5756
Langmuir 1996, 12, 5756-5762
Articles Flocculation of Polystyrene Latex with Mixtures of Poly(p-vinylphenol) and Poly(ethylene oxide) Robert Pelton,* Huining Xiao, Michael A. Brook, and Archie Hamielec McMaster Centre for Pulp and Paper Research, Department of Chemical Engineering, McMaster University, Hamilton, Ontario L8S 4L7, Canada Received July 28, 1995. In Final Form: March 12, 1996X Poly(p-vinylphenol) (PVPh) was evaluated in conjunction with high molecular weight poly(ethylene oxide) (PEO) as a flocculant for monodisperse surfactant-free polystyrene latex. PVPh was effective with performance similar to that of a commercial phenol formaldehyde resin (PFR). Flocculation was not sensitive to PVPh molecular weight in the range 1500-30 000. PVPh was most effective at intermediate pH values where the polyphenol was present as dispersed particles in the colloid size range. It is proposed that the flocculation involved hydrogen bond complexes between the PVPh particles and PEO. Molecular mechanics calculations of relative energies of PEO/PVPh complexes with varying hydrogen bonding sequences indicated that hydrogen bonding with adjacent or alternating polyether oxygens along the PEO chain was less favorable than that for complexes with bonding of every fourth or fifth oxygen. Polyacrylic acid, another polymer capable of hydrogen-bonded complexes with PEO, gave no latex flocculation. Kinetic analysis indicated that flocculation was an orthokinetic collision process.
Introduction High molecular weight poly(ethylene oxide) (PEO) has been widely used as a flocculant in the manufacture of newsprint and related paper grades.1 In many cases it is necessary to use a second polymer, we call a cofactor, in conjunction with PEO. Water soluble phenol formaldehyde resin (PFR) is a common cofactor; however, kraft lignin2 and tannic acid3 are also effective cofactors. The common feature of all effective cofactors is the presence of phenolic hydroxyl groups which are presumed to form hydrogen-bonded complexes with PEO. The interactions between PFR and PEO have not been completely characterized because PFR, which is prepared by condensation of phenol and formaldehyde using acid or base catalysts, can have complicated branched structures with a broad molecular weight distribution. Poorly defined cofactor structures limit our ability to probe the PEO/cofactor complexes which appear to be crucial for flocculation. Described in this work is the use of poly(p-vinylphenol) (PVPh)4-6 as a cofactor. Although PVPh has been used in solid polymer blends,7 this is the first reported study of complex formation between poly(vinyl phenol) and PEO in water. After this work was completed, a European patent application by Hercules was published describing the use of PVPh as a cofactor.8 The impetus for this work was our interest in understanding the flocculation mechanism for the PEO/cofactor * To whom correspondence should be addressed. X Abstract published in Advance ACS Abstracts, May 15, 1996. (1) Pelton, R. H.; Allen, L. H; Nugent, H. Tappi, 1981, 64 (11), 89. (2) Pelton, R. H.; Allen, L. H.; Nugent, H. U.S. Patent 4,313,790 (1982). (3) Lindstrom, T.; Glad-Nordmark, G. Colloids Surf. 1984, 8, 337. (4) Pomposo, J. A.; Cortazar, M.; Calahorra, E. Macromolecules 1994, 27, 252. (5) Moskala, E. J.; Howe, S. E.; Painter, P. C.; Coleman, M. M. Macromolecules 1984, 17, 1671. (6) Suzuki, T.; Pearce, E. M.; Kwei, T. K. Polym. Commun. 1992, 33, 198. (7) Zhang, X.; Takegoshi, K.; Hikichi, K. Macromolecules 1991, 24, 5756. (8) Echt, E. European Patent Application 621 369 A1 (1995).
