustosizer titration. The average molecular weight of the (12) Hoffman, R. L. Trans. Soc. Rheol. 1972, 16, 155–173. (13) Hoffman, R. L. J. Colloid Interface Sci. 1974, 46, 491–506. (14) Napper, D. H. Polymeric Stabilization of Colloidal Dispersions; Academic Press Inc.: London, 1983.
Published on Web 03/26/2010
Langmuir 2010, 26(11), 7946–7952
Eriksson et al. PVOH is ∼205000 g/mol and the degree of polymerization ∼4200. The degree of hydrolysis is 88%. In all cases where the PVOH concentration is mentioned, it is expressed relative to the amount of solids in the suspensions. The water used was of Millipore (A10) quality. 2.2. Adsorption Isotherms. The adsorption of PVOH was determined on 40 wt % (20 vol %) calcite suspensions. The amount adsorbed was determined by spectrophotometric analysis of the supernatants. This method is based on the formation of a blue/ green complex between triiodide and PVOH in the presence of boric acid.15,16 The wavelength of maximum adsorption (657 nm) was determined by performing scans over several different PVOH concentrations. Joshi et al.16 found a wavelength of maximum adsorption of 658 nm for 88% hydrolyzed PVOH, which agrees very well with our determination (657 nm). The same scans were used to construct a calibration curve for determining PVOH concentrations between ∼2 and 80 mg/L. The analyses were performed with a Shimadzu UV-2501PC spectrophotometer. The suspensions were allowed to equilibrate for about 15 min before removal of the supernatants. The removal of the supernatant was done by a two stage centrifugation. The first stage was centrifugation at 3000 rpm for 30 min followed by collection of the liquid phase (usually still containing particles). The first stage was done in order to minimize mechanical settling of PVOH. The liquid phase from the first stage was thereafter centrifuged at 20000 rpm for 30 min after which the clear supernatant was collected and analyzed. 2.3. ζ Potential. ζ potential was measured with a Colloidal Dynamics Acoustosizer II instrument. The PVOH titration was made using the Acoustosizers automatic titration software. The pH varied during the titration from an initial value of 9.1-8.4 with increasing PVOH concentration. At PVOH concentrations higher than 1 pph the pH stayed at a constant value of 8.4. The temperature was 25 °C. 2.4. Rheology. The rheological measurements were performed with a Bohlin VOR rheometer. The measuring system used (C25) is of Couette type consisting of a bob (25 mm diameter) and a rotating cup (27.5 mm diameter). The frequency range for viscoelastic measurements lies between 10-3 and 20 Hz and the maximum amplitude is 20 mrad. The volume of sample needed is about 13 mL. All measurements were made at 25 °C.
3. Results and Discussion 3.1. Adsorption and Surface Potential. The PVOH adsorption isotherm on calcite particles is shown in Figure 1. The amounts adsorbed are characteristic for polymers.17-19 Water is not a very good solvent for fully hydrolyzed PVOH,20 which might lead to increased adsorption onto some surfaces. However, fully hydrolyzed PVOH has been reported to only adsorb in very small quantities on certain surfaces21-23 and others have found that partially hydrolyzed PVOH adsorbs in larger amounts than fully hydrolyzed PVOH.24,25 The hydrophobicity of the solid surface will also affect adsorption, and calculations confirm that the calcite surface is (15) Finley, J. H. Anal. Chem. 1961, 33, 1925–1927. (16) Joshi, D. P.; Lan-Chun-Fung, Y. L.; Pritchard, J. G. Anal. Chim. Acta 1979, 104, 153–160. (17) Greenland, D. J. J. Colloid Sci. 1963, 18, 647–664. (18) Kocherga, I. I.; Baran, A. A. Theor. Exp. Chem. 1977, 12, 497–503. (19) Backfolk, K.; Rosenholm, J. B.; Husband, J.; Eklund, D. Colloids Surf. A 2006, 275, 133–141. (20) Koopal, L. K. Colloid Polym. Sci. 1981, 259, 490–498. (21) Tadros, Th. F. J. Colloid Interface Sci. 1978, 64, 36–47. (22) Khan, A. U.; Briscoe, B. J.; Luckham, P. F. Colloids Surf. A 2000, 161, 243–257. (23) Mikkola, P.; Lev€anen, E.; Rosenholm, J. B.; M€antyl€a, T. Ceram. Int. 2003, 29, 393–401. (24) Kavanagh, B. V.; Posner, A. M.; Quirk, J. P. Faraday Discuss. Chem. Soc. 1975, 59, 242–249. (25) Koopal, L. K.; Lyklema, J. J. Electroanal. Chem. 1979, 100, 895–912. (26) Kerisit, S.; Parker, S. C.; Harding, J. H. J. Phys. Chem. B 2003, 107, 7676– 7682. (27) Kvamme, B.; Kuznetsova, T.; Uppstad, D. J. Math. Chem. 2009, 46, 756–762.
