Amphiphilic derivatives of carboxymethylcellulose: Evidence for intra- and intermolecular hydrophobic associations in aqueous solutions.

INTRODUCTION

Hydrophobically modified water-soluble polymers (both synthetic and natural) or amphiphilic polymers have received great attention because of their potential use in various industrial applications such as cosmetics, paints, drilling fluids, and oil recovery and owing to the advantage of both the hydrophilic and hydrophobic groups (1-9). However, from the practical point of view, the hydrophobically modified polysaccharides are important in many applications because of the nontoxicity of the polysaccharides as well as their biocompatibility and biodegradability (9-13). The associating polymers form a major part of the hydrophilic backbone and a small proportion of the hydrophilic backbone and small proportion of the hydrophobic groups (1), (4), (6-8), (13-16). In aqueous solution, the hydrophobic groups aggregate to minimize their exposure to water and thereby form intra or intermolecular association resulting in hydrophobic microdomains (1), (7), (8), (16). In the semidilute regime, where intermolecular aggregates prevail over intramolecular ones, clusters of hydrophobic domains are formed (1), (8). The addition of salt enhances this interaction because of the increased in the solvent polarity, which results in good salt-tolerance (3), (8), (16). However, in the diluted regime the added salt can screen out the polyelectrolyte effect, while, at the same time, enhance the hydrophobic intramolecular association (7).

Amphiphilic polymers can be prepared either by chemical modification of the precursor polymers or by free radical copolymerization of the appropriate monomers (3-9), (11), (12), (16-19). These methods produce random, block, alternating, or graft copolymers (7), (17). In the case of the polysaccharides, the hydrophobic groups are grafted onto a hydrophilic chain by chemical modification of the precursor (3), (9), (11), (18), (200. Free radical copolymerization has been widely used because of its simplicity (5), (6), (8), (16), (19), (21). The micellar copolymerization is the most convenient way to ensure the solubilization of the hydrophobic monomer within the surfactant micelles and the copolymerization occurring in the continuous water medium (5), (6), (8), (16), (19), (21).

The presence of hydrophobic groups greatly influences the physicochemical properties of amphiphilic polymers in aqueous solution and their properties are directly related to the involvement of intra or intermolecular associations between the hydrophobic moieties, depending on the investigated range of concentration (16), (22), (23). In an amphiphilic polyelectrolyte, if the attractive hydrophobic interaction between the hydrophobic groups prevails over the electrostatic repulsion between the charged groups, the hydrophobic groups can self-aggregate intramolecularly to form a micelle-like structure in aqueous solution. This compact conformation is reflected by a sharp decrease in solution viscosity. Such intramolecular micellization largely depends on chemical structure and composition (7). This effect has been observed at low polymer concentrations. In this case, macromolecular coils behave as individual species and intramolecular interactions are almost exclusively favored. These lead to more compact, shrunken conformations. Furthermore, the effects become more marked as ionic strength increases because progressive screening of the repulsive charges overcomes their opposite expansion-generating effect (7), (23). However, at high concentration solutions, the coils are entangled, favoring intermolecular hydrophobic interactions, which lead to the formation of a three-dimensional network of polymer chains resulting in a rapid increase in apparent viscosity (1), (8), (16), (23). Moreover, the balance between the intra or intermolecular interactions depends on the structural parameters of the polymer, such as the nature, length and content of the hydrophobic groups, their distribution along the backbone, hydrophilic segment length and hydration, charge density, degree of polymerization and polymer concentration (2-5), (7-9), (11), (18-20) and on environmental parameters such as salinity (3), (8), (11), (16), pH (1), (18), organic solvent (11), and temperature (3), (8), (18). For this reason, unmodified and hydrophobically modified polymers are very different.

One of the aims of the present study was to examine the formation of the intra or intermolecular interactions of a series of hydrophobically modified carboxymethylcelluloses (CMC) containing different amounts of the N,N-dihexyl-acrylamide (1 to 6 mol %) in aqueous solutions. CMC is a chemically modified cellulose derivative with high water solubility, widely used for its low cost, lack of toxicity, and biodegradability (9), (18). The solution behavior of the modified polymers is expected to result from a balance between intra and intermolecular interactions due mainly to electrostatic repulsions and hydrophobic attractions. The relative contribution of these interactions was investigated as a function of the content of the hydrophobic group (N,N-dihexylacrilamide), polymer concentration, salt, and temperature. Another aim of this study was to evaluate the effect of introducing hydrophobic groups of various lengths on the solution behavior of copolymers and on temperature tolerance.

EXPERIMENTAL

Materials

The commercial sample of sodium carboxymethylcellulose (CMC) was provided by Petrobras S/A, Brazil. Its weight-average molar mass of 6.8 X [10.sup.5] g/mol was determined by static light scattering in 0.5 M NaC1 aqueous solution at 25[degrees]C. The content of carboxyl groups was determined by (1)H NMR (24) and was found to be 1.05 carboxyl groups per anydroglycose unit (DS = 1.05). Prior to use, the CMC was purified by polymer dissolution in [10.sup.-2] mol/L (0.58 g/L) NaCl aqueous solution under constant stirring at room temperature for at least two days. After complete dissolution of the polymer, the solution was centrifuged (15 min, 15,000 rpm, 25[degrees]C), filtered through a 3-[micro]m Millipore membrane, and lyophilized. Acryloyl chloride (Merck), N,N-dihexylamine (Merck), tetrahydrofuran (THF) (Merck), sodium dodecyl sulphate (SDS) (Labsynth, 90% purity), potassium persulphate ([K.sub.2][S.sub.2][O.sub.8]) (Vetec), and acetone (AQEEL) were used without further purification.

Synthesis of N,N-Dihexylacrylamide Monomer

The N,N-dihexylacrylamide was prepared by a reaction of acryloyl chloride with N,N-dihexylamine, according to the procedure described in the literature for N-alkylacryla-mide monomers (4), (25). In a 250-mL three-necked round-bottom flask equipped with a thermometer, a condenser, a magnetic stirrer, and addition funnel, the N,N-dihexylamine (0.199 mol) was dissolved in 100 mL of tetrahydrofuran (THF) and placed in the flask. The solution was then cooled to 0[degrees]C. Acryloyl chloride (0.099 mol) dissolved in 40 mL of THF was added slowly to the reaction flask over a period of 2.5 h at such a rate that the temperature did not exceed 5[degrees]C. The resulting mixture was then stirred at 10[degrees]C for 2 h. The resulting product was filtered to remove the N,N-dihexylamide hydrochloride, and the THF was removed by vacuum distillation at 60[degrees]C. At the end of this process, a yellow viscous liquid hydrophobic monomer was obtained.

Synthesis of CMC-g-Poly(N,N-Dihexylacrylamide) Copolymers

The CMC-g-poly(N,N-dihexylacrylamide) copolymers were prepared by an aqueous micellar copolymerization technique between CMC and N,N-dihexylacrylamide. This copolymerization reaction was adapted from the copolymerization method used for acrylamide monomers described in the literature (4), (5), (19), (20). The associating copolymers were obtained by inserting a low amount (1 to 6 mol %) of N,N-dihexylacrylamide into the CMC chain. They were obtained through a micellar radical copolymerization in water, with sodium dodecyl sulfate (SDS) as the surfactant and potassium persulfate ([K.sub.2][S.sub.2][O.sub.8]) as the initiator. The presence of the surfactant allows the solubilization of the hydrophobic monomer within the micelles dispersed in the water. The initial concentration of CMC and N,N-dihexylacrylamide in water (total polymer and hydrophobic monomer) was 3% (w/w) (1.2 X [10.sup.-1] mol/L) and the initiator concentration was 0.03% (w/w) (1.12 X [10.sup.-3] mol/L) relative to the monomer feed. The surfactant concentration was 1.5% (w/v) (5.2 X [10.sup.-2] mol/L) in the feed, based on the water volume. The temperature was fixed at 50[degrees]C. The copolymerization was carried out in various amounts of hydrophobic monomer ranging from 1 to 6 mol %.