S0743-7463(95)00628-7 CCC: $12.00
system as a retention aid for newsprint manufacture. For this work we studied mixtures of polystyrene latex and bleached kraft pulp fibers. The latex serves as a model for colloids in pulp; this has been used in our laboratories 9 and others.3,10 The objectives of this work were (1) to investigate the factors influencing the PVPh/PEO performance in flocculating latex particles, which included pH, PVPh concentration, PVPh molecular weight, salt concentration, and order of reagent addition, and (2) to compare the performance of PVPh with PFR in latex flocculation. Experimental Section Materials. Poly(ethylene oxide) (PEO) with a weight average molecular weight of 8 × 106 was obtained from Union Carbide (Polyox 309). Methoxypoly(ethylene glycols) with molecular weights of 2000 and 5000 were also supplied by Union Carbide. Phenol formaldehyde resin (PFR) CASCOPHEN C271 (40% in aqueous solution), a standard cofactor for dual-polymer retention systems, was supplied by Borden Chemical. The molecular weight of PFR was approximately 13 000.11 Concentrated PFR was initially diluted to 3.7 wt % aqueous solution, which was further diluted to 250 mg/L for the flocculation experiments. Polyacrylic acid with a molecular weight of 2 × 105, obtained from PolySciences, was also used as a cofactor. The bleached fibers were softwood kraft pulp supplied by Noranda Forest Co. Three poly(p-vinylphenol) (PVPh) samples (PolySciences) with different molecular weights were used: PVPh-1, molecular weight 1500-7000; PVPh-9, molecular weight 9000; and, PVPh-30, molecular weight 30 000. PVPh was used as 50/50 wt % methanol/water solutions. Potentiometric and conductometric titrations of PVPh were carried out with an ABU93 Triburette Radiometer (Copenhagen) and a CDM 83 conductivity meter (Copenhagen) controlled by Aliquot software (McMaster University). 0.1 N NaOH was added first, and the forward titration was stopped at pH 11.5; 0.1 N HCl was added for the back titration (9) Xiao, H.; Pelton, R.; Hamielec, A. J. Colloid Interface Sci. 1995, 175, 166. (10) Couture, L.; van de Ven, T. G. M. Colloids Surf. 1991, 54, 245. (11) Xiao, H. N, Pelton, R. H.; Hamielec, A. E. J. Interface Colloid Sci. 1995, 175, 166-172.
© 1996 American Chemical Society
Flocculation of Polystyrene Latex
Langmuir, Vol. 12, No. 24, 1996 5757
to give 70 data points for the titration curves. The hydroxyl content of PVPh was calculated from the reverse titration curve. The particle size distributions of colloidal PVPh suspensions were measured using a BI-DCP (Disc Centrifuge Photosedimentometer) Particle Sizer (Brookhaven Instrument Co.). PVPh was initially dissolved in methanol at a concentration of about 1.9-2.9 wt %, and then dispersed in water to give a colloidal dispersion with PVPh concentrations ranging from 200 to 700 mg/L. The BI-DCP was operated in the homogeneous mode, and the density of the PVPh particles was 1.19 ( 0.01 g/cm3 measured by using a specific gravity bottle. Volumetric mean diameters are reported. The mobility measurements for the precipitated PVPh particles in aqueous solution were made with a Coulter DELSA 440 using the software version 1.36. The suspension was prepared by dispersing 0.1 mL of PVPh methanol solution (1.9-3) wt % into 20 mL of 0.001 M NaCl aqueous solution, and pH was adjusted to 5 by HCl. An improved parabola method was used for data analysis.12 Polystyrene latex was synthesized in our laboratories. Details of the preparation of the latex have been reported elsewhere.9 The diameter of the latex particles was 685 nm (weight average) measured using BI-DCP. Some flocculation experiments were conducted using a PSL size of 606 nm (intensity average with 17% of standard deviation), measured by dynamic light scattering using a fixed angle NICOMP 370 sub-microparticle sizer (Pacific Scientific). Precipitate calcium carbonate, Albacar HO, was supplied by Specialty Minerals Inc. Polystyrene Latex Flocculation. Both batch and continuous flocculation experiments were conducted. In a typical batch flocculation experiment, 35 mL of Milli-Q treated distilled water, 5 mL of bleached fiber pulp of 1.0 wt %, and 1 mL of 0.25 wt % latex were added to a 50 mL test tube. The pH of the solution was adjusted to approximately 5 by adding 0.1 mL of 0.02 M HCl. To maintain the electrolyte at a concentration of 2.0 × 10-3 M, 0.1 mL of 1 M NaCl was added followed by 0.4 