Tailoring HASE rheology through polymer design: effects of hydrophobe size, acid content, and molecular weight.

Hydrophobically modified alkali-soluble emulsion (HASE) polymers are an important class of rheology modifiers for waterborne coatings applications. They are typically prepared as terpolymers by emulsion polymerization of ethyl acrylate (EA), methacrylic acid (MAA), and an associative macromonomer. The viscosity development and shear responses of HASE solutions depend on a number of factors. This article presents rheological data reflecting the impacts of three key variables: hydrophobe size, acid content, and molecular weight, on model HASE thickening and rheological performance. The relative contributions of hydrophobic association, chain expansion, and polymer chain length are discussed.

In steady shear flow, all thickener solutions approached some constant low-shear viscosity at small deformation rates. At the same molar composition, larger hydrophobe size resulted in higher viscosity development and greater shear thinning behavior.

The amount of acid monomer in HASE polymers can influence the balance between hydrophobic attraction and electrostatic repulsion forces. It was found that a minimum of 15 wt% MAA was required to effect dissolution and thickening of the model HASE polymers. Increasing the MAA level yielded higher zero-shear viscosity and storage modulus G' with maximal values being obtained at 40% MAA.

The molecular weight of the model thickeners was controlled by the amount of chain transfer agent (CTA) added during polymerization. When the CTA level was below 0.1 wt% based on total monomers, the polymer solutions displayed shear-thinning behavior. A small increase in CTA concentration beyond 0.1% resulted in a dramatic change to Newtonian flow, and the solution viscosity was nearly two orders of magnitude lower.

The model thickeners were also tested in a vinyl acrylic architectural paint formulation. The effects of each individual factor on paint thickener efficiency, high-shear, and low-shear properties are discussed and compared with solution rheology for predictive relationships.

Keywords: Rheology/flow control, thickeners, rheology, flow, leveling, waterborne, latex, water-based

**********

Hydrophobically modified alkali-soluble emulsion (HASE) polymers have become an important class of rheology modifiers for waterborne coatings applications as they offer improved rheology, ease of handling, and attractive economics. Compared to conventional thickeners, some favorable rheological attributes of HASE polymers are flow and leveling, film build, gloss development, and spatter resistance. HASE thickeners are typically prepared by emulsion polymerization as terpolymers of ethyl acrylate (EA), methacrylic acid (MAA), and an associative hydrophobic macromonomer (HMM). They are supplied as acidic, low viscosity emulsions. When neutralized with a suitable base, carboxylic acid groups along the polymer backbone are ionized to yield water-soluble polymers. The dissolution of particles results in interesting changes in polymer conformation and solution rheology. (1-6) A number of factors contribute to the thickening effect of HASE polymers: association of hydrophobic groups, topological entanglements, and chain expansion of high molecular weight polyelectrolyte backbone. (7-10)

Hydrophobic Association and Solution Rheology

Hydrophobic association, the primary thickening mechanism in HASE and other associative polymers, is a result of hydrophobe exclusion by water molecules. (11) Strong hydrogen-bonding forces between the water molecules drive the hydrophobic groups into clusters to minimize the disruption of highly ordered water structure. The formation of hydrophobic junctions constructs a transient network structure, which effectively builds up solution viscosity. The transient network theory (12,13) developed for telechelic hydrophobically modified ethoxylate urethane (HEUR) thickeners has been used to elucidate some rheological properties of HASE systems. According to this model, the network can be distinguished by a single characteristic time, [tau], that is controlled by the average lifetime of hydrophobe association in a junction. The solution viscosity is determined by this network relaxation time, [tau], and high frequency modulus, [G.sub.[infinity]], in the undisturbed state

[[eta].sub.0] = [tau]G[infinity] = [tau][v.sub.eff]kT (1)

where k is the Boltzman constant and T is the absolute temperature. The number density of elastically effective chains, [v.sub.eff], is defined as the number of chains per unit volume that bridge junctions involved in the network. The contribution to network junctions by chain entanglement is considered small relative to hydrophobic association in low molecular weight HEUR systems. (14) The reciprocal of [tau] can then be approximated by the hydrophobe dissociation rate, [[beta].sub.0], characterized by a bonding potential, [E.sub.m],

[[beta].sub.0] = [[omega].sub.0]exp([E.sub.m]/kT) (2)

where [[omega].sub.0] is the characteristic frequency of thermal vibration, estimated to be of the order of 0.1 ns. When the total number of association junctions is not affected by shear, the transient network theory predicts a constant viscosity as shear rate approaches zero.

[[eta].sub.0] = [[[v.sub.eff]kT]/[[beta].sub.0]] = [[[v.sub.eff]kT]/[[omega].sub.0]]exp([E.sub.m]/kT) (3)

The magnitude of [[eta].sub.0], depending on the amount of network structure at rest, is indicative of overall network strength. Equation (3) demonstrates that zero-shear viscosity is an exponential function of [E.sub.m] since the pre-exponential factor is a constant for a given system at a given temperature. The activation energy can be determined experimentally from the Arrhenius temperature dependence of zero-shear viscosity. (14)

The transient network theory has successfully described the rheological behaviors of low molecular weight, nonionic associative polymers such as HEUR systems. (12) HASE polymers, however, differ considerably from the telechelic HEUR thickeners: HASE polymers have higher molecular weights, and multiple hydrophobes are attached to the polyelectrolyte backbone through EO side chains. Therefore, HASE polymers are under greater influence of additional forces transmitted along the backbone. Deviation of HASE behavior from the transient network model is not surprising given the coexistence of many forces, such as hydrophobic association, topological entanglements, and electrostatic repulsion. (7)