The sequence distribution of the hydrophobic monomer in the polymer chain depends mainly on the initial number of hydrophobic monomers per micelle, [N.sub.H] [20], calculated as follows:

[N.sub.H] = [[M.sub.H]]/[SDS] - cmc/[N.sub.agg] (1)

where [[M.sub.H] is the initial molar concentration of N,N-dihexylacrylamide, [SDS] the molar surfactant concentration, cmc its critical micellar concentration, and [N.sub.agg] its aggregation number ([cmc.sub.SDS] = 9.2 X [10.sup.-3] mol/L and [N.sub.agg] = 60 at the polymerization temperature of 50[degrees]C) (20), (26). [N.sub.H] values can be modified by varying the initial concentration of the hydrophobic monomer or the concentration of SDS (4). Keeping the hydrophobic concentration constant and decreasing the surfactant concentration leads to an increase in [N.sub.H], which results in an increase in the length of the hydrophobic pendant group on the polymer chain (20).

Polymer characteristics are given in Table 1. The sample code of the copolymers refers to the polymer precursor and to the initial concentration of the hydrophobic monomer (HM1) used in the copolymerization reaction. For example, CMCHM1 refers to a CMC copolymer containing approximately 1 mol % of N,N-dihexylacrylamide.

TABLE 1. Initial molar concentration of hydrophobic monomer
N,N-dihexylacrylamide ([M.sub.H]), initial number of hydrophobic
monomers per micelle ([N.sub.H]), intrinsic viscosity, and Huggins
constant of the CMC and CMC-g-poly(N,N-dihexylacrylamide) in 0.5
NaCl aqueous solution and synthetic formation water (SFW), at
25[degrees]C.

                                            0.5 M NaCI
        [[M.sub.H]]
Sample    (mol %)    [[N.sub.H]]  [[eta]] (mL/g)   K'
CMC         --           --            336        0.20
CMCHM1     1.12         1.96           333        0.67
CMCHM3     3.36         5.90           334        0.47
CMCHM6     5.94        10.41           323        0.45

                    SFW (a)

Sample  [[eta]] (mL/g)  K'

CMC           577       0.19
CMCHM1        797       0.33
CMCHM3        724       0.34
CMCHM6        622       0.33

(a) SFW, Synthetic formation water (5.4 X [10.sup.-3] M NaCl,
4.9 X [10.sup.-4] M Ca[Cl.sub.2] and 7.3 X [10.sup.-5]
M Mg[Cl.sub.2]).

Sample Preparation

Polymer solutions were prepared by dissolving the desired polymer concentration in distilled water, Milli-Q water, and synthetic formation water (5.4 X [10.sup.-3] M NaCl, 4.9 X [10.sup.-4] aqueous solution under gentle stirring at room temperature for at least three days. The ionic strength of the solutions is estimated to be 6.83 X [10.sup.-3] for synthetic formation water and 0.5 for 0.5 M NaCl aqueous solution.

Viscosity Measurements

Viscosities of the unmodified CMC and the hydrophobic-modified CMC solutions were measured using an Ubbelohde capillary viscometer at a temperature of 25 [+ or -] 0.05[degrees]C maintained with a thermostatic bath. Stock solutions were prepared by dissolving the polymers in distilled water, synthetic formation water, and in 0.5 M NaCl solutions. In distilled water, the concentration of the stock solution used was equal to 1.0 g/L for the copolymer and 2.5 g/L for the unmodified CMC, while in synthetic formation water and 0.5 M NaCl the polymer concentration was 2.5 g/L for all the polymers. The solvents and the stock solutions were then filtered through a 0.22- and 0.45-[micro]m Millipore type membrane for light-scattering measurements, respectively.

The reduced viscosity was calculated as the ratio (t--t.sub.0])/([t.sub.0]c), where [t.sub.0] and t are the flow times of the solvent and of the polymer solution at concentration c, respectively. The intrinsic viscosity [[eta]] was determined from the extrapolation of the reduced viscosity curve to zero concentration using the Flory-Huggins equation as shown below:

[[eta].sub.red] = [[eta].sub.sp]/c = [[eta]] + k'[[eta]].sup.2]c (2)

where k' is the Huggins coefficient, a constant for a series of polymers of different molecular weights in a given solvent and temperature.

Light Scattering Measurement

Static light scattering (SLS) and dynamic light scattering (DLS) experiments were performed by ALV laser goniometer, which consists of a 22-mW HeNe linear polarized laser operating at a wavelength of 632.8 nm and an ALV-5000/EPP multiple [tau] digital correlator with 125 ns initial sampling time. Data were acquired using ALV Correlator Control software. For the SLS experiments, the polymer solutions were maintained at a constant temperature of 25 [+ or -] 0.1[degrees]C and for the DLS experiments the temperature of the polymer solutions varied between 25 and 65[degrees]C. The minimum sample volume of 2 mL was used in the SLS and DLS experiments and the volume of the solution was placed in 10 mm diameter glass cells. Before the SLS and DLS measurements, all solvents and polymer solutions were filtered through 0.22-[micro]m and 0.45-[micro]m Millipore filters, respectively, to remove dust and any non-dissolved particles.

In the SLS experiments, the time average light scattering intensity by the sample (I) is recorded as a function of the scattering wave vector q = (4[pi]n/[lambda]sin([theta]/2) (27), where n is the refractive index of the solvent (1.33 for the water at 25[degrees]C), [lambda] is the wavelength of the incident light in the vacuum, and [theta] is the scattering angle. In the SLS experiments one measures the excess of scattering intensity I(q) with respect to the solvent. In our experiments, the scattering angle [theta] varied from 45 to 150[degrees] with a 5[degrees] stepwise increase. Toluene was used as a calibration standard. Weight-average molar mass [M.sub.w], z-average radius of gyration [R.sub.g], and second virial coefficient [A.sub.2] were determined by double extrapolation using the Zimm method, based on the following relation:

Kc/[DELTA][R.sub.[theta] = 1/[M.sub.w][1 + [q.sub.2][R.sub.g.sup.2]/3] + 2[A.sub.2]c (3)

where c is the polymer concentration, [DELTA][R.sub.q] is the excess absolute time-average light scattering intensity (excess Rayleigh ratio), and K = 4[[pi].sup.2] [n.sup.2] (dn/dc) (2) /[N.sub.A] [[lambda].sup.4] is a scattering optical constant, where [N.sub.A] is Avogadro's number and dn/dc is the refractive index increment. The dn/dc of the unmodified CMC and hydrophobic-modified CMC in aqueous solution at a wavelength of 632.8 nm was established at 0.163 mg/L and this value did not depend on DS or on [M.sub.w] (28).

Molecular parameters [M.sub.w], [R.sub.g], and [A.sub.2] were determined in 0.5 M NaCl aqueous solution. The SLS measurements were carried out in a dilute regime and the concentration of the polymer solution varied from 0.5 to 2.0 g/L. These concentrations were chosen in such a way that the overlap concentration was not exceeded, but large enough to obtain a measurable light scattering intensity.

In the DLS experiment, the fluctuation of the scattering light because of the Brownian motion of the particles is analyzed in terms of an autocorrelation function, which contains the distribution of relaxation times, [tau], and scattering amplitudes of the examined components (27). In the DLS experiments, the homodyne intensity autocorrelation function [G.sup.2] (t) was measured within the range of delay times from 1 X [10.sup.-4] to 10 X [10.sup.4] ms. The homodyne intensity autocorrelation function [g.sup.2](t) is related to the electric field time autocorrelation function [g.sup.1] (t), according to the Siegert relation (29):

[g.sup.2](t) = 1 + B|[[g.sup.1](t)|.sup.2] (4)

where [g.sup.2] (t) = [G.sup.2] (t)/[G.sup.2] ([infinity]), [G.sup.2] ([infinity]) is an experimentally determined baseline, B ([less than or equal to]1) is a coherence factor depending on the geometry of the detection and the ratio of the intensity scattered by the polymer to that scattered by the solvent. In a dilute solution of monodisperse particles, the field autocorrelation function is connected with the translation diffusion coefficient D as follows:

[g.sup.1](t) = exp(-t/[tau]) = exp(-[GAMMA]t) = exp(-D[q.sup.2]t) (5)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)

where [tau] is the decay time of the time autocorrelation function of the scattered electric field [g.sup.1] (t), [GAMMA] = 1/[tau] is the decay rate, and q is the wave vector. If there is a large number of independent decay processes in the system, [g.sup.1] (t) is a weighed sum of individual contributions. Generally, [g.sup.1] (t) may be expressed by a continuous distribution of decays:

[g.sup.1](t) = [[integral].sub.0.sup.[infinity]]A([tau])exp(-t)/[tau])d[tau] (7)

where A([tau]) is a distribution of scattered light over decay times. The CONTIN program (30) based on the inverse Laplace transform was used to determine this distribution function. From the relaxation time, diffusion coefficient D and hydrodynamic radius [R.sub.H] were calculated using relations D = 1/[q.sup.2][tau], D = kT/6[pi][eta][R.sub.H] (Stockes-Einstein Equation), where k is the Boltzmann constant, T is the absolute temperature, and [eta] is the viscosity of solvent at the same temperature. The true values of D and [R.sub.H] were obtained by extrapolation to q = 0 and c = 0.