mL of 250 mg/L phenolic resin to give a concentration of 2.0 mg/L. The fibers were allowed to settle and the transmittance for the supernatant was measured with a HP 8452A UV spectrophotometer (HewlettPackard) at a wavelength of 500 nm. This measurement gave Tc, the control transmittance. After redispersing the fibers, 0.4 mL of 250-500 mg/L aqueous flocculant (i.e. copolymer or PEO) was added. The samples were vigorously shaken by hand for 4 s, and the suspended solids were allowed to settle for 1 h at room temperature. The supernatant was decanted and filtered through a 200 mesh screen to remove suspended fiber fragments. The extent of latex removal by flocculation and sedimentation was determined by the relative turbidity, τR, calculated from the following expression:
τR ) log(100/Tg)/log(100/Tc)
(1)
where Tc and Ts were the percent transmittance of the control and the supernatant after flocculation, respectively. Dilution experiments were used to confirm that, for stable latex, τR was proportional to latex concentration. Furthermore, the standard deviation of Tc was 2.3% of the mean value based on six duplicates, and the standard deviation of Ts for one copolymer sample was 4% of the mean based on five duplicates. In continuous flocculation experiments, latex concentrations were continuously monitored by a photometric dispersion analyzer (PDA-2000, Rank Brother Inc.). Latex (0.5 mL of 4.0 wt %) was added to a 400 mL beaker containing 250 mL of Milli-Q water and equipped with a mechanical stirrer. NaCl was added to maintain the salt concentration at 10-3 M. The colloidal particles were kept mixed at 300 rpm. A peristaltic pump was used to pump the suspension at a flow rate of 65 mL/min through the PDA. The dc output of the PDA was proportional to the transmittance, and it was assumed that the corresponding turbidity was proportional to the concentration of stable latex particles. Similar assumptions have been used in studies of clay particle deposition on pulp fibers.13 (12) Pelton, R.; Miller, P. McPhee, M.; Rajaram, S. Colloids Surf. 1993, 80, 181. (13) Al-Jabri, M.; Van Heiningen, A. R. P.; van de Ven, T. G. M. J. Pulp Paper Sci. 1994, 20 (10), 289.
Figure 1. Potentiometric and conductometric titration curves for PVPh-30. Initial total volume ) 150 mL, PVPh concentration ) 0.29 g/L.
Figure 2. Particle size distributions of colloidally dispersed PVPh in water. Measurements were made with a Brookhaven BI-DCP centrifugal particle sizer. The specific gravity of the PVPh particles was 1.19. Table 1. Effect of PVPh Dispersed Particle Size on the Flocculation of Latexa volumetric conc in conc in mean conc in relative PVPh methanol water particle size flocculation turbidity samples (g/L) (mg/L) (mm) ( std (mg/L) (tR) PVPh-9 PVPh-30
29.4 19.0
690 285
0.55 ( 0.21 0.76 ( 0.25
6 6
0.35 0.40
a PVPh was dissolved in methanol and then dispersed in water. PEO-309 concentration was 2 mg/L.
The same procedure was used for experiments with precipitated calcium carbonate (PCC). PVPh (2 mg/L) was added to a 1 g/L PCC suspension at pH 8.3. PEO 309 (2 mg/L) was added 50 s after the PVPh.
Results Conductometric and potentiometric titration curves for PVPh-30 are shown in Figure 1. The potentiometric curve shows that pH values greater than 10 were required to give significant dissociation of the phenolic groups. The total hydroxyl group content, from the conductometric curve, was 8.2 mequiv/g of PVPh compared with a theoretical value of 8.33 mequiv/g of PVPh. PVPh is not water soluble at pH values less than 9, and for most of the experiments in this work PVPh was present as colloidal dispersions. The properties of the dispersions are illustrated in the following experiments. A concentrated PVPh solution (29 g/L) in methanol was added to water to give a 0.69 g/L PVPh suspension which was split into two parts. The first was characterized with a centrifugal particle size apparatus whereas the second was used in latex flocculation experiments. The two procedures were conducted as quickly as possible, so it was reasonable to assume that the dispersion characterized in the disk centrifuge was the same as that in the flocculation experiment. The particle size distributions for two PVPh samples are shown in Figure 2, and the flocculation results are summarized in Table 1 together with the average diameters. Relative turbidity values
5758 Langmuir, Vol. 12, No. 24, 1996
Pelton et al.