Hydrophobic modifications are expected to have a profound impact on the thickening power and rheology of HASE solutions. The influence of hydrophobe structure can be estimated by the difference in chemical potential between monomeric and clustered hydrophobes using the Scott-Hildebrand equation (8, 15)

[DELTA][mu] = 2RT - ([V.sub.s] + [V.sub.p])([[delta].sub.s] - [[delta].sub.p])[.sup.2][X.sup.2]/2 (4)

where R is the universal gas constant and T is the absolute temperature. [V.sub.s] and [V.sub.p] are the molar volumes of the solvent and the hydrophobe, respectively. [[delta].sub.s] and [[delta].sub.p] are the solubility parameters of the solvent and the hydrophobe, respectively. X is the volume fraction of the hydrophobe in solution. Equation (4) illustrates the impact of the molar volume and solubility of the hydrophobe on the chemical potential of hydrophobe cluster formation. Increasing the molar volume or decreasing the solubility of the hydrophobe in aqueous media both promote intermolecular associations. Studies on HASE thickeners have confirmed that the strength of hydrophobe association and, consequently, the viscosity enhancement, increase with hydrophobe length. (9,16,17)

The rheology of a HASE thickener as a function of shear rate is of tremendous practical interest to a paint formulator. The application of shear forces causes the disruption of hydrophobic junctions, resulting in viscosity drop commonly referred to as the shear-thinning phenomenon. It has been found that shear-thinning behavior of HASE polymers can be described by the power-law fluid model of Ostwald and de Waele. (18,19) This model expresses viscosity profiles with a power law index n and a constancy index m:

[eta]([dot.[gamma]]) = m[dot.[gamma]][.sup.n-1] (5)

Shear thinning materials have n < 1 and shear thickening materials have n > 1. The constancy index m is a measure of thickener response. The higher the value, the greater the viscosity development of the thickener. The power law index n can be obtained from the double logarithmic plot of viscosity versus shear rate.

Other than hydrophobic association, greater hydrodynamic volume also enhances solution viscosity. The contribution of hydrodynamic volume is more complicated in HASE systems because of their polyelectrolyte backbones. Chain expansion results not only from a high molecular weight polymer, but also from electrostatic repulsion between the ionized acid groups in the polymer backbone. The influences of polymer composition and chain length on the dynamics and topology of the network, and, therefore, rheological responses, are not well understood at present in spite of the many technological advantages of HASE thickeners. (20-22) This lack of fundamental understanding testifies to the complexity of HASE systems.

Thickener Performance in Latex Paints

Ultimately, a coating chemist must be concerned with the rheology of formulated paints over a wide range of shear rates representative of storage and application conditions. (23) The rheology of some commercial HASE products in latex paint formulations has been examined. (20-22, 24-27) There is still a surprising lack of comprehensive studies on model HASE polymers. Since thickening arises from multicomponent interactions, associative thickeners are more sensitive to formulation changes. In addition to intermolecular associations in the aqueous phase, hydrophobic groups on HASE polymers can be adsorbed on dispersed components in a coatings formulation. A multitude of interactions in a pigmented waterborne latex coating often makes the quantitative data interpretation difficult. (27) Predictive relationships are, thus, highly useful if the rheological correlations between thickener solutions and formulated paints can be established.

In this article, the model thickeners were evaluated in a semigloss vinyl acrylic paint formulation. (20) The shear-dependent paint properties such as low-shear viscosity (LSV) and high-shear viscosity (HSV) are related to a coating's performance characteristics such as flow and leveling and film build.

Objectives

The objectives of this article are to examine the effects of three important variables: hydrophobe size, acid content, and molecular weight on rheology of model HASE thickeners. The synthesis parameters are individually varied in order to decouple the contributions of hydrophobic association, charge density and chain entanglements of high molecular weight HASE to viscosity development, and shear responses in aqueous solutions as well as latex paint formulations. The relative contributions of hydrophobic association, chain expansion due to charge repulsion, and polymer chain length will be discussed.

This article also attempts to correlate the paint thickening efficiency with thickener solution rheology.

EXPERIMENTAL

Sample Preparation

The model HASE polymers were prepared from associative macromonomers (HMM) based on linear alcohol ethoxylates. The macromonomers used in this study were derived from dodecanol, octadecanol, and docosanol. The details of stepwise synthesis are described elsewhere. (9,16,17)

The thickener latexes were synthesized by conventional semicontinuous emulsion polymerization according to the method previously described by Jenkins et al. (9) All thickener samples contained 0.22 mol% of macromonomer. The macromonomer was varied in one study to include three hydrophobe sizes, C12, C18, and C22. Only one set of samples had varying acid content ranging from 10-50 wt%, while the acid level was kept at 40% for the other two series. No chain transfer agent (CTA) was added to regulate the molecular weight except for the sample set used in the evaluation of molecular weight effect. The proprietary chain transfer agent used in the model samples was varied from 0 to 3.0 wt% based on total monomers. The compositions of three sets of model HASE samples are summarized in Table 1.

The latexes were dialyzed using degenerated hydroxylethyl cellulose membrane over three weeks with daily water change. The dialyzed latexes were then converted into solutions by dilution and neutralization to pH of 9 with sodium hydroxide solution (from Fisher Scientific) or 2-amino-2-methyl-1-propanol (AMP-95[TM] from ANGUS Chemical Company). The weight concentration of thickener solutions was chosen to maintain the same rheometer geometry for viscosity measurements of the entire sample series: 1.0% for HASE1-HMM series, 0.75% for HASE2-MAA series, and 2.0% for HASE3-CTA series.