The hydrodynamic radius [R.sub.H] values of the CMC and CMC copolymers were determined in Milli-Q water, synthetic formation water or aqueous 0.5 NaCl solutions at different polymer concentrations (1, 2, 3 and 4 g/L) and at different temperatures (25, 45, 55 and 65[degrees]C). The measurements in the DLS experiments were performed at 60, 90, and 120[degrees]. To avoid excessive information, we only show the results at 90[degrees], since they are representative of all the light scattering angles analyzed.

In the low concentration range, the autocorrelation function of scattered electric field [g.sup.1] (t) is a single exponential, whereas at higher concentrations, [g.sup.1] (t) can be described by a sum of two relaxations widely separated in time, with [A.sub.f] + [A.sub.s] = 1.

[g.sup.1](t) = [A.sub.f]exp(-1/[[tau].sub.f]) + [A.sub.s]exp[[ - (t/[[tau].sub.se]).sup.[beta]]] (8)

The parameters [A.sub.f] and [A.sub.s] are the amplitudes for the "fast" and "slow" relaxation modes, respectively. Analyses of the time correlation functions in the semidilute concentration regime (where q[R.sub.H] < 1) have shown that the first term (short-time behavior) on the right-hand side of Eq. 8 is related to a collective diffusion coefficient [D.sub.c] ([tau].sub.r.sup.-1] = [D.sub.c][q.sup.2], where [[tau].sub.f] is the "fast" relaxation time), which reflects a concerted motion of the polymer chains relative to the solvents. The second term (long-time behavior) is expected to be associated with disengagement relaxation of individual chains (1), (31) or cluster relaxation (31), (32). The variable [[tau].sub.se] is some effective slow relaxation time, and the stretched exponent [beta](0 < [beta] [less than or equal to] 1) is an indication of the width of the distribution of relaxation times. In this work, the value of [beta] depends on the hydrophobicity of the polymer. The mean slow relaxation time is given by

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (9)

where [GAMMA] ([beta].sup.-1] is the gamma function of [beta].sup.-1].

Rheological Measurements

Rheological experiments were carried out using a Haake RheoStress RS150 rheometer equipped with DG41 coaxial cylinder sensor. Solution viscosities of the unmodified and modified polymers with 1 and 3 mol % of the hydrophobic groups (CMCHM1 and CMCHM3) at 10 g/L polymer concentration were measured at shear rates ranging from 0.1 to 100 [s.sup.-1] at temperature T = 55[degrees]C.

RESULTS AND DISCUSSION

Synthesis and Characterization of the Hydrophobic Monomer N,N-Dihexylacrylamide

The hydrophobic monomer N,N-dihexylacrylamide used in this study was chosen based on literature data (19), which present evidence that N,N-disubstituted acrylamide derivatives used in micellar copolymerization favor the obtaining of copolymers with a more homogeneous chemical structure than that obtained for N-monosubstituted acrylamide monomers. The homogeneity of the copolymer structure is related to the fact that the N,N-disubstituted acrylamide monomers do not form mtermolecular hydrogen bonding due to the absence of the N--H amino group. Furthermore, at the lower copolymer concentration, it is expected that the presence of the two hexyl chains, in the case of N,N-dihexylacrylamide monomer, with six carbon atoms in each chain and two terminal methyl groups, will make the polymer more hydrophobic, favoring strong associative properties and a significant viscosity increase of the copolymer solutions in water. In this study, the characterization of N,N-dihexylacrylamide monomer microstructure by [.sup.1] NMR in CDC[l.sub.3] (not shown here), as previously described (33), was similar to that reported by Volpert et al. (19).

Synthesis and Characterization of the Copolymers CMC-g-poly(N,N-dihexylacrylamide)

The polymers were obtained by micellar copolymerization of carboxymethylcellulose with the length of hydrophobic monomer N,N-dihexylacrylamide ranging from 1.1 to 5.9 mol %, as shown in Table 1. The sequence distribution of the hydrophobic monomers in the copolymers was controlled by varying the number of hydrophobes per micelles, [N.sub.H], and it was assumed that the hydrophobe content corresponds to the initial feed composition. In this study, the length of the hydrophobic chain grafted onto the CMC backbone was increased by increasing the hydrophobic monomer concentration, making the surfactant concentration constant in the medium, i.e., the [N.sub.H] values varied by modifying the initial concentration of hydrophobic monomer, [[M.sub.H]]. Polymerization was carried out in an [N.sub.H] range from 1.96 to 10.41 (Table 1). For associating polyacrylamides, the micellar copolymerization process allows one to control the microstructure of the resulting copolymer by varying the hydrophobe to micelle ratio going from a very blocky structure (high [N.sub.H] values) to an almost random copolymer ([N.sub.H] = 1) [19]. It must be emphasized that, in the case of CMC copolymers, the hydrophobic groups are grafted onto the CMC backbone. In this case, it was observed that the [N.sub.H] values were approximately equal to or greater than 2, so that the N,N-dihexylacrylamide monomers are grafted onto the CMC of the hydrophobic groups having different lengths. In this condition, the hydrophobic association in the hydrophobically modified polymers increase as [N.sub.H] increases. On the other hand, high [[M.sub.H]] and consequently high [N.sub.H] values can make the copolymers either partially or completely insoluble in water. All copolymer solutions used in this study were turbid and this is due to the presence of the hydrophobic groups, and it was observed that this turbidity increased with the length of the hydrophobic groups and polymer concentration in the medium. In spite of this, after several months under equilibrium conditions, no precipitation was observed in the polymer solutions. This result seems to indicate that all the polymers studied have good solubility in aqueous medium.

NMR spectroscopy is a commonly used method for determining copolymer structure. Nevertheless, in the present case, the technique is not sensitive enough owing to the very low hydrophobe content. However, information on copolymer microstructure was indirectly obtained from a study on their solution properties of viscosity, static and dynamic light scattering and rheological measurements, as will subsequently be shown. For hydrophobically modified polyacrylamide copolymers, the very low percentage of hydrophobic monomers leads to difficulties in evaluating its incorporation level in the polymer. For this reason, the hydrophobic monomer level is assumed to be equal to the initial feed composition (5), (19), (25). This probably leads to some error because the composition of hydrophobic copolymers is prepared from the micellar process. Some studies using N-disubstituted hydrophobic acrylamides as hydrophobic comonomers in the synthesis of the associating polyacrylamides prepared by micellar copolymerization indicate that the copolymers with hydrophobe content lower than 1 mol % do not allow the determination of composition by [H.sup.1] NMR, since the end methyl signal is too low, making this method insufficiently sensitive, or by UV spectroscopy because these copolymer samples do not contain phenyl or other aromatic groups (2), (5). In this study, hydrophobe content was larger than that usually used in the synthesis of hydrophobically modified polyacrylamides; however, this content was much lower than that generally used in the chemical modification of polysaccharides (9). For this reason, we were unable to determine the copolymer microstructure by NMR.

Viscosity Measurements

The variation of the reduced viscosity of the CMC and CMC-g-poly(N,N-dihexylacrylamide) in distilled water at 25[degrees]C against the polymer concentration is plotted in Fig. 1. In distilled water, the reduced viscosity reached a maximum for all systems, but this maximum was greater for hydrophobically modified polymers than that for the polymer precursor, suggesting an increase in viscosity as the length of the hydrophobic chain grafted onto the CMC backbone increases. For the unmodified CMC, this behavior can be explained by the electrostatic repulsions between charged carboxylic groups located along the CMC backbone, showing typical polyelectrolyte behavior (34). For the hydrophobically modified polymers, this phenomenon can be explained either by the flexibility of the chain or by the decrease in the solvent quality, which contributes to the increase or decrease of the reduced viscosity of these polymers in solution. In the case of flexible polymer chains, such as hydrophobically modified polyacrylamides, the decrease in the reduced viscosity occurs because of the strong intramolecular hydrophobic interactions, which promote the polymer-polymer interactions in aqueous medium due to the decrease in the solvent quality. This behavior suggests that the flexibility of the polymer backbone facilitates the entanglement of the polymeric chain. However, in the case of semi-rigid polymer chains such as CMC, the increase in the reduced viscosity for the hydrophobically modified CMC copolymers occurs because aggregate formation was favored by the decrease in the solvent quality (35). In this case, the rigidity of the polymer chain hinders the entanglement of the polymer chain, contributing to aggregate formation and consequently to the increase in reduced viscosity of these polymers in solution. Moreover, the electrostatic repulsions between charged carboxylic groups (CO[O.sup-]) located along the CMC hydrophilic chain also contribute to the increased hydrodynamic volume in distilled water. However, the presence of hydrophobic groups seems to contribute more significantly to the increase in the reduced viscosity of the copolymer solutions. This effect can be observed with the increased length of the hydrophobic group grafted onto the CMC backbone, as shown in Fig. 1.