Figure 4. Effect of fiber concentration on latex flocculation. Experimental conditions: [690 nm latex] ) 50 mg/L; [PEO] ) 2.0 mg/L; [PVPh-30] ) 6.0 mg/L; pH ) 5.0; [NaCl] ) 0.002 M.
Figure 3. Photomicrograph of large latex flocs attached to a cellulose fiber. The fiber diameter is approximately 20 µm.
are assumed to be proportional to the concentration of latex not removed by flocculation. The particle size distribution for the higher molecular weight phenolic polymer was broader and had a larger mean value than that for the lower molecular weight PVPh. The corresponding flocculation results indicate that the smaller PVPh colloidal particles (0.55 mm for PVPh-9) produced slightly better flocculation (tR ) 0.35) than the larger PVPh particles (0.76 mm from PVPh-30, tR ) 0.40). In separate experiments, electrophoretic mobility values for the dispersed PVPh particles were measured at pH 5 in 0.001 M NaCl The results were -1.44 × 10-8 m2/(V s) for PVPh-9 and -1.18 × 10-8 m2/(V s) for PVPh-30. These values indicate that the surface potential should be high enough to confer colloidal stability. Two latex flocculation procedures were used. In the first, colloidally stable aqueous mixtures of latex, cellulose fibers, and colloidal PVPh were aggregated with PEO in a hand-shaken test tube. The extent of latex removal was estimated from the supernatant turbidity after sedimentation of the fibers and flocs. In the second, mixtures of colloidal PVPh and latex were flocculated in a stirred beaker by PEO addition and the extent of flocculation was estimated by turbidity measurements with a photodispersion analyzer. Control experiments confirmed that PEO alone and PVPh alone did not give latex flocculation with or without fibers. Furthermore, the presence of colloidally dispersed PVPh is a requirement for good flocculation. When coarse PVPh powder, instead of an alcoholic solution, was added to water, a coarse suspension was produced which gave poor flocculation. For example, 40 mg/L of PVPh powder followed by 5 mg/L of PEO gave a tR value of 0.89, indicating little latex flocculation. There was no evidence of deposition of either latex or PVPh particles onto the cellulose fibers in the absence of PEO. Shown in Figure 3 is a photomicrograph of flocs formed after PEO addition. The flocculated latex appears as large black masses which are attached to a wood fiber. Single latex or PVPh particles are not discernible at this resolution; however, it is clear that the flocs contain many particles. Therefore the dominant process was heteroflocculation of the latex with PVPh particles induced by PEO. Flocs containing many particles subsequently
Figure 5. pH effect on flocculation of latex for PEO-309/PVPh combinations. [latex] ) 50 mg/L; diameter of latex ) 685 nm; bleached fiber ) 1.0 g/L. Polymer concentrations were as follows: 2, PEO/PVPh-1 ) 1.0 /2.3 mg/L in 0.002 N NaCl; b, PEO/PVPh-1 ) 2/10 mg/L 0.002 N NaCl; and, 9, PEO/PVPh-30 ) 2/6 mg/L dissolved in 0.01 M NaCl.
deposit onto the fibers. SEM analysis showed no evidence of the deposition of individual latex particles onto the fibers. Wood fibers were used to help mixing and to enhance sedimentation of flocs. Figure 4 shows the degree of latex flocculation as a function of the concentration of wood fibers. Although latex flocculation was observed in the absence of fibers, more latex was flocculated, as judged by turbidity, when fibers were added. Similar results have been reported for latex flocculation with PEO/PFR mixtures.14 Latex flocculation experiments were conducted as a function of solution pH, and the results are shown in Figure 5. For PVPh-1 (10 mg/L) and PVPh-30 (6 mg/L), the maximum latex flocculation occurred at pH 5. By contrast, when PVPh-1 was used at a low concentration (2.25 mg/ L) and the PEO concentration was also reduced by half, latex flocculation increased as the pH decreased. In all cases, flocculation was poor when the pH was raised to the point where the PVPh dissolved. Presumably under these conditions there were insufficient phenolic hydroxyls for complex formation, or perhaps the PVPh was too negatively charged to adsorb onto the latex/water interface. Figure 6 shows latex flocculation with the three PVPh samples together with a standard cofactor, phenolic resin C-271 (PFR). The PFR gave the highest latex removal. At low cofactor concentrations (2 mg/L) PVPh-9 was the most effective PVPh whereas PVPh-30 was the best cofactor at 6 mg/L. In this work, the dispersed PVPh cofactors were less sensitive to concentration than was the soluble phenolic resin. The role of PVPh molecular weight is not obvious. Since the PVPh was present as colloidal particles, flocculation dynamics depend upon the PVPh particle concentration (14) Xiao, H.; Pelton, R.; Hamielec, A. J. Pulp Paper Sci., 1996, 22 (12), to appear.