Rheological Measurements

The steady state flow experiments were performed on a controlled stress rheometer AR1000 (TA Instruments) using a 2[degrees] and 60 mm cone-and-plate. The experiment temperatures ranged from 10[degrees]-40[degrees]C. Temperature control to within [+ or -] 0.1[degrees]C was achieved through the use of a peltier plate. Oscillatory strain amplitude sweep at 25[degrees]C was carried out at two frequencies, 6.28 and 50 rad/sec, respectively. The strain amplitude sweep at low frequency was conducted to determine the linear viscoelastic (LVE) region for each sample. The high frequency experiments were performed to probe the structure of the polymers since the high frequency modulus is related to the number of elastically effective chains. (28)

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Paint Thickener Performance Test

An interior semigloss vinyl acrylic paint formulation with 20% pigment volume concentration (PVC) and 30% volume solids (VS) was employed for thickener testing. (20) The thickener/water mixture was added slowly to the paint base under agitation, and then the paints were subsequently neutralized to pH 8.5 ~ 9.5 with ammonium hydroxide. Depending on thickener efficiency, varying amounts of thickener were added to bring the paints to a constant mid-shear viscosity, as determined on a Stormer viscometer (Brookfield KU-1). The paints were equilibrated for 24 hr prior to viscosity measurements. The paint thickener efficiency (PTE) was expressed as the number of dry pounds of thickener needed to thicken 100 gal of the test paint formulation to 95 KU. Two other viscometers were then utilized to assess the rheology of thickened paints. The LSV and HSV were measured on Brookfield LV and ICI cone-and-plate viscometers, respectively.

RESULTS AND DISCUSSIONS

Solution Rheology -- Effect of Hydrophobe Size

STEADY SHEAR EXPERIMENTS: Table 2 summarizes the calculated values of the molar volume and solubility parameters for the three hydrophobes utilized in the synthesis of HASE1-HMM model thickeners. (17) The solubility parameter varies only slightly with the size of alkyl groups. The increase of carbon length, however, leads to a large change in molar volume of the hydrophobe, from 227.1 to 399.2 [cm.sup.3]/mol. The important rheological consequences of hydrophobe molar volume is evident in Figure 1, which compares steady shear viscosity of three HASE thickeners containing C12, C18, and C22 hydrophobes.

Irrespective of hydrophobe size, all flow curves in Figure 1 seem to approach some constant zero-shear viscosity at low shear rates, as predicted by the transient network theory. In this unperturbed condition, the HASE polymer forms a stable and extended network through the intermolecular association between the hydrophobes of neighboring polymer chains. Figure 1 demonstrates that the rheological properties of thickener solutions are drastically altered by the hydrophobe size. Larger hydrophobe size or molar volume gives rise to higher viscosity development. The C12-based model thickener results in only modest viscosity enhancement. When the hydrophobe length is increased from C12 to C22, the zero-shear viscosity goes up by four orders of magnitude. The increase of solution viscosity with hydrophobe length is consistent with the results of other publications. (9,16,17) The zero-shear viscosity has been shown previously to increase exponentially with the molar volume of the hydrophobe. (17)

Larger hydrophobe also tends to impart more shear-thinning rheology. (9,17) The viscosity reduction under shear manifesting network disruption is an attractive feature of associative thickeners. Figure 1 shows that the shear-thinning behavior is also a strong function of hydrophobe size. The HASE1-C22 thickener exhibits the most pronounced shear-thinning character. The solution viscosity declines rapidly when the shear rate exceeds 0.1 [sec.sup.-1]. The flow behavior of HASE1-C12 solution is substantially less shear sensitive and the viscosity reduction occurs at much higher shear rates. In the shear thinning regions, the viscosity reduction with shear rate obeys the power-law fluid model [equation (5)]. The calculated power law index n decreases with increasing hydrophobe length, confirming that larger hydrophobe exhibits stronger shear-thinning behavior. (17)

ACTIVATION ENERGY FROM STEADY SHEAR: The model thickener solutions exhibit typical viscosity loss with temperature, following the Arrhenius temperature dependence. Figure 2 demonstrates nice linear relationships when the extracted [[eta].sub.0] from the steady shear experiment is displayed against 1/T in a semi-log plot. The activation energy estimated from the slope increases with alkyl length up to C18. The calculated values for C12, C18, and C22 are 5.3, 13.0, and 11.6 kcal/mole.K, respectively. A higher [E.sub.m] value indicates stronger association junction. It is interesting that the C22 hydrophobe generates the highest solution viscosity but has lower activation energy than the C18 hydrophobe. Brown has proposed that the magnitude of viscosity is a function of the chain conformation and the hydrophobe length. (29) Thus, the decrease of activation energy from C18 to C22 suggests that chain conformation is influenced by the hydrophobe length in a complex manner and that hydrophobic association is not totally independent of chain conformation as assumed by the transient network theory.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Solution Rheology -- Effect of Acid Content

WATER SOLUBILITY OF HASE POLYMERS: Carboxylic acid monomer is a basic component of HASE thickeners as it renders polymers water-soluble upon neutralization. The amount of acid monomer present in HASE polymer determines not only the electrostatic repulsion between charged carboxylate groups, but also the thermodynamic interactions between the monomeric components and water. A minimum acid level is required to facilitate the solubilization of HASE polymers and this critical value depends on hydrophobe size and concentration. (9) For the model HASE polymers containing C18 hydrophobe, a minimum of 15 wt% MAA is required to afford dissolution and thickening. The model thickeners with greater than 15% MAA yields optically clear solution at 0.75% weight concentration. The thickener containing 10% MAA remains turbid beyond pH 10, suggesting ionization of 10% MAA does not provide adequate water solubility.