[FIGURE 1 OMITTED]

Table 1 shows the values of intrinsic viscosity and Huggins constant for the CMC and CMC-g-poly(N,N-dihexylacrylamide) copolymers, in 0.5 M NaCl aqueous solution and synthetic formation water (SFW), at 25[degrees]C. In the 0.5 M NaCl, the copolymers had a tendency to lower intrinsic viscosity compared to CMC. In the presence of salt, two effects can contribute to the contraction of the polyelectrolyte associative coil in solution: (i) screening of electrostatic interactions and (ii) intramolecular hydrophobic interactions. The latter is promoted by decreasing the polymer-solvent interactions, favored by hydrophobic groups, and explaining the increase in the Huggins constant values. However, only a slight increase was observed in these values for the copolymers. The Huggins constant values for copolymers remained in the 0.3-0.7 range expected for non-associating polymers. This behavior can be explained by the very small content of the hydrophobic groups grafted onto the semi-rigid CMC backbone, favoring a weak reduction in the polymer-solvent interactions. In SFW, the values of intrinsic viscosity for copolymers were higher than that for unmodified CMC. This is due to intra and/or intermolecular hydrophobic interactions promoted by the salts (NaCl, Mg[Cl.sub.2], and Ca[Cl.sub.2]) present in the synthetic formation water (SFW) even at lower concentrations and carboxilate groups ([COO.sup.-]) in the CMC chain. Moreover, for copolymers, the slightly contraction of the polymer coil became more evident as the length of the hydrophobic groups grafted onto the CMC backbone increased. Thus, the copolymer with the largest size of the hydrophobic groups (CMCHM6) had the lowest value of intrinsic viscosity, suggesting strong intramolecular hydrophobic interactions. This effect was also observed for this copolymer in SFW. This behavior is in agreement with the results reported for CMCHMs with a small content of the hydrophobic groups (9).

The values of intrinsic viscosity of the copolymers in SFW were greater than that of the polymer precursor and in 0.5 M NaCl. The larger hydrodynamic volume can be explained by the interactions between the carboxylic groups (CO[O.sup.-]) in the hydrophilic CMC backbone and the divalent cations (C[a.sup.2+] and M[g.sup.2+]) present in the synthetic formation water. In this condition, the negatively charged groups in the CMC backbone are available to interact with the divalent cations present in the synthetic formation water. Moreover, this result suggests that in this solvent the interactions between charged groups of opposite charges are more significant than the intramolecular hydrophobic interactions. However, the contraction of the polymer coil, likely promoted by hydrophobic intramolecular associations, was also observed in the copolymers by their intrinsic viscosity reduction. Furthermore, the Huggins constant values were maintained around 0.3, suggesting good interaction between the macromolecules and the solvent. This behavior could explain the stability of the solutions, i.e., the absence of the precipitate as a function of time.

The difference between the values of intrinsic viscosity for unmodified CMC in 0.5 M NaCl and SFW can be explained by the ionic strength of the solvents. The hydrodynamic volume of the unmodified CMC decreases as ionic strength increases owing to the reduction of electrostatic interactions and consequently intrinsic viscosity. This effect was more significant in the solvent of higher ionic strength (0.5 M NaCl). However, the values of the Huggins constant not changed in both solvents, indicating good interaction between the unmodified CMC chain and the solvents.

Static Light Scattering

Physicochemical characteristics such as weight-average molar mass ([M.sub.w]), radius of gyration ([R.sub.g]) and the second virial coefficient ([A.sub.2]) of unmodified CMC and CMCHMs with different amounts of the hydrophobic group (1, 3 and 6 mol % of the initial feed composition), obtained from static light scattering by Zimm plot in 0.5 M NaCl, at 25[degrees]C, are presented in Table 2. The weight-average molar mass of the CMCHM copolymers is a parameter of utmost importance since it controls the viscosity of the polymer solutions, mainly in the semidilute solutions. In this study, a 0.5 M NaCl aqueous solution was used as solvent for the polymers to screen the electrostatic interactions between the carboxylate groups. Polymer concentrations of 0.5, 0.8, 1.2, 1.6, and 2.0 g/L were chosen to ensure a measurement of light scattering intensity. Although we could not determine critical viscosity values, i.e., the transition from the dilute to semi-dilute regime, owing to the transition from the dilute to semi-dilute regime, owing to the low polymer concentrations used, we relied on literature data to ensure that our study was carried out in a dilute regime, not exceeding the overlap concentration for the unmodified CMC of similar degree of substitution (DS = 0.9) and molar mass ([M.sub.w] = 3 X [10.sup.5] g/mol) investigated by Charpentier et al. (9). In the copolymer case, some concentrations were slightly higher than the critical hydrophobe concentration of the CMCHM grafted with 6% hexadecylamine (HDA) ([C.sub.16]) = ([C.sub.cr] = 1.5 g/L) [9] and of the bigrafted CMC with 30% [C.sub.4] and 3% [C.sub.12-18] ([C.sub.cr] = 1.4 g/L). Nevertheless, they were lower than those of the monografted CMC sample containing 9% [C.sub.12-18] ([C.sub.cr] = 2.5 g/L) [21]. In the two latter studies, the copolymer samples had larger molar mass (4.8 X [10.sup.6] and 4.9 X [10.sup.6] g/mol, respectively) than those of the CMCHM samples studied in this work (Table 2). It is likely for this reason that the CMCHM copolymers studied here had higher critical concentrations. Therefore, it is believed that all copolymer concentrations used should be below the critical concentration values observed by Charpentier-Valenza et al. [21]. It must be emphasized that, in the case of CMC copolymers, the transition from a dilute to semidilute regime occurs not only because of the close packing, but also because of the specific association between the coil and/or aggregates. Hence, it is more appropriate to use the term critical hydrophobe concentration than overlap concentration for the hydrophobically modified copolymers. A similar behavior was observed by a number of researchers [5, 10, 21]. In addition, the static light scattering study must be carried out in dilute polymer concentrations to guarantee the macromolecule disassociate state in solution. Thus, the [M.sub.w] obtained is just a single molecule. However, to determine the [M.sub.w] of the aggregates, it is suggested that higher polymer concentrations be used [36].

TABLE 2. Molar mass ([M.sub.w]), radius of gyration ([R.sub.g]), and
the second virial coefficient ([A.sub.2]) of unmodified CMC and
CMC-g-poly(N,N-diheylacrylamide) copolymers with different lengths
of the hydrophobic groups obtained from static light scattering by
Zimm plot (dn/dc = 0.163 mL/g) in 0.5 M NaCl at 25[degrees]C.

Sample  [M.sub.w] (g/mol)  [R.sub.g] (nm)  [A.sub.2] (mol.mL/[g.sup.2])

CMC     6.8 x [10.sup.5]       125.4            5.2 x [10.sup.-4]
CMCHM1  7.3 x [10.sup.5]       139.8            2.4 x [10.sup.-4]
CMCHM3  8.9 x [10.sup.5]       82.88           -2.5 x [10.sup.-4]
CMCHM6  5.9 x [10.sup.5]       96.66            7.1 x [10.sup.-5]

It can be seen from Table 2 that the molar mass of the hydrophobically modified CMC with 1 or 3 mol % of N, N-dihexyacrylamide was slightly larger than that of the unmodified CMC; however, the [M.sub.w] of the copolymer with 6 mol % of hydrophobic monomer was slightly lower. This result seems to indicate the existence of aggregating structures. This behavior agrees with the results reported by Charpentier-Valenza (21). The increase of [M.sub.w] is confirmed by the increasing light scattering intensity. It occurs because light scattering intensity is proportional to polymer molar mass (37). The greater the copolymer molar mass, the greater the light scattering intensity. Moreover, an increase in [M.sub.w] is observed as the hydrophobic. group length increases. However, the lowest [M.sub.w] value was observed for the copolymer of the hydrophobic groups with the largest length (CHCHM6), indicating strong intramolecular associations that induce macromolecular compactness. This occurred because the solvatation of the polymer chain decreased because of the increasing hydrophobe content. These results clearly show the presence of intramolecular hydrophobic interactions resulting from the largest content of the N,N-dihexylacrylamide groups in 0.5 M NaCl. This behavior agrees with results reported in the literature (37). In contrast, for all copolymers, the [R.sub.g] values decrease as the length of hydrophobic groups increases. Furthermore, they are lower than that of the precursor polymer, except for the one with 1 mol % of hydrophobic groups. This result seems to indicate that the hydrophobic interactions are favored by a reduction in the solvent quality and this effect becomes more significant with the increase in hydrophobic group length grafted onto the CMC backbone. This behavior suggests that the greater the hydrophobic group length, the stronger will be the intramolecular hydrophobic interactions and consequently the greater will be the tendency to contraction of the polymeric chain into a more compact coil when compared to unmodified CMC. It is also shown in Table 2 that all the copolymers had lower [A.sub.2] values than that of unmodified CMC, suggesting an increase in polymer-polymer interactions because of the reduction of the solvent quality caused by the presence of the hydrophobic groups grafted onto the CMC backbone in salt solution, which, as previously suggested, favor aggregate formation. However, only CMCHM3 had negative [A.sub.2] value, suggesting easier hydrophobic aggregate formation as a result of the reduction in polymer-solvent interactions in the presence of salt. Moreover, this copolymer also had the lowest [R.sub.g] and highest [M.sub.w] values. This result clearly indicates that this behavior is related to hydrophobic chain length, considering the solvent effect in the hydrophilic CMC backbone, which will be able to either facilitate or hinder the hydrophobic aggregate formation.

Dynamic Light Scattering

The DLS experiments were carried out at different polymer concentrations (1, 2, 3, and 4 g/L) and scattering angles (60[degrees], 90[degrees], and 120[degrees]) in Milli-Q water, synthetic formation water (SFW) and 0.5 M NaCl aqueous solution, at 25, 45, 55, and 65[degrees]C. The ionic strength of the two latter solutions was 6.83 X [10.sup.-3] and 0.5, respectively. However, we will show the results of typical autocorrelation function and distribution function of the hydrodynamic radius obtained by CONTIN analysis for two polymer concentrations (1 and 3 g/L) in 0.5 M NaCl, at a scattering angle of 90[degrees], at temperatures of 25 and 65[degrees]C, because these conditions represent the overall behavior observed for the copolymers studied.

Effect of Polymer Concentration and Solvent. Figure 2 shows the effect of polymer concentration and solvent on the typical autocorrelation function (see Fig. 2a and b) and the distribution function of the hydrodynamic radius (see Fig. 2c and d) for unmodified CMC and hydrophobically modified CMC with different lengths of the N,N-dihexylacrylamide groups in 0.5 M NaCl. The most important feature of the decays is the presence of long-time tails of the autocorrelation function for all polymers under the conditions analyzed (Fig. 2a and b). This suggests the existence of aggregates. This behavior agrees with the results reported by Kj[empty set]niksen et al. for hydrophobically modified EHEC at different levels of SDS addition (31). It is important to emphasize that the aggregates observed in the unmodified CMC backbone are hydrophilic and produced by hydrogen bonding. In addition, the size of these aggregates decreases as ionic strength increases owing to the reduction of electrostatic interactions (Table 3).

[FIGURE 2 OMITTED]

TABLE 3. Hydrodynamic radii for the CMC and CMCHM copolymers
obtained at polymer concentrations of 1 and 3 g/L, scattering
angle of 90[degrees], and temperature of 25[degrees]C.

                        [R.sub.H polymer]    [R.sub.H aggregate]
                         (nm) (fast mode)      (nm) (slow mode)

Solvent        Sample  [C.sub.p]  [C.sub.p]  [C.sub.p]  [C.sub.p]
                         = 1g/L     = 3g/L     = 1/L     = 3 g/L

Milli-Q Water  CMC         --         --        179        620
               CMCHM1      66         75        620        796
               CMCHM3      85         35        902       1310
               CMCHM6      40         85        484        548

SFW            CMC         17         --        294        213
               CMCHM1      35         85        377        901
               CMCHM3      21         31        484        703
               CMCHM6      45         24        294        548

0.5 M NaCl     CMC         35         21         --        294
               CMCHM1      24         24        549        294
               CMCHM3      17         40        703        377
               CMCHM6      17         19        333        202

An inspection of the autocorrelation function for the CMC and CMCHM systems shows that the relaxation process for the copolymers occurred at longer times than that of the unmodified CMC, i.e., the copolymers take longer to relax than the precursor polymer does. This probably occurs because of the hydrophobic aggregate on the copolymer chain, making the molecule denser and relaxation difficult. However, the shift towards a more prolonged relaxation process is more pronounced for a copolymer solution containing 3 mol % of the hydrophobic groups (CMCHM3), indicating that this copolymer needs more time to relax than the copolymers with 1 mol % (CMCHM1) or 6 mol % (CMCHM6) of N,N-dixehylacrylamide groups. This behavior may be related to the size of the aggregate formed. The greater the aggregate size, the longer time the polymer chain needs to relax. This result also clearly indicates that the copolymer of the hydrophobic groups with the largest length (CMCHM6) forms the smallest hydrophobic aggregates and consequently needs the shortest relaxation process. This behavior suggests the presence of strong intra and/or intermolecular hydrophobic interactions for this copolymer considering both regimes analyzed: dilute regime (Cp = 1 g/L) and semidilute regime (Cp = 3 g/L).

The relaxation process for the copolymer solutions in Milli-Q water and SFW (not shown here) occurred over slightly longer periods, with the increase of polymer concentration. This result seems to indicate that the extension of the hydrophobic aggregates increases as polymer concentration increases, making the macromolecule more denser. In the solvent of highest ionic strength (0.5 M NaCl) (Fig. 2a and b), the copolymer solutions exhibited the inverse behavior. In this case, the relaxation process for copolymers occurred at shorter times, indicating the smallest size of the hydrophobic aggregate formed. This could be explained by the reduction of electrostatic interactions, which favors the closer polymer chain, or of solvent quality, which promotes strong intramolecular and/or intermolecular associations, suggesting the contraction of the polymeric coil due to the presence of the hydrophobic groups. This result indicates that the size of the hydrophobic aggregates is reduced with the increase of ionic strength in the medium.

The autocorrelation functions are better described by the distribution function of the hydrodynamic radius obtained from CONTIN analysis for CMC and CMCHMs with different lengths of the hydrophobic groups (1 to 6 mol %). The analysis conditions are shown in Fig. 2c and d. It can be observed from these figures that all copolymers had two relaxation modes. This comparison can be made here because these curves are similar to those obtained by distribution of the relaxation times (not shown here). The fast mode of the relaxation process occurs in a shorter time while the slow mode takes place over a longer time. The former may reflect a motion of the polymer chains relative to the solvent (1) and the latter may indicate the existence of aggregates (31). The unmodified CMC exhibited monomodal relaxation time distribution in Milli-Q water (ionic strength zero, free salt) (not shown here); nevertheless, in the presence of the salts (SFW (not shown here) and 0.5 M NaCl (Fig. 2c and d), there is a tendency to bimodal distribution, whereas the copolymers are bimodally distributed in all the solvents and polymer concentrations studied here. This behavior suggests that the polymer solutions have structures with different sizes. Moreover, polymer chain length and aggregate size depend strongly on the polymer concentration and ionic strength of the solvent, as shown in Table 3.

Table 3 also shows that the [R.sub.H] values attributed to the fast mode of relaxation (polymer chain) varied from 17 to 35 nm for CMC and from 17 to 85 nm for copolymers and that the [R.sub.H] values of the slow mode (aggregates) ranged from 179 to 620 for CMC and from 202 to 1480 nm for copolymers. The [R.sub.H] variation for the polymer chain likely occurred because the macromolecules are polydisperse, i.e., they have chains of different sizes. Moreover, the [R.sub.H] of the aggregates depends on polymer concentration, ionic strength and hydrophobic group length. The [R.sub.H] values lower than 10 nm (see Fig. 2d) may be due to the oligomers generally present in the polymer samples and the [R.sub.H] values around [10.sup.4] nm may be attributed to the analysis conditions used in the DLS experiments, where q.[R.sub.H] < 1. In this case, we can observe the relaxation of all the polymeric chain. However, when observing the system at q.[R.sub.H] > 1, one observes the relaxation of macromolecular and internal rotations.

The [R.sub.H] values shown in Table 3 also indicate that, in general, for the fast relaxation mode (polymer chain), [R.sub.H] increases with increasing polymer concentration and decreases with increasing ionic strength in the medium. Nevertheless, for the low relaxation mode (aggregates), [R.sub.H] increases with higher polymer concentration in Milli-Q water and SFW, but decreases in 0.5 M NaCl. This occurs as a result of the reduction in electrostatic interactions and solubility promoted by solvent of high ionic strength, suggesting strong intra and/or intermolecular associations. Furthermore, this effect was more significant with the increased polymer concentration (3 g/L), owing to the highest concentration of polymer chains in the medium, promoting closeness between the molecules and the formation of larger aggregates.

An interesting feature can be observed from the results of [R.sub.H], shown in Table 3. The CMCHM3 copolymer had the largest [R.sub.H] of the aggregates in all the polymer concentrations and solvents studied here, except in SFW at 3 g/L. However, the lowest [R.sub.H] of the aggregates was observed for the CMCHM6 copolymer, with the longest hydrophobic chain length of all the conditions analyzed. These results seem to indicate that up to 3 mol % of hydrophobic groups show equilibrium between enhanced hydrophobic aggregate size and reduced solvent quality. The polymer-polymer interactions do not seem to interfere in the aggregate formation, but above a certain concentration of hydrophobic groups, the polymer-polymer interactions become more significant. In this case, the intramolecular associations prevail over the intermolecular ones.

Temperature Effect. Figures 3 and 4 illustrate the effect on hydrophobic groups in the copolymers with the increase in temperature obtained by the typical autocorrelation function (see Figs. 3a and b, 4a and b) and the distribution function of the hydrodynamic radius (see Figs. 3c and d, 4c and d) in 0.5 M NaCl for polymer concentrations of 1 g/L (see Fig. 3) and 3 g/L (see Fig. 4). Figures 3a and b show that an increase of temperature reduces the relaxation process time. This behavior indicates that the size of aggregates decreases as the temperature increases; this takes place because of increased molecular agitation, favoring the polymer-solvent interactions and consequently polymer solubilization in the medium, but making the formation of more compact structures more difficult. For this reason, the molecule relaxes more easily. Moreover, even with the increased temperature, which favors the reduction of aggregates, the formation of major hydrophobic aggregates for the CMCHM3 copolymer was still observed. This result indicates that intermolecular associations are more significant than intramolecular ones. It also suggests that the copolymer with the largest aggregate size is more stable at higher temperatures than the other copolymers studied here, because this behavior persisted even at higher temperatures.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Figure 4a and b show that an increase in polymer concentration (3 g/L) and temperature reduced the relaxation process time in 0.5 M NaCl. However, the hydrophobic aggregates formed are larger than those at 1 g/L polymer concentration owing to the higher polymer concentration in the medium (see Table 4). This result seems to indicate that hydrophobic interactions are favored with an increase in temperature and polymer concentration in the presence of salt (0.5 M NaCl). It is well known that hydrophobic hydration is an exothermic process whereas the interaction between hydrophobic chains is an endothermic process (8). Thus, hydrophobic aggregate formation is favored by increased temperatures. This effect, together with hydrophilic chain hydration, may contribute to increasing the [R.sub.H] of the aggregates in the copolymer chains. However, at very high temperatures, molecular motion can destroy the intermolecular hydrophobic interactions (3).

An interesting effect was observed for the CMCHM1 copolymer at temperatures of 45 and 55[degrees]C (Table 4). This copolymer had a slightly higher relaxation time (or hydrodynamic radius) than that of the larger copolymer of the hydrophobic groups (CMCHM3). This effect seems to indicate that with an increase in polymer concentration and temperature, hydrophobic aggregate formation is also favored for the copolymer with the lowest content of the hydrophobic groups. This may have occurred because of increased polymer solubility in the medium and intermolecular hydrophobic interactions. However, for all polymers at 65[degrees]C (see Fig. 3d and Table 4), a shift in the relaxation process for very short times was still observed. This may have occurred because of the rupture of intermolecular associations that takes place at very high temperatures.

TABLE 4. Hydrodynamic radii for the CMC and CMCHMs copolymers
obtained in 0.5 M NaCl at different temperatures,polymer
concentrations of 1 and 3 g/L and scattering angle of
90[degrees]

                            Temperature ([degrees]C)

                         25                    45

c(g/L)  Sample  [R.sub.H]  [R.sub.H]  [R.sub.H]  [R.sub.H]
                 polymer   aggregate   polymer   aggregate

1       CMC         35          --        35          --
        CMCHM1      24          549       21         200
        CMCHM3      17          703       21         227
        CMCHM6      17          333       17         156

3       CMC         21          294       21         289
        CMCHM1      24          294       24         478
        CMCHM3      40          377       27         423
        CMCHM6      19          202       11         291

                        Temperature ([degrees]C)

                         55                    65

c(g/L)  Sample  [R.sub.H]  [R.sub.H]  [R.sub.H]  [R.sub.H]
                 polymer   aggregate   polymer   aggregate

1       CMC         30        --          33          --
        CMCHM1      27        195         26         188
        CMCHM3      23        250         42         309
        CMCHM6      21        138         18         114

3       CMC         21        249         26         213
        CMCHM1      11        527         23         213
        CMCHM3      23        466         42         396
        CMCHM6      13        186         18         166

Nevertheless, with increased polymer concentration (3 g/L), the unmodified CMC had two relaxation modes (Fig. 4c and d). The fast relaxation mode of the polymer chains and the slow mode of the hydrophobic aggregates. This behavior shows once again that CMC chemical modification took place. This result was also observed in Fig. 2d. Furthermore, we can observe more clearly that with increasing temperature, the shift in the relaxation process for aggregates occurs in less time, indicating a reduction in aggregate size.

Effect of Polymer Concentration on [R.sub.H]. Figure 5 illustrates the typical [q.sup.2] dependence on the relaxation frequency [GAMMA] for CMC and CMCHM copolymers in SFW, at polymer concentration of 1 g/L and temperature of 25[degrees]C. The [q.sup.2] values were obtained by q = (4[pi]n/[lambda]) sin ([theta]/2), where n is the refractive index of the solvent, [lambda] is the wavelength of the incident light in the vacuum and [theta] is the scattering angle. The relaxation frequency was calculated by [GAMMA] = 1/[tau], where [tau] is the maximum slow mode relaxation time obtained by G(t) curves as a function of relaxation time (not shown here), which in turn, was obtained during data treatment by decay time spectra. The coefficient angle of the relaxation frequencies for the CMC and CMCHM copolymer corresponds to the diffusion coefficient value for each system when [q.sup.2] tends to zero (Equation 6). The true [R.sub.H] values of the aggregates were obtained by D = kT/6[phi][eat][R.sub.H] (Stockes-Einstein Equation). This same route was used to determine true [R.sub.H] values for CMC and CMCHM copolymers in all polymer concentrations and temperatures used here.

[FIGURE 5 OMITTED]

The CMCHM3 copolymer had the lowest diffusion coefficient. This behavior indicates that CMCHM3 has a higher [R.sub.H] than that of its polymer precursor and other copolymers. Furthermore, the lower diffusion coefficient suggests aggregate growth (38), (39). Therefore, the lower the diffusion coefficient, the larger the hydrodynamic radius. In Fig. 5, we can observe that the other copolymers also have a diffusion coefficient lower than that of unmodified CMC, indicating that the hydrophobic aggregates are larger than the hydrophilic ones on the CMC backbone. This effect depends strongly on solvent quality and on hydrophobic group length. However, Fig. 5 shows that the [R.sub.H] value of the CMC aggregates is slightly lower than that of the CMCHM6 copolymer, while in Table 3 (1 g/L polymer concentration, 90[degrees] scattering angle, 25[degrees]C) equal values can be observed. It is important to remember that the [R.sub.H] obtained at 90[degrees] depicted in Figs. 2c and d, 3c and d, 4c and d showed only the apparent [R.sub.H] value of polymer backbone length in solution, while the [R.sub.H] value obtained by the diffusion coefficient from Stockes-Einstein Equation characterizes the real [R.sub.H] of the polymer in solution.

Figures 6a--c show the variation of the real hydrodynamic radius of the aggregates with polymer concentrations in different solvents at 25[degrees]C. The real [R.sub.H] of the aggregates increases with the polymer concentration in Milli-Q water (Fig. 6a) and SFW (Fig. 6b), but decreases in 0.5 M NaCl (Fig. 6c). This behavior also was observed in Table 3 and in Fig. 2c and d. In Milli-Q water, the increase in the real [R.sub.H] of the aggregates for the CMC solution with increasing polymer concentration is a result of two effects: the electrostatic repulsions between the carboxylate groups and the formation of intermolecular hydrophilic aggregates, which promotes the increase of the hydrodynamic volume in solution. The former is more significant at lower polymer concentrations and the latter at higher ones. In the first case, the chains are further apart from one another, in such a way that the presence of the charges hinders approximation between the chains and between the chain segments, making the molecules more stretched and more rigid. In the second case, the chains are closer to one another, in such a way that electrostatic repulsion becomes less significant. In this situation, intermolecular interactions are favored. In the case of copolymers, the increase of real [R.sub.H] of the aggregates takes place because of the electrostatic repulsion and hydrophobic aggregate formation. For this reason, the largest [R.sub.H] values are observed in Milli-Q water. Moreover, the hydrophobic aggregates are generating in this solvent owing to the strong interactions that occur between the hydrophobic groups to minimize their exposure to water.

[FIGURE 6 OMITTED]

In SFW (Fig. 6b), for all the polymer concentration ranges used here, it was observed that the copolymers had lower [R.sub.H] values than those obtained in the absence of salt. This is due to the reduction of electrostatic interactions promoted by the salts (NaCl, MgC[l.sub.2] and CaC[l.sub.2]) present in the synthetic formation water (SFW) even at lower concentrations. However, the [R.sub.H] of the aggregates increases as polymer concentration increases. This behavior was observed for CMC and CMCHM copolymers. For copolymers, this occurs because of intra and/or intermolecular hydrophobic aggregate formation that depends on solvent quality and interactions between the divalent cations and carboxylate groups. For CMC, hydrophilic aggregate formation takes place mainly owing to the reduction of electrostatic repulsions, which promotes greater interaction between the OH groups present on the CMC backbone. However, intermolecular hydrophobic and/or hydrophilic interactions are favored with increased polymer concentration.

In 0.5 M NaCl (Fig. 6c), the [R.sub.H] of the copolymer aggregates decreases with increased polymer concentration whereas that of CMC increases. For the copolymers, the salt, in addition to reducing electrostatic repulsions, may favor closeness between the polymer chains and reduce copolymer solubility, generating strong intra and/or intermolecular interactions. However, in this case, the intramolecular hydrophobic interactions seem to be more significant than the intermolecular hydrophobic ones, given the reduction of hydrodynamic volume observed. Charpentier-Valenza et al. (21) also observed reduced CMCHM copolymer solubility with increased ionic strength in the medium. They considered that this behavior was more significant for CMCHM than for amphiphilic polymers based on acrylamides, because CMCHM copolymers have lower charge density than that of associating synthetic polymers. For unmodified CMC, reduced electrostatic interactions promote the formation of intermolecular hydrophilic interactions. Thus, hydrophilic aggregate formation is more evident with higher polymer concentrations.

Another interesting factor to be observed is the effect of hydrophobic groups on the polymer backbone. With an increase in hydrophobic groups, intra and/or intermolecular hydrophobic interactions may become more significant owing to reduced solvent quality, which favors increased polymer--polymer interactions and consequently hydrophobic aggregate formation. Therefore, hydrophobic aggregate size depends strongly on the groups grafted onto the hydrophilic CMC backbone. This behavior was discussed in section Effect of Polymer Concentration and Solvent.

Temperature Effect on the [R.sub.H] of the Aggregates. Figure 7 shows the variation of the real hydrodynamic radius of the aggregates with temperature in 0.5 M NaCl at polymer concentrations of 1 g/L (Fig. 7a) and 3 g/L (Fig. 7b). An interesting effect was observed for the [R.sub.H] values of all the copolymer aggregates in the highest ionic strength solvent (0.5 M NaCl) with an increase in temperature. For polymer concentration of 1 g/L, all copolymers showed reduced [R.sub.H] of the aggregates with an increase in temperature. This result indicates that the NaCl salt in the concentration used here causes reduced copolymer solubility due to the increase of the polymer-polymer interaction favored by the hydrophobic groups. An opposite effect occurs for unmodified CMC, indicating that in a diluted regime greater polymeric chain solubility caused by increased polymer-solvent interactions seems to prevail over the reduction of electrostatic interactions favored by salt. Nevertheless, even with increasing temperature, the CMCHM3 copolymer had the highest [R.sub.H] values while CMCHM6 had the lowest. This result is shown in Fig. 6 and remained unchanged as temperature increased. In these two cases, intramolecular hydrophobic interactions prevail over intermolecular hydrophobic interactions prevail over intermolecular hydrophobic ones, but this effect was more significant for the CMCHM6 copolymer owing to its longer length of hydrophobic groups.

[FIGURE 7 OMITTED]

However, in Fig. 7b, the [R.sub.H] of the copolymer aggregates increases with higher temperatures, suggesting that intermolecular hydrophobic interactions are favored by the presence of high ionic strength salt solutions, high polymer concentrations and high temperatures. This effect was more significant for the copolymer with the shortest hydrophobic group length (CMCHM1). For this copolymer, an increase in [R.sub.H] was observed with an increase in temperature from 25 to 45[degrees]C. [R.sub.H] decreased slightly at 55[degrees]C, where intermolecular hydrophobic interactions are promoted by an endothermic process, and dropped significantly after 55[degrees]C, indicating the break of intermolecular associations due to fast molecular movement. For CMCHM3, the [R.sub.H] increases slightly with increasing temperature, suggesting only intermolecular associations and for CMCHM6 an increase of [R.sub.H] from 25 to 55[degrees]C was observed, indicating intermolecular associations. This was followed by falling [R.sub.H] after 55[degrees]C, suggesting a rupture of intermolecular hydrophobic interactions. This result suggests that the CMCHM3 copolymer is more stable thermally, since the rupture temperature had not yet been reached. Even though intermolecular hydrophobic interactions are favored by the presence of high ionic strength salt solutions, high temperatures and high polymer concentrations, in the presence of salt, the [R.sub.H] values of the aggregates were still lower than those observed in Milli-Q water and SFW (not shown here). However, [R.sub.H] decreases with an increase in temperature for the CMC, which indicates that in a semidilute regime the effect of reduced electrostatic repulsions is more significant than that of the increased polymer solubility favored by increasing temperatures.

Rheological Measurements

The influence of shear rate on the apparent viscosity of unmodified CMC and hydrophobically modified CMC with 1 and 3 mol % of hydrophobic groups at constant polymer concentration (Cp = 10 g/L) in SFW, at 55[degrees]C, is shown in Fig. 8. All polymers exhibited pseudoplastic behavior, i.e., a reduction of apparent viscosity with an increase of shear rate. This is due to two factors: (1) orientation of the polymer chains in the flow direction as the stress rate increases, the viscosity (resistance to flow) decreases; (2) disruption of the entanglement present in the polymeric solutions as the shear rate increases (40). Moreover, the copolymers had higher viscosities than those of the unmodified CMC. This behavior indicates that intermolecular hydrophobic interactions can occur even in light saline solutions, but it is important to remember that this effect can also be caused by the interactions between divalent cations and carboxylate groups. However, the sum of the two effects was insufficient to increase the size of the aggregates or lead to a significant viscosity increase. Furthermore, the increase in copolymer viscosities in solution confirms the efficacy of incorporating the hydrophobic groups onto the CMC backbone. However, no significant difference in the viscosities of the aqueous copolymer solutions (CMCHM1 and CMCHM3) was observed.

[FIGURE 8 OMITTED]

CONCLUSIONS

The synthesis of CMC-g-poly(N,N-dihexylacrylamide) copolymers with 1 to 6 mol % of hydrophobic groups was not confirmed by [H.sub.1] NMR, likely because this technique is not sensitive enough to detect the presence of very low hydrophobe content in a medium. However, the results obtained by viscosity, static and dynamic light scattering and rheological measurements for all copolymers showed significant changes in their solution properties, indicating that a chemical modification of the cellulose derivative actually occurred.

In a dilute regime, intramolecular hydrophobic associations promoted by hydrophobic groups became evident with a slight reduction of the intrinsic viscosity of the copolymers generated by the increase of ionic strength in the medium. This effect was more pronounced with the increased hydrophobic group length grafted onto the CMC backbone. Copolymer formation was also confirmed by SLS experiments, by increasing the [M.sub.w] and decreasing the [R.sub.g] and [A.sub.2] when compared to unmodified CMC, indicating the presence of hydrophobic structures on the CMC backbone that favor the polymer-polymer interactions caused by reduced solvent quality.

The results obtained by DLS experiments illustrates that the [R.sub.H] of the copolymer aggregates increases with increased polymer concentrations in low ionic strength systems (Milli-Q water and SFW), which suggests the formation of intermolecular hydrophobic associations. However, in 0.5 M NaCl (the highest ionic strength solvent) [R.sub.H] decreased as polymer concentration increased, indicating the formation of intramolecular hydrophobic interactions. The copolymer with the greatest hydrophobic group length had the lowest [R.sub.H] values in three solvents and in all polymer concentrations used, showing that the intramolecular hydrophobic associations became more significant as hydrophobic group length increased. The results held true at higher temperatures. Although, the intermolecular hydrophobic interactions in the copolymers were observed in the presence of salt of high ionic strength, high temperatures and high polymer concentrations, the [R.sub.H] values of the aggregates were still lower than those observed in Milli-Q water and SFW. The intermolecular hydrophobic interactions were also shown by the rheological behavior of copolymers in SFW at 55[degrees]C in a semidilute, since the two copolymers studied (CMCHM1 and CMCHM3) had higher viscosities than that of the unmodified CMC. However, the aggregates formed by the intermolecular hydrophobic interactions promoted by divalent cations and carboxilate groups were not enough to significantly increase the size of the aggregates.

ACKNOWLEDGMENTS

The authors are grateful to the PRH30-ANP/MCT and Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (CNPq) for the financial support toward the attainment of a Ph.D undertaken partly in Brazil and partly in France. The Experiments in this study were carried out in Brazil and in LCPO, Bordeaux, France.

REFERENCES

(1.) B. Nystrom, A.-L. Kjoniksen, and C. Iversen, Adv. Colloid Interface Sci., 79, 81 (1999).

(2.) W. Xue, I.W. Hamley, V. Castelletto, and P.D. Olmsted, Eur. Polym. J., 40, 47 (2004).

(3.) J. Zhang, L-M. Zhang, and Z-M. Li, J. Appl. Polym. Sci., 78, 537 (2000).

(4.) M. Camail, A. Margaillan, I. Martin, A.L. Papailhou, and J.L. Vernet, Eur. Polym. J., 36, 1853 (2000).

(5.) F. Candau and J. Selb, Adv. Colloid Interface Sci., 79, 149 (1999).

(6.) Y.A. Shashkina, Y.D. Zaroslov, V.A. Smirnov, O.E. Philippova, A.R. Khokhlov, T.A. Pryakhina, and N.A. Churochkina, Polymer, 44, 2289 (2003).

(7.) Y. Feng, L. Billon, B. Grassl, A. Khoukh, and J. Francois, Polymer, 43, 2055 (2002).

(8.) J. Ma, P. Cui, L. Zhao, and R. Huang, Eur. Polym. J., 38, 1627 (2002).

(9.) D. Charpentier, G. Mocanu, A. Carpov, S. Chappelle, L. Merle, and G. Muller, Carbohydr. Polym., 33, 177 (1997).

(10.) S. Simon, J.Y. Dugast, D. Le Cerf, L. Picton, and G. Muller, Polymer, 44, 7917 (2003).

(11.) I. Bataille, J. Huguet, G. Muller, G. Mocanu, and A. Carpov, Int. J. Biol. Macromol., 20, 179 (1997).

(12.) E. Rotureau, E. Dellacherie, and A. Durand, Eur. Polym. J., 42, 1086 (2006).

(13.) Y. Yamanaka and K. Esumi, Colloids Surf. A Physicochem. Eng. Aspects, 122, 121 (1997).

(14.) S. Nilsson, K. Thuresson, P. Hansson, and B. Lindman, J. Phys. Chem. B, 102, 7099 (1998).

(15.) T. Aubry, F. Bossard, and M. Moan, Polymer, 43, 3375 (2002).

(16.) G.A. Stahl and D.N. Schulz, Water-Soluble Polymer for Petroleum Recovery, Plenum Press, New York, 147 (1988).

(17.) K. Podhajecka, K. Prochazka, and D. Hourdet, Polymer, 48, 1586 (2007).

(18.) G. Bokias, Y. Mylonas, G. Staikos, G.G. Bumbu, and C. Vasile, Macromolecules, 34, 4958 (2001).

(19.) E. Volpert, J. Selb, and F. Candau, Macromolecules, 29, 1452 (1996).

(20.) E. Volpert, J. Selb, and F. Candau, Polymer, 39, 1025 (1998).

(21.) D. Charpentier-Valenza, L. Merle, G. Mocanu, L. Picton, and G. Muler, Carb. Polym., 60, 87 (2005).

(22.) A. Sinquin, P. Hubert, and E. Dellacherie, Langmuir, 9, 3334 (1993).

(23.) A. Sinquin, P. Hubert, P. Marchal, L. Choplin, and E. Dellacherie, Colloids Surf. A Physicochem. Eng. Aspects, 112, 193 (1996).

(24.) F.F.-L. Ho and D.W. Kloslewlcz, Anal. Chem., 52,913 (1980).

(25.) C.L. McCormick, T. Nonaka, and C.B. Johnson, Polymer, 29, 731 (1988).

(26.) A. Hill, F. Candau, and J. Selb, Macromolecules, 26, 4521 (1993).

(27.) C. Follmer, F.V. Pereira, N.P. da Silveira, and C.R. Carlini, Biophys. Chem., 111, 79 (2004).

(28.) C.W. Hoogendam, A. de Keizer, M.A.C. Stuart, B.H. Bijsterbosch, J.A.M. Smit, J.A.P.P. van Dijk, P.M. van der Horst, and J.G. Batelaan, Macromolecules, 31, 6297 (1998).

(29.) A.J.F, Siegert, Corretation Functions in Dynamic Light Scattering, MIT Rad. Lab. Rep. no. 465 (1943).

(30.) S.W. Provencher, J. Hendrix, L. de Maeyer, and N.L.J. Paulussen, J. Chem. Phys., 69, 4273 (1978).

(31.) A-L. Kjoniksen, B. Nystrom, and B. Lindman, Langmuir, 14, 5039 (1998).

(32.) A-L. Kjoniksen, S. Nilsson, S. Nilsson, K. Thuresson, B. Lindam, and B. Nystrom, Macromolecules, 33, 877 (2000).

(33.) A.M.S Maia, M. Costa, R. Borsali, and R.B. Garcia, Macromolecular Symposia, 229, 217 (2005).

(34.) M. Rinaudo, M. Milas, N. Jouon, and R. Borsali, Polymer, 34, 3710 (1993).

(35.) A. Tager; Physical Chemistry of Polymers, Cap. 15, Mir. Publishers, 2nd ed., Moscow (1978).

(36.) E.F. Lucas, B.G. Soares, and E. Monteiro, Serie Instituto de Macromoleculas: Caracterizacao de Polimeros - Determinacao de peso molecular e analise termica, Cap. 5, e-papers. Rio de Janeiro (2001).

(37.) W. Fred and JR. Billemeyer, Measurement of Molecular Weight and Size (Light Scattering), Textbook of Polymer Science, 3rd ed., A Wiley-Interscience Publication, New York: John Wiley & Sons (1984).

(38.) Z-G. Wang, Langmuir, 6, 928 (1990).

(39.) S Guillot, M. Delsanti, S. Desert, and D. Langevin, Langmuir, 19, 230 (2003).

(40.) O.H. Lin, R.N. Kumar, H.D. Rozman, M. Azemi, and M. Noor, Carbohydr. Polym., 59, 57 (2005).

Correspondence to: Rosangela Balaban; e-mail: [e-mail] DOI 10.1002/pen.21180

Published online in Wiley InterScience (www.interscience.wiley.com).

[c] 2008 Society of Plastics Engineers

Rosangela Regia Lima Vidal, (1) Rosangela Balaban, (1) Redouane Borsali (2), (3)

(1) LAPET, Chemistry Department, Federal University of Rio Grande do Norte, 59078-970, Natal/ RN- Brazil

(2) LCPO-CNRS- ENSCPB & Universite Bordeaux 1, 16 Avenue Pey Berland 33607 PESSAC Cedex, France

(3) CERMAV, CNRS and Joseph Fourier University, BP53, 38041 Grenoble cedex 9, France