Flocculation of Polystyrene Latex
Langmuir, Vol. 12, No. 24, 1996 5759 Table 2. Effects of Addition Modes on the Latex Flocculation Induced by PEO-309/PVPh-30a first addition
second addition
PVPh dosage (mg/L)
relative turbidity
PVPh-30 PEO PVPh-30 PEO
PEO PVPh-30 PEO PVPh-30
6 6 10.8 10.8
0.64 0.63 0.48 0.51
a Experimental conditions: [PEO] ) 2 mg/L, pH ) 5.0, [NaCl] ) 0.002 M, [fiber] ) 1.0 g/L, and [latex, 606 nm] ) 50 mg/L.
Figure 6. Influence of cofactor concentration and molecular weight on latex flocculation. Cofactors were PVPh-1 (b); PVPh-9 (2); PVPh-30 (9); and PFR-C271 ([). Conditions: [PEO-309] ) 2 mg/L; pH ) 5.0; [NaCl] ) 0.002 M; [fiber] ) 1.0 g/L; [685 nm latex] ) 50 mg/L.
Table 3. Effect of Addition Order of PEO, PVPh, and Latex on the Flocculationa first addition
second addition
third addition
relative turbidity
PS latex PEO PS latex PEO
PEO PVPh-30 PEO PFR C271
PVPh-30 PS latex PFR C271 PS latex
0.33 0.60 0.24 0.89
a
Figure 7. Effect of conductivity on the latex stability. Conductivity was varied by HCl and NaCl addition. Turbidity was calculated as log(100/T), where T is the transmittance of dilute latex. The pH was controlled between 3.2 and 3.8, and the remainder of the conditions were the same as given in Figure 6.
and size distribution, which in turn depends upon the molecular weight in a complicated way. For example, different polymer samples may have different concentrations of charge groups, which, in turn, would influence the nucleation of charge-stabilized particles when alcoholic PVPh solution was added to water. Clarification of the PVPh particle formation mechanism requires a detailed investigation. Flocculation efficiency was measured as a function of ionic strength, and the results are presented in Figure 7. The ionic strength was varied by adding NaCl, and conductivity was used to characterize the solutions. Turbidity, log(100/T), is plotted as a measure of the concentration of stable latex. The data labeled “no polymer” were from a control experiment involving neither PEO nor cofactor. As expected, at log(conductivity/(µmS/ cm)) values greater than 2, the latex coagulated due to double-layer compression. Comparison of the control and PEO/PVPh data sets showed that the polymeric flocculants were not sensitive to electrolyte until sufficient NaCl was added to coagulate the latex in the absence of polymer. Effects of the addition order in PEO/PVPh flocculation systems were evaluated and the results are summarized in Tables 2 and 3. Two series of experiments were conducted. In the first series (Table 2), latex, water, salt, HCl, and fibers were mixed first. Then, either PEO or PVPh was added. The results indicated that flocculation was not sensitive to the relative order of addition of PEO and PVPh. For example, at 6.0 mg/L concentration of PVPh-30, the relative turbidity was 0.64 when PEO (2.0 mg/L) was added last whereas the relative turbidity was 0.63 when PVPh was added last. In the second series of experiments, the latex was added at different stages. The results in Table 3 indicate that a much poorer flocculation occurred when the PS latex
Experimental conditions were the same as in Table 2.