STEADY SHEAR EXPERIMENTS: Figure 3 presents the steady shear viscosity of thickener solutions at 25[degrees]C. It is evident that thickening is also effected at MAA content greater than 15% and that the magnitude of viscosity development increases with methacrylic acid content. A 5% change in acid content from 15% to 20% results in a 20-fold jump in low shear viscosity. The ascending trend continues up to 40% MAA content, and the zero-shear viscosity recedes thereafter when MAA level reaches 50%. Higher acid level improves water solubility and at the same time increases the charge density along the polymer backbone. Both effects can enhance thickener solution viscosity as chains unfold to assume greater hydrodynamic volume. However, increasing MAA content also results in higher [T.sub.g] or chain stiffness, which decreases the hydrodynamic volume of the solubilized polymer coil and, therefore, the solution viscosity. The competing effects of chain stiffness and hydrophilicity yield an optimum thickening efficiency with respect to acid content. (9) The optimum acid content for HASE2-MAA is approximately 40%, at which solution viscosity reaches a maximum.

The conformation of polyelectrolyte backbone apparently plays an important role in controlling the number of intermolecular association junctions and, hence, network structures. Presumably, smaller junctions at very low acid content are formed with predominately intramolecular association. Greater chain expansion and conversion of intramolecular to intermolecular junctions due to charge repulsion produce a continuous viscosity increase with higher acid content. The aggregation number in each junction and the number of junctions in the network change with acid content. The large rise in solution viscosity when MAA concentration is increased from 15% to 20% is probably a combined effect of a larger aggregation number and higher junction density. Larger aggregation number, greater junction density, or both, are responsible for ascending viscosity up to 40% MAA. Eventually, inter-chain charge repulsion exceeds hydrophobic attraction. Fewer junctions are formed and each contains a smaller number of hydrophobes. The viscosity downturn at 50% MAA reflects the diminution of network strength. Hydrophobic association strength can also be affected by electrostatic repulsion and chain stiffness. The interplay of these factors determines the optimal thickening power with respect to methacrylic acid level.

STRAIN AMPLITUDE SWEEP EXPERIMENTS: Only the high frequency data shown in Figure 4 will be discussed here since the oscillation strain amplitude sweeps yield similar results at both frequencies. Both storage and loss moduli remain unchanged at low strains when no disruption of the polymer network occurs. The low strain section with constant G' values is termed the linear viscoelastic region (LVE). Figure 4 shows that the LVE regions are limited to small strain amplitudes. A more extended LVE region is observed for samples containing 40% and 50% MAA. This is because the network structure that has higher elastic modulus G' is more resistant to the applied stress. According to transient network theory, the G' data at high frequency is indicative of the relative magnitude of intermolecular hydrophobic junctions [equation (1)]. Like the zero-shear viscosity data, the modulus of the polymer solution at low strain amplitude increases progressively with increasing acid content and peaks at 40% MAA.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Figure 5 compares the zero-shear viscosity and high frequency storage modulus of the model thickeners. The maximum for both zero-shear viscosity and elastic modulus occurs at 40% MAA. Above this critical acid content, the solution viscosity and G' both decrease. These results imply that rising viscosity with MAA content is largely due to a higher number of mechanically active junctions. Continuous increase of acid content opens up the polymer chain, promoting more inter-chain association versus intra-chain association. The additive effects of chain expansion due to charge repulsion and a larger amount of inter-chain associations improve the overall strength of the network and, thus, enhance the HASE thickening power. Exceedingly high charge density and chain stiffness values at 50% MAA hinder the hydrophobic associations. Increasing chain stiffness results from greater charge repulsion and heightened glass transition temperature. Lower G' and G" values are evidence for a reduction in the number of associative junctions and the network volume. A lower zero-shear viscosity at 50% MAA thus reflects a weaker network overall. Figure 5 demonstrates the usefulness of the transient network theory for predicting the viscosity in relation to the number of elastically effective chains in the network.

[FIGURE 7 OMITTED]

ACTIVATION ENERGY FROM STEADY SHEAR: The activation energy is plotted in Figure 6 as a function of acid content. The data confirms that acid content indeed affects the hydrophobe dissociation kinetics, despite the fact that both hydrophobe length and thickener concentration are constants for the HASE2-MAA model thickeners. The maximum activation energy obtained at 20% MAA indicates that HASE2-MAA20 forms the strongest hydrophobic junctions, possibly arising from the optimal aggregation number in each junction. As mentioned earlier, the ratio of inter-chain to intra-chain associations increases with acid content as the chain unfolds due to greater electrostatic repulsion. The chain conformation of this polymer in aqueous solution allows the maximum number of hydrophobes to participate in association junctions. However, further increase in acid content actually lowers the activation energy.

The difference in association strength influences both the number of the hydrophobes associating to form network junctions as well as the number of junction points. Since the total number of hydrophobes per chain is the same for the model HASE2-MAA series, the only explanation for the decrease of activation energy is that the number of junctions increases at the expense of junctions becoming weaker and smaller.

Solution Rheology -- Effect of Molecular Weight

The molecular weight of HASE polymers has a dramatic effect on the type of rheology that a HASE polymer imparts. This influence, however, has not been well studied, and, to our knowledge, this is the first article investigating the dependence of rheological performance on HASE molecular weight. As previously mentioned, the polymer molecular weight is primarily controlled by the amount of chain transfer agent (CTA) added during polymerization.

STEADY SHEAR EXPERIMENTS: The molecular weight effect will be discussed in terms of the concentration of chain transfer agent. Figure 7 demonstrates that a small variation of CTA level can result in amazingly different rheology. The polymer prepared without CTA exhibited a typical shear-thinning behavior. At 0.1% CTA concentration, the polymer has a longer first Newtonian plateau followed by a lesser degree of viscosity reduction. When 0.2% of the CTA is added during polymerization, the resulting polymer solution yields an even more extended Newtonian rheology and the solution viscosity was nearly two orders of magnitude lower than that of HASE without CTA (HASE3-CTA0). Above 0.2%, the solution viscosity declines almost linearly with further increase of CTA level. In other words, solution viscosity of HASE polymers increases with increasing molecular weight, a behavior similar to that of unassociating polymer systems. This result contradicts the findings on telechelic HEUR thickeners. Extensive studies by Jenkins demonstrated that lower hydrophobe density due to increasing HEUR molecular weight results in lower solution viscosity. (12) However, nonionic HEUR polymers are generally lower in molecular weight than HASE thickeners. The results suggest that the contributions of chain entanglement and network topology as moderated by molecular weight cannot be neglected in HASE systems.

The solution viscosity of a polymer with the same monomer ratio as HASE3-CTA0, but without hydrophobe modification, is also included in Figure 7 for comparison. The zero-shear viscosity of this polymer (ASE-CTA0) is three orders of magnitude lower than that of HASE3-CTA0 thickener, even though the molecular weights are similar. The tremendous difference in viscosity development emphasizes the impact of hydrophobic association. At the same time, the unmodified polymer at the same concentration yields a zero-shear viscosity in between those of HASE3-CTA0.4 and HASE3-CTA0.6. Clearly, molecular weight and electrostatic repulsion are the main contributing factors to the solution viscosity of ASE-CTA0 polyelectrolyte. In the case of hydrophobically modified HASE3-CTA thickeners, the molecular weight effect is even more amplified. This is because longer polymers have a proportionally higher number of hydrophobic groups attached to the backbone since the monomer molar ratios were kept the same within the sample series. Thus, polymers of higher molecular weight bearing more hydrophobes can span the distance between the association junctions to form a more extended network. This bridging effect combined with topological entanglements results in greater viscosity enhancement. Viscosity development of shorter HASE polymers, on the other hand, benefits little from chain entanglement and hydrodynamic volume. More critically, a weaker network is formed because short polymer chains with fewer hydrophobes are less capable of forming "superbridges." Consequently, solution viscosity decreases quickly with decreasing molecular weight or increasing CTA level.

The response of the HASE polymer solutions under shear is complicated by the coexistence of both hydrophobic associations and topological entanglements at higher polymer concentration or higher molecular weight. Figure 7 illustrates that lower molecular weight polymers seem to have longer Newtonian plateaus. Lower efficiency thickeners are less sensitive to shear because they do not generate the same high level of network structures. (30) The initial increase of shear rate induces similar changes in the flow curves of the two highest molecular weight samples, as observed with the unmodified ASE-CTA0 solution. The shear thinning profile of the unmodified polymer solution can be attributed to the reduction of chain entanglements and alteration of chain conformation caused by shear stress. The same cause can be assigned to the weak shear-thinning behavior noted for HASE3-CTA0 and HASE3-CTA0.1. Three lower molecular weight samples do not display such initial viscosity reduction owing to the absence of chain entanglements, and, consequently, much wider Newtonian regions are observed.

All thickener solutions experience the catastrophic viscosity decline at higher shear stress, a signature of network breaking. The critical stress also is shifted to lower value for smaller molecular weight samples. This result confirms that large molecular weight HASE form stronger networks that can withstand higher disruption force.

[FIGURE 8 OMITTED]

ACTIVATION ENERGY FROM STEADY SHEAR: Figure 8 displays the activation energy as a function of CTA concentration ranging from 0% to 0.6%. The respective values are 13.9, 12.9, 11.6, 11.4, and 10.7 kcal/mole.K, reflecting changing hydrophobe dissociation rate with molecular weight. The strong decrease of solution viscosity with molecular weight can be partially explained by the relative values of activation energy. The solution viscosity continues to decline with increasing CTA concentration, but the activation energy of hydrophobe dissociation decreases only slightly when CTA level is greater than 0.1%. The small difference in activation energy indicates that junction strength depends strongly on hydrophobe size and concentration, and much less on molecular weight when the HASE polymer is below a certain chain length.

[FIGURE 9 OMITTED]

[FIGURE 10 OMITTED]

Since the transient network theory based upon HEUR polymers assumes that topological entanglement has no consequences, viscosity development is purely dictated by hydrophobic associations. The viscosity and activation energy results presented in this article suggest obvious deviation of HASE behavior, which is consistent with observations made by other groups. (7) For example, broadening of the relaxation spectrum and weaker frequency dependence of dynamic moduli in the terminal zone have been attributed to the combined effects of topological entanglements and hydrophobic association. (31, 32) The concentration dependence of HASE solution viscosity (7) also seems to support the "hindered reptation" model (33) proposed for entangled networks. This model assumes that the disengagement time of a single chain from the network is much longer compared to the residence time of the hydrophobe within a junction. Therefore, the macroscopic stress relaxation is not only affected by the disengagement rate of a hydrophobe from a junction, but also by physical constraints arising from entanglements of the polymer chains. The data in this study provides appealing evidence for the existence of topological entanglements and its impact on HASE thickening and shear-dependent rheological properties.

Paint Thickener Test -- Effect of Hydrophobe Size

The viscometers utilized to assess the paint rheology correspond to three shear rates: Brookfield at ~0.4, Stormer at 70, and ICI cone-and-plate at 10000 [sec.sup.-1]. (19) The vertical lines in Figure 1 mark these three shear rates encountered in the paint applications. When the solution viscosity is used to predict the efficiency of individual thickeners at intermediate shear rate, it is clear that the smallest hydrophobe, C12, yields the least efficient thickener. Figure 1 also indicates that HASE1-C18 and HASE1-C22 result in similar PTE values because C22 hydrophobe exhibits more dramatic shear-thinning behavior.

The paint thickener efficiency results correlate very well with the predictions based on rheology of thickener solutions. Table 3 summarizes the rheological performances of model thickeners in the paint. The amount of thickener required to thicken 100 gal of paint to 95 KU is 9.9, 5.1, and 5.0 dry pounds for HASE1-C12, HASE1-C18, and HASE1-C22, respectively. Since the PTE value is inversely related to thickening power, the thickener efficiency is almost doubled when the hydrophobe size increases from C12 to C18. Because of its strong shear-thinning nature, the thickener based on C22 is only slightly more efficient than the C18-based thickener at medium shear rate.

Consistent with the solution viscosity profiles, HASE1-C22 generates the highest low-shear viscosity in spite of the lowest thickener concentration. Although the solution viscosity is quite different, the thickeners containing C12 and C18 hydrophobes provide similar low-shear viscosity. This probably reflects the concentration effect, because less thickener is present in the paints as the medium size hydrophobe imparts higher efficiency.

During rheological characterization, the steady state flow experiments were terminated before the rheometer reached the shear rate that is typical on the ICI cone-and-plate viscometer. Hence, the data points at high-shear rates are not available in Figure 1, although the trend is clear. The HSV of the paints confirms that the trend indeed continues into high-shear rates: the ICI viscosity is inversely dependent on the hydrophobe size. The sharp decline of ICI viscosity with increasing hydrophobe size is more than what the decreasing thickener concentration can account for. Larger hydrophobe imparts greater shear thinning and, thus, lower ICI viscosity. The application results on the thickened paints demonstrate that the thickener performance in this paint formulation reproduces the flow behaviors of thickener solutions at respective shear rates.

Paint Thickener Test -- Effect of Acid Content

The paint thickener efficiency (PTE) and HSV data of the HASE2-MAA model thickener series are both displayed in Figure 9. The thickening power, as shown in Figure 9, increases sharply with initial increase of MAA content up to 30% and levels off beyond 40% MAA. The balancing acts of hydrophobic association, charge repulsion, and chain stiffness determine the ultimate thickening efficiency with respect to the MAA content in the polymer backbone. The PTE results mirror the solution viscosity data at intermediate shear rate and the maximum thickening occurs at 40% MAA in both cases.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

Figure 9 shows that the ICI viscosity, or HSV, is remarkably similar to the PTE result. The HSV of the paints thickened with the model HASE polymers appears to decrease with increasing acid content as well. Because the polymer with higher acid content is more efficient, it requires less thickener to bring the paints to the same KU. Not surprisingly, the HSV of the paints turns out to be a linear function of PTE or thickener concentration.

The Brookfield viscosity, or LSV, displayed in Figure 10 decreases linearly with increasing acid content. This trend is opposite to the solution viscosity result where low shear viscosity increases continuously with acid content up to 40% (Figure 3). As for the ICI viscosity, lower thickener concentration in the paint is the primary cause for the Brookfield viscosity decrease with increasing MAA level. It is noted, however, that LSV depends more strongly on MAA level than HSV when Figure 10 is compared to Figure 9. A great number of interactions are present under low-shear conditions. The polymers with higher acid content are stiffer and more hydrophilic, hindering chain entanglement and thickener adsorption onto dispersed particles. Thus, the LSV of formulated paint is lower.

The PTE data illustrates that acid content of HASE polymer affects thickener efficiency and an optimum exists. The plateau in PTE indicates that a very high concentration of methacrylic acid is counterproductive for a thickener to achieve balanced properties. For instance, water sensitivity may become a concern at higher acid level while efficiency is only marginally improved. There are more effective ways to improve thickener efficiency, such as appropriate selections of the length and concentration of hydrophobe in polymer, and chain transfer agent level. The LSV and HSV results demonstrate the rheological impact of MAA content in HASE thickeners: high MAA level improves flow and leveling but reduces film build due to increased thickening efficiency or less thickener requirement. Such declines of both LSV and HSV indicate that manipulation of MAA content alone would not allow good flow and leveling and film build simultaneously.

Paint Thickener Test -- Effect of Molecular Weight

The PTE value, as shown in Figure 11, increases with CTA concentration. In other words, higher CTA level or lower molecular weight reduces the thickening power of HASE3-CTA model samples. The linear decrease in efficiency at CTA concentration above 0.1% is consistent with solution viscosity data shown in Figure 7.

The ICI viscosity, or HSV, also shown in Figure 11, increases with CTA level in a similar fashion. The increase of solution viscosity with higher CTA level is not apparent from Figure 7 since the steady shear data at equivalent high shear rates were not available for comparison. The positive slopes of both ICI and PTE, in relation to CTA level, suggest the significance of thickener concentration effect on paint film build. It requires a greater amount of a smaller molecular weight sample to achieve 95 KU. Figure 12 indeed confirms the linear dependence of HSV, or film build, on thickener concentration.

The LSV is compared with zero-shear viscosity of thickener solutions in Figure 13. It is clear that the CTA concentration has a stronger influence on the zero-shear viscosity of the straight thickener solutions than the LSV of the formulated paints. The zero-shear viscosity decreases sharply and continuously with increasing CTA level. In formulated paints, the LSV decreases initially with CTA concentration but approaches a plateau at a CTA concentration greater than 0.2%. Again, the increasing thickener concentration in the paint slows the LSV attenuation due to decreasing molecular weight.

The paint rheology data suggest that molecular weight is a critical consideration in HASE polymer design. Increasing chain transfer agent level results in lower molecular weight polymer and decreasing thickening efficiency. A low molecular weight HASE is a less efficient thickener, but it can afford good flow and leveling and film build. In other words, a more Newtonian type rheology can be obtained by adjusting the molecular weight of HASE polymers.

[FIGURE 13 OMITTED]

CONCLUSIONS

This article examined the effects of hydrophobe size, acid content, and molecular weight on the rheology of model HASE thickeners in solution and paints. The results demonstrate that these three variables can produce significant differences in thickener solution behavior and paint application properties. A wide range of rheology and, hence, performance optimization, can be achieved by a judicious choice of key synthesis parameters. However, these factors are not of equal importance to HASE rheology.

Hydrophobe size undoubtedly has a profound impact on thickening power and shear responses of model thickeners. A larger hydrophobe provides higher viscosity and greater shear-thinning behavior, reflecting changing association strength and network structure.

Acid content plays an important role in the alkali-solubility and rheology of HASE thickeners, but it does not have the same degree of impact as hydrophobe size or molecular weight. As the polymer backbone becomes more hydrophilic and extended at higher acid level, inter-chain associations are promoted and thickening power is enhanced. The solution viscosity and elastic modulus increases steadily with acid content up to 40%, despite the formation of weaker, smaller, hydrophobic junctions. Greater chain stiffness and inter-chain electrostatic repulsion at 50% MAA results in losses in viscosity and elastic modulus.

Molecular weight can drastically alter the rheology of HASE thickeners. When the polymer chain is shortened to some critical length, the thickener solution changes from a shear-thinning to a Newtonian fluid. Although the strength of association junctions does not vary greatly, the absence of chain entanglements and "superbridges" produces weaker networks and substantially lower solution viscosity.

When the model thickeners were evaluated in a vinyl acrylic architectural paint, surprisingly good correlation was found between thickener efficiency and solution viscosity at the corresponding intermediate shear rate. This predictive power renders rheological characterization of HASE thickener solution highly useful. Larger hydrophobe size, higher acid content, and greater molecular weight all enhanced the thickener efficiency. The HSV, or film build, is primarily dependent on thickener concentration in the paint; therefore, opposite effects were observed. The LSV, also a function of thickener concentration, is more sensitive to the interactions in the paint matrix, charge density, and polymer entanglements. In general, smaller hydrophobe, higher acid content, and lower molecular weight improved flow and leveling by decreasing the LSV of HASE polymer-thickened paints.

Table 1 -- Compositions of Model HASE Thickeners

                          HMM       MAA Content  CT
Sample ID     Hydrophobe  (mole %)  (wt%)        (wt%)

HASE1-C12     C12         0.22      40           0
HASE1-C18     C18         0.22      40           0
HASE1-C22     C22         0.22      40           0
HASE2-MAA10   C18         0.22      10           0
HASE2-MAA15   C18         0.22      15           0
HASE2-MAA20   C18         0.22      20           0
HASE2-MAA30   C18         0.22      30           0
HASE2-MAA40   C18         0.22      40           0
HASE2-MAA50   C18         0.22      50           0
HASE3-CTA0    C18         0.22      40           0
HASE3-CTA0.1  C18         0.22      40           0.1
HASE3-CTA0.2  C18         0.22      40           0.2
HASE3-CTA0.4  C18         0.22      40           0.4
HASE3-CTA0.6  C18         0.22      40           0.6
ASE-CTA0      --          --        41           0

Table 2 -- Calculated Molar Volume and Solubility Parameter of
Hydrophobes

                                        Solubility Parameter
            Molar Volume                [[delta].sub.p] (cal/
Hydrophobe  [V.sub.p] ([cm.sup.3]/mol)  [cm.sup.3])[.sup.1/2]

C12         227.1                       7.04
C18         327.5                       7.71
C22         399.2                       7.77

Table 3 -- Performance Test Results of HASE1-HMM Thickeners in Paints

Hydrophobe  PTE (dry lb/100 gal)  ICI (P)  BF @0.3 rpm (P)

C12                 9.9             1.2        718
C18                 5.1             0.4        609
C22                 5.0             0.3       1277

Presented at the 81st Annual Meeting of the Federation of Societies for Coatings Technology, November 11-12, 2003, in Philadelphia, PA.

References

(1) Siddiq, M., Tam, K.C., and Jenkins, R. D., "Dissolution Behavior of Model Alkali-Soluble Emulsion Polymers: Effects of Molecular Weight and Ionic Strength," Colloid Polym. Sci., 277, 1172-1178 (1999).

(2) Kumacheva, E., Rharbi, Y., Winnik, M.A., Guo, L., Tam, K.C., and Jenkins, R.D., "Fluorescence Studies of an Alkali-Swellable Associative Polymer in Aqueous Solution," Langmuir, 13, 182 (1997).

(3) Horiuchi, K., Rharbi, Y., Yekta, A., Winnik, M.A., Jenkins, R.D., and Bassett, D.R., "Dissolution Behavior in Water of a Model Hydrophobic Alkali-Swellable Emulsion (HASE) Polymer with C20H41 Groups," Can. J. Chem., 76, 1779-1787 (1998).

(4) Wang, C., Tam, K.C., Jenkins, R.D., and Bassett, D.R., "Potentiometric Titration and Dynamic Light Scattering of Hydrophobically Modified Alkali Soluble Emulsion (HASE) Polymer Solutions," Chem. Phys., 2, 1967-1972 (2000).

(5) Wang, C., Tam, K.C., and Jenkins, R.D., "Dissolution Behavior of HASE Polymers in the Presence of Salt: Potentiometric Titration, Isothermal Titration Calorimetry, and Light Scattering Studies," J. Phys. Chem. B, 106, 1195-1204 (2002).

(6) Knaebel, A., Skouri, R., Munch, J.P., and Candau, S.J., "Structural and Rheological Properties of Hydrophobically Modified Alkali-Soluble Emulsion Solutions," J. Polym. Sci. Part B: Polym. Phys., 40, 1985-1994 (2002).

(7) English, R.J., Gulati, H.S., Jenkins, R.D., and Khan, S.A., "Solution Rheology of a Hydrophobically Modified Alkali-Soluble Associative Polymer," J. Rheol., 41(2), 427-444 (1997).

(8) Hoy, K.L. and Hoy, R.C., "Polymer with Hydrophobe Bunches," U.S. Patent 4,426,485.

(9) Jenkins, R.D., Delong, L.M., and Bassett, D.R., "Influence of Alkali-Soluble Associative Emulsion Polymer Architecture on Rheology," in Hydrophilic Polymers: Performance with Environmental Acceptability, Glass, J.E. (Ed.), ACS Symposium Series No. 248. American Chemical Society, Washington, D.C., Chap. 23, 1996.

(10) Bassett, D.R., Jenkins, R.D., and Shay, G.D., "Aqueous Latexes Containing Macromonomers," U.S. Patent 5,486,587.

(11) Tanford, C., The Hydrophobic Effect: Formation of Micelles and Biological Membranes, John Wiley & Sons, New York, NY, 1980.

(12) Jenkins, R.D., "The Fundamental Thickening Mechanism of Associative Polymers: A Rheological Study," Ph.D. Thesis, Lehigh University, Bethlehem, PA, 1990.

(13) Tanaka, F. and Edwards. S.F., "Viscoelastic Properties of Physically Crosslinked Networks (Parts 1, 2, and 3)," J. Non-Newtonian Fluid Mech., 43, 247-272, 273-288, and 289-309 (1992).

(14) Annable, T., Buscall, R., Ettelaie, R., and Whittlestone, D., "The Rheology of Solutions of Associating Polymers: Comparison of Experimental Behavior with Transient Network Theory," J. Rheol. 37, 695-726 (1993).

(15) Hildebrand, J.H. and Scott, R.L., The Solubility of Non-Electrolytes, Dover Publication, Inc., New York, NY, p. 253, 1964.

(16) Tirtaamadja, V., Tam, K.C., and Jenkins, R.D., "Rheological Properties of Model Alkali-Soluble Associative (HASE) Polymers: Effect of Varying Hydrophobe Chain Length," Macromolecules, 30, 3271-3282 (1997).

(17) Wu, W., Grieb, R.S., and Shay, G.D., "Rheological Characterization of Model HASE Thickeners," in Proc. 29th International Waterborne, High-Solids, and Powder Coatings Symp., New Orleans, LA, 343-355, 2002.

(18) Bird, R.B., Armstrong, R.C., and Hassager, O., Dynamics of Polymer Liquids, Vol. 1: Fluid Mechanics, Wiley, New York, NY, 1987.

(19) Darby, R., Viscoelastic Fluids: An Introduction to Their Properties and Behavior, Marcel Dekker, New York, NY, 1976.

(20) Shay, G.D., Bassett, D.R., and Rex, J.D., "A New Class of Alkali-Swellable Associative Thickeners," Surf. Coat. Intern., November, 446-453 (1993).

(21) Shay, G.D., "Flow Control with HEURASE Rheology Modifiers," APCJ, December 20, 44-53 (1993).

(22) Gambino, J.J. and Schaller, E.J., "Rheology Modifiers For Latex Paints," Mod. Paint Coat., July, 36-40 (1982).

(23) Schoff, C.K., Rheology, Federation Series on Coatings Technology, Federation of Societies of Coatings Technology, Blue Bell, PA, 1991.

(24) Shay, G.D., and Rich, A.F., "Urethane-Functional Alkali-Soluble Associative Latex Thickeners," JOURNAL OF COATINGS TECHNOLOGY, 58, No. 732, 43 (1986).

(25) Shay, G.D., "Alkali-Swellable and Alkali-Soluble Thickener Technology: A Review," in Polymers in Aqueous Media, Performance Through Association, Glass, J.E. (Ed.), Advances in Chemistry Series 223, American Chemical Society, Washington, D.C., 457-494, 1989.

(26) Toshio, M., Fernando, R.J., and Glass, J.E., "The Influence of Hydrophobically-Modified, Alkali-Swellable Emulsions on the Rheology of Pigmented Coatings," Surf. Coat. Intern., 76(1), 8 (1993).

(27) Glass, J.E., "Influence of Water-Soluble Polymers on Rheology of Pigmented Latex Coatings," in Water-Soluble Polymers, Beauty with Performance, Glass, J.E. (Ed.), Advances in Chemistry Series 213, American Chemical Society, Philadelphia, PA, 391-416, 1984.

(28) Tirtaamadja, V., Tam, K.C., and Jenkins, R.D., "Superposition of Oscillations on Steady Shear Flow as a Technique for Investigating the Structure of Associative Polymers," Macromolecules, 30, 1426-1433 (1997).

(29) Brown, G. and Chakrabarti, A., "Structure Formation in Self-Associating Polymer and Surfactant Systems," J. Chem. Phys., 96, 3251 (1992).

(30) Howard, P.R., Leasure, E.L., Rosier, S.T., and Schaller, E.J., "Systems Approach to Rheology Control," in Polymers as Rheology Modifiers, Glass, E.J. (Ed.), ACS Symposium Series 462, American Chemical Society, Washington, D.C, 211, 1991.

(31) Graessley, W.W., "The Entanglement Concept in Polymer Rheology," Adv. Polym. Chem. 17, 19-151 (1974).

(32) Aubry, T. and Moan, M., "Rheological Behavior of a Hydrophobically Associating Polymer Solutions," J. Rheol. 38, 1681-1692 (1994).

(33) Leibler, L., Rubinstein, M., and Colby, R.H., "Dynamics of Reversible Networks," Macromolecules, 24, 4701-4707 (1991).

Wenjun Wu and Greg D. Shay -- The Dow Chemical Company*

*UCAR Emulsion Systems, 410 Gregson Dr., Cary, NC 27511.