Langmuir 2010, 26(11), 7946–7952
Article
Figure 1. Adsorption isotherm for PVOH on calcite particles. The inset shows the first four points for clarity.
Figure 2. ζ potential of 10 wt % calcite particles as a function of PVOH concentration. Also shown is the percentage of adsorbed PVOH. The shaded area marks the approximate region where more than 90% of the added PVOH adsorbs onto the particles.
hydrophilic in nature.26,27 Although the acetate groups are more hydrophobic than the hydroxyl groups in the bulk PVOH chains, the fairly high electron density of the acetate groups leads to a partial negative charge on these groups. Thus, the most likely situation is that the residual acetate groups have a higher affinity for the positively charged calcite surface28 than the bulk polymer chains. This may lead to a rather thin adsorption layer with vinylalcohol loops and tails extending outward from the surface in between acetate groups that are adsorbed to the surface. Napper14 described the vinylacetate groups as anchors on any solid surface with the vinylalcohol chains usually acting as stabilizing moieties. The ζ potential of the calcite particles as a function of PVOH concentration is shown in Figure 2. The ζ potential decreases smoothly as a function of PVOH concentration, which is believed to be due to a shift of the slipping plane further away from the surface leading to a screening of the effective charge.29-31 These (28) Eriksson, R.; Merta, J.; Rosenholm, J. B. J. Colloid Interface Sci. 2007, 313, 184–193. (29) Sidorova, M.; Golub, T.; Musabekov, K. Adv. Colloid Interface Sci. 1993, 43, 1–15. (30) Bakandritsos, A.; Psarras, G. C.; Boukos, N. Langmuir 2008, 24, 11489– 11496. (31) Koopal, L. K.; Hlady, V.; Lyklema, J. J. Colloid Interface Sci. 1988, 121, 49–62.
DOI: 10.1021/la9048117
7947
Article
Figure 3. ζ potential of 10 wt % calcite particles as a function of adsorbed amount of PVOH. The adsorbed layer thickness is calculated from eq 1.
Eriksson et al.
Figure 4. Shear viscosity of 40 wt % calcite particle suspensions at different PVOH concentrations. Flow curves are shown for PVOH concentrations up to 0.7 pph.
results can be used to estimate an “electrophoretic thickness” of the adsorbed layer,32,33 which is discussed below. Also shown in Figure 2 is the percentage of adsorbed PVOH, calculated from the equilibrium amount in solution. Background corrections with corresponding PVOH solutions were made, and the results do not differ from the noncorrected measurement, confirming that PVOH does not contribute to the signal at a significant level. Up to ∼0.1 pph PVOH more than 90% of the PVOH is adsorbed, and this corresponds to the initial high-affinity region in Figure 1 (up to ∼0.1 mg/m2 adsorbed). Between ∼0.1 and ∼1 pph PVOH an intermediate adsorption region is observed. The adsorbed amount of PVOH increases linearly in this region (up to ∼0.65 mg/m2), and the ζ potential gradually starts to level off (Figure 2). In this region, the adsorbed polymer layer probably reorganizes and becomes denser. If it is assumed that PVOH does not affect the surface charge density, then the polymer chain layer also becomes thicker (see Figure 3). This means that some of the adsorbed acetate groups either desorb or simply that the density of adsorbed acetate groups at the surface increases. Both of these situations would allow for longer loops and tails extending from the surface, thus increasing the effective layer thickness. Above 1 pph PVOH, the adsorbed amount still increases, but the fraction of added PVOH that adsorbs to the surface decreases. At 5 pph PVOH less than 50% of the added PVOH is adsorbed (Figure 2). It is possible that, due to the high concentration of PVOH, the polymer chains are incorporated in the particle-liquid network and are only mechanically removed from the suspension during centrifugation. An apparent increase in the adsorbed amount would be observed in this case, although the polymers technically speaking are not adsorbed to the surface. The rheological measurements are reinforcing this assumption based on the nonelasticity of PVOH solutions. A nonelastic solution indicates only weak interactions between PVOH molecules, which makes multilayer adsorption less likely. The electrophoretic thickness of the adsorbed polymer layer is calculated using Gouy-Chapman double layer theory.34 An assumption is that the adsorbed PVOH does not alter the surface charge density and thus the potential decay from the surface is easy to calculate provided the surface potential is known. The surface
where ψδ is the Stern layer potential at a distance δ (=0.4 nm) from the surface. The ionic strength is 0.0025, which is based on the dissolved amounts of calcium and carbonate ions.28 The calculated thickness as a function of adsorbed PVOH is shown in Figure 3. Similar values for the layer thickness can be found elsewhere.31,35,36 However, the electrophoretic thickness can only be used as a very rough estimation of the adsorbed layer thickness, most notably because it is only an expression for the statistical ion distribution near a charged surface. It does not consider polymer-particle or polymer-solvent interactions at all. Basically, two factors in eq 1 govern the layer thickness, of which the first one is the difference in ζ (or surface) potential at zero adsorbed polymer and a presumably constant low ζ potential at high amounts of adsorbed polymer. Obviously, the layer thickness increases as a function of absolute difference in potential, meaning that usually a higher initial surface potential will lead to a larger calculated layer thickness. The second, and more important factor, is the ionic strength. For example, in our case, if the ionic strength would be 1 decade lower, the calculated layer thickness would be more than three times larger (all other things being equal). This is not to say that it is impossible, but it is nevertheless highly unlikely. The electrophoretic thickness is useful for estimating the slipping plane displacement, which does not necessarily have to coincide with the polymer layer thickness. One thing is for sure though; the polymer layer cannot be thinner (on average) than the electrophoretic thickness. It can be thicker, however, in the same sense as the diffuse layer of ions extends a significant distance outward from the slipping plane. 3.2. Viscosity. The flow curves of calcite particle suspensions at different PVOH concentrations are shown in Figures 4 and 5.
(32) Eremenko, B. V.; Malysheva, M. L.; Baran, A. A. Colloids Surf. 1992, 69, 117–124. (33) Barany, S.; Shilov, V.; Madai, F. Colloids Surf. A 2007, 300, 353–358. (34) Hiemenz, P. C. Principles of Colloid and Surface Chemistry, 2nd ed., Marcel Dekker Inc.: New York, 1986.
(35) Cosgrove, T.; Crowley, T. L.; Vincent, B. In Adsorption from Solution; Ottewill, R. H., Rochester, C. H., Smith, A. L., Eds., Academic Press Inc.: London, 1983; pp 287-297. (36) Chibowski, S.; Paszkiewicz, M.; Krupa, M. Powder Technol. 2000, 107, 251–255.
7948 DOI: 10.1021/la9048117
potential is approximated at the Stern layer (∼0.4 nm from the physical surface33), and the ζ potential at zero PVOH concentration is used as an approximate Stern layer potential. The electrophoretic thickness, h, can be calculated with the following equation:33 eζ eψδ ¼ tanh expð - Kðh - δÞÞ ð1Þ tanh 4kT 4kT
Langmuir 2010, 26(11), 7946–7952
Eriksson et al.
Figure 5. Shear viscosity of 40 wt % calcite particle suspensions at different PVOH concentrations. Flow curves are shown between PVOH concentrations 0.85 and 10 pph.
One characteristic of calcite particle suspensions is that they tend to form a network of flocs in aqueous suspensions,37 therefore exhibiting a fairly high low shear rate viscosity. Calcite particle suspensions are highly shear thinning, however, which results in very low viscosity at higher shear rates (Figure 4). The shear thinning phenomenon of colloidal suspensions is a very wellknown property and has been discussed extensively in the literature.11 In short, any existing particle flocs and networks are broken down by the shear into significantly smaller units or even single particles, which align along the field of flow into layers of particle-rich zones with solvent rich layers in between. This leads to a significant reduction in viscosity. In the initial region of high affinity adsorption (up to ∼0.1 pph PVOH), there is no major change in the flow curve (Figure 4). Particle interactions govern the flow properties and typical shear thinning flow is observed. Above about 0.5 pph PVOH a change in the flow curve occurs. The overall viscosity goes down, and shear thinning is not as pronounced anymore. In addition, the floc network starts to break down also under zero shear conditions, since the low shear rate viscosity is reduced (Figure 4). This is the transition region between a flocculated system and a sterically stabilized system. The PVOH concentration is still not high enough to attain maximum stability in the system and some weak interactions between particles still exist. These will be further discussed in connection with the viscoelasticity measurements. Following the transition region there is a stable region where the viscosity is very low (Figure 5). This occurs between PVOH concentrations ∼0.8 and 1.5 pph. Almost all of the flocs have been broken, and the viscosity is nearly Newtonian (Figure 5). Steric repulsion between the adsorbed polymer layers keep the particles in a well dispersed state, although the larger particles settle fairly quickly under no-flow conditions. The most probable high shear rate flow configuration for the system consists of layers of particle-rich zones with intermediary liquid layers in between. This is also the case for the highest shear rates measured in Figure 4. Moreover, it is important to keep in mind that the discussion is about layers with a certain volume, which is not to be confused with well organized planes. The association into flow layers is by no means self-evident and requires some justification. Although a number of different flow (37) Eriksson, R.; Pajari, H.; Rosenholm, J. B. J. Colloid Interface Sci. 2009, 332, 104–112.
Langmuir 2010, 26(11), 7946–7952
Article
configurations can be visualized, the most important fact to remember is that the difference in viscosity is considerable between a flocculated system and a well-dispersed system. In other words, the internal friction of the system is significantly reduced which certainly rules out a substantial migration of particles perpendicular to the direction of flow. Therefore, the suspension structure is definitely organized in a directional arrangement during flow. Pioneering work in this field was made by Hoffman.12,13 Here, the layers of particles are pictured as planes in which the particles are arranged in a hexagonal pattern. A regular particle arrangement was confirmed by studying light diffraction patterns during flow.12,13 When a critical shear stress (or rate) was attained, a disruption in the flow layers was detected as a considerable jump in viscosity. This signifies the onset of more disordered flow, where the hexagonal arrangement of particles largely disappears and is replaced by flow layers of irregularly spaced particles. Just above the critical shear stress significant overlap between layers occur, and this is the cause for the viscosity jump. A gradual decrease in viscosity as a function of increasing shear rate was observed after the jump, which means that the layers reform, although in a more disordered state.12,13 Hoffman12,13 studied monodisperse systems of a high volume fraction (>0.47), which differs significantly from our system. The same hydrodynamic forces apply, however, and thus the same arguments can be used for the formation of flow layers. The hexagonal flow arrangement at low shear stresses (rates) does not apply for our system, since the volume fraction is too low and a random floc network exists in the system at rest (with no added polymer). Instead, the structure gradually breaks down with increasing shear rate, and the smaller units arrange in the direction of flow. This can be compared to the region following the viscosity jump in Hoffmans work.12,13 Most likely, layers containing irregularly spaced particles and/or flocs form under the influence of shear which are separated by intermediary solvent-rich layers. At longer distances from the plane of shear the structure is probably increasingly disorganized, and particle migration perpendicular to the flow direction will increase in frequency. The free PVOH in solution actually helps to maintain the layer structure because of steric effects. Although the distribution of PVOH is unclear, it would still seem likely that there is a certain accumulation of PVOH in solvent-rich layers between the particle-rich layers. This is likely because the experiments indicate no attractive interactions between PVOH molecules, which implies that at least PVOH does not accumulate in the particle rich zones since they can also be considered PVOH rich zones due to adsorbed PVOH. Whatever the concentration of PVOH in the solvent-rich layers, it will act as an additional steric barrier toward particle migration in the perpendicular flow direction, thus stabilizing the particle-rich layers. At even higher concentrations of PVOH (g2 pph), the viscosity starts to increase again. However, the flow curves are completely different in shape compared to low PVOH concentration (Figure 4-5). Almost Newtonian flow curves are observed in the high PVOH concentration region (Figure 5). Only at the highest PVOH concentrations measured a slight tendency to shear thinning can be seen (Figure 5), but not nearly as pronounced as at low PVOH concentrations (Figure 4). In effect, the flow properties have shifted from a system governed by particle interactions to a system governed by polymer interactions. In this region the particles are well dispersed and do not contribute to a large degree to the overall viscosity. In addition, sedimentation is slower due to the increased viscosity of the continuous phase. With increasing PVOH concentration the frequency of polymer chain collisions DOI: 10.1021/la9048117
7949
Article
Eriksson et al.
Figure 6. Summary of viscosities for the different characteristic flow regions. Shear viscosities for one chosen “low” shear rate (4.61 s-1) and one chosen “high” shear rate (184 s-1) are presented.
and entanglement increases, which leads to an increased viscosity. Moreover, PVOH binds a lot of water and therefore reduces the volume fraction of bulk water, which leads to a further increase in viscosity. The fact that the viscosity is nearly Newtonian implies that the polymers do not align very well along the field of flow, at least not within the studied shear rate range. A summary of the viscosities at two different shear rates as a function of PVOH concentration are shown in Figure 6. Four different flow regions (I-IV) are identified in Figure 6 and schematic representations of regions I-III are shown in Figure 7. In Figure 6, the difference between particle interaction controlled (I) and polymer interaction controlled (IV) systems are clear. While there is a big difference between low and high shear rate viscosity in region I, the viscosities in region IV almost coincide. 3.3. Viscoelasticity. The network structure that forms during zero shear conditions was previously found to be elastic in nature.37 Increasing amounts of added PVOH leads to a gradual weakening of the structure and eventually a collapse of the network. This is demonstrated by a series of strain sweep measurements with increasing PVOH concentration (Figure 8). Only the storage modulus is shown since the loss modulus was very low in all measurements shown in Figure 8. Two different characteristic responses are seen in Figure 8, the first one at PVOH concentrations up to 0.1 pph, and the second one at PVOH concentrations between 0.5 and 0.6 pph. These correspond to region I and II, respectively, in Figures 6 and 7. In region I, the particles still form an elastic network in the suspension, although it becomes weaker with increasing PVOH concentration. This is observed both as a decreasing storage modulus in the linear viscoelastic region and as a smaller strain amplitude at which the structure breaks as a function of increasing PVOH concentration (Figure 8). The PVOH concentration in this region is still too low to cover a significant part of the total surface area, but leads in particular to a weakening of the bridges between flocs, and thus a reduced network strength. The response in region II is the most difficult to evaluate. A region with an elongated elastic response is observed which is much weaker than the original network structure (Figure 8). The complete viscoelastic response at 0.6 pph PVOH is shown in Figure 9 as an example. At low strain amplitudes (up to ∼1 mrad, Figure 9), the response is more or less constant and the phase angle stays at values below 10 deg. The amount of PVOH in solution is still not 7950 DOI: 10.1021/la9048117
Figure 7. Schematic pictures of flow regions I-III. The left column illustrates the situation when the suspension is at rest (no external force). The right column illustrates the situation when a relatively high (>100 s-1) shear rate is applied on the system. The direction of flow is assumed to be horizontal.
Figure 8. Strain sweep measurements of 40 wt % calcite suspensions at different PVOH concentrations. Only the storage modulus is shown. The frequency was 1 Hz.
enough to cover the entire surface area of the particles, and individual flocs covered by PVOH are formed instead. Most likely, some of these flocs are bound together by PVOH bridges which leads to a weak elastic structure that easily breaks down under shear. The ability to withstand higher strain amplitudes than the original network (Figure 8) also indicates that polymeric bridges impart elasticity on the system as opposed to the original network which was held together by van der Waals or electrostatic Langmuir 2010, 26(11), 7946–7952
Eriksson et al.
Figure 9. Viscoelastic response for 40 wt % calcite with 0.6 pph PVOH.
Figure 10. Loss modulus and phase angle for 40 wt % calcite suspensions as a function of PVOH concentration. The frequency was 1 Hz in all measurements.
forces (mechanism is not determined). The elastic response for flexible polymeric bridges is expected to work over longer distances than electrostatic or van der Waals forces. This could explain the intermediate region between 1 and 3 mrad where the structure apparently becomes weaker with increasing strain while still maintaining its elasticity (Figure 9). In this region, the polymeric bridges are gradually broken down, and individual flocs are formed. During this process, a number of “free” polymer chain ends will emerge in the solution as the weaker points of contact of the polymeric bridges are desorbed. These loose ends will most likely adsorb on the original floc and may also lead to individual particles or smaller flocs breaking away from the original floc. When the concentration of PVOH in solution is high enough to break up most of the flocs, i.e. when the particles are well dispersed, the viscoelastic response is too weak to yield any reliable data. This corresponds to region III in Figure 6, and in fact the nonviscoelastic behavior continues into region IV. The response is too weak to yield any viscoelastic data up to a PVOH concentration of 3 pph. At higher than 3 pph PVOH the response is more viscous than elastic and constant up to the maximum appliable strain of the instrument (20 mrad). A summary of the loss moduli as a function of PVOH concentration is shown in Figure 10. Langmuir 2010, 26(11), 7946–7952
Article
Figure 11. Strain sweep of a 10% PVOH solution.
Up to 7 pph PVOH the viscosity of the suspension is still fairly low (Figure 5) and does not show signs of any strong structure. The response is mainly viscous with phase angles around 80 degrees (Figure 10). The PVOH concentration is not high enough for a significant number of collisions to occur, and therefore the suspension flow properties are still good. Above 7 pph PVOH there is a change in the response, however, with a drop in phase angle and a more distinct increase in loss modulus. The reason for the increase in elasticity is probably not increased entanglement or specific interactions between PVOH molecules, because the response of a 10% PVOH solution is very nonelastic in nature (Figure 11). Instead, the polymer chains form a matrix in the solution in which the particles become trapped. With a decreasing volume fraction of bulk water, the space available for the particles to move diminishes, and the viscosity of the system increases sharply. The observed elastic response at PVOH concentrations higher than 7 pph most likely originates from the particles oscillating in the polymer matrix.
4. Conclusions The calcite/PVOH system was found to evolve through several different stages as a function of PVOH concentration. PVOH was found to adsorb to the surface of calcite, most likely with the acetate groups of the partially hydrolyzed PVOH attaching to the surface while the chains containing hydrolyzed groups form loops and tails that extend outward from the surface. The flow properties depend heavily on the concentration of PVOH, and goes through several regions with characteristic rheological behavior. They can be summarized as follows: I. PVOH Concentration 1.5 pph. Viscosity starts to increase again due to an increasing amount of free polymer chains in solution. The particles are well dispersed, and very low elasticity is observed. At PVOH concentrations above 7 pph the mobility of the particles is diminished as a result of confinement in the polymer matrix. A small increase in elasticity
7952 DOI: 10.1021/la9048117
Eriksson et al.
is observed which most likely originates from particles oscillating in the polymer matrix. Acknowledgment. The Technical Research Center of Finland and the Academy of Finland are acknowledged for financial support. Supporting Information Available: Figures showing the complete shear viscosity data for 0.05 pph PVOH and 5 pph PVOH as well as the frequency dependent measurement for 10 pph PVOH. This material is available free of charge via the Internet at http://pubs.acs.org.
Langmuir 2010, 26(11), 7946–7952