Figure 8. Latex flocculation induced by PEO/PVPh-1. Data were compared with the results obtained by Pelssers et al.10. Experimental conditions: PEO/PVPh system: [PEO-309] ) 2 mg/L; [PVPh-1] ) 10 mg/L; pH ) 5.0; [NaCl] ) 0.002 N; the bleached fiber 1.0 g/L; [latex] ) 50 mg/L (i.e. N0 ) 2.8 × 1011/L for the latex with diameter 685 nm). Pelssers system: [latex] ) 426 mg/L (i.e. N0 ) 2.3 × 1012/L for the latex with diameter 696 nm); PEO molecular weight ) 6 × 105; PEO addition amount ) 0.83 mg/m2 (i.e. 8.34 mg/L); [KNO3] ) 0.01 M.
was added last. For example, about 67% of the latex particles were removed by PEO/PVPh-30 when the latex was added before the flocculant. By contrast, only 40% of the latex particles were removed when latex was added last. With PFR as cofactor, the difference was more pronounced (τR ) 0.24 vs τR ) 0.89). The kinetics of the PEO/PVPh induced flocculation process were determined by measuring turbidity as a function of time and the results are summarized in Figure 8. Also shown are the published data of Pelssers et al. for polystyrene latex flocculation with PEO but without cofactor or wood fibers.15 N1 and N0 are the number of stable latex particles after and before flocculation, respectively; we assumed that N1/N0 = relative turbidity. The latex concentration in this work was 2.8 × 1011/L, which was nearly an order of magnitude less than that used by Pelssers et al. Flocculation with the PEO/PVPh system was completed in the first 40 s, which was much faster than Pelsser’s results for PEO without cofactor. This comparison illustrates that the PEO/cofactor system is much more efficient than PEO alone. A latex sample was circulated through a rank particle dispersion analyzer which continuously measures turbidity. Figure 9 shows the dimensionless concentration of stable latex (estimated from turbidity) as a function of time. PVPh-30 was added at 50 s with no effect on latex (15) Pelssers, E. G. M.; Cohen Stuart, M. A.; Fleer, G. J. Colloids Surf. 1989, 38, 15.
5760 Langmuir, Vol. 12, No. 24, 1996
Pelton et al.
is polyacrylic acid, which is known to form hydrogenbonded complexes with PEO16 and yet showed no significant latex flocculation. For example, when 10 mg/L of 105 molecular weight polyacrylic acid at pH 5 was evaluated with 2 mg/L of PEO-309, the relative turbidity was close to 1.0. Discussion
Figure 9. Comparison of latex (606 nm) and precipitated calcium carbonate (PCC) flocculation. Latex/PVPh/PEO: 13.6 mg/L of PVPh-30 was added at time 50 s followed by 1.2 mg/L of PEO-309 at 100 s. [NaCl] ) 0.001 M and pH ) 5. PCC/ PVPh/PEO: 1 g/L, 2 mg/L of PVPh-30 was added at 50 s, 2 mg/L of PEO-309 was added at 120 s. PCC/PEO: 1 g/L, 2 mg/L of PEO-309 was added at 60 s.
Figure 10. Influence of poly(ethylene glycol) on latex flocculation with PEO 309/cofactor mixtures. The poly(ethylene glycol) molecular weights were 5000 for the PFR results and 2000 for the PVPh experiments. Conditions: PEO ) 2.0 mg/L; [PFR] ) 2.0 mg/L; [PVPh-30] ) 6.0 mg/L; pH ) 5.0; [NaCl] ) 0.002 M; [fiber] ) 1.0 g/L; and [606 nm latex] ) 50 mg/L.
stability. However, subsequent addition of PEO-309 at 100 s resulted in rapid flocculation. Latex concentration was reduced by 50% in about 9 s. For comparison, the diffusion-limited half-life of the latex, assuming no barrier to coagulation, was estimated to be 600 s. Also shown in Figure 9 are the results of two flocculation experiments with precipitated calcium carbonate (PCC). With PEO-309 alone there was no flocculation whereas flocculation was observed when PVPh was present. The small overshoot in the PCC/PVPh/PEO curve perhaps indicates that some flocs initially formed were partially redispersed. The PCC results show that flocculation with PVPh/PEO is not limited to latex. Also, the PCC results emphasize that PEO is a much better flocculant in the presence of PVPh. A well-documented feature of the PEO/cofactor system is that the PEO molecular weight must be greater than 106 to get good results. However, it is possible that short PEO chains could interfere with flocculation by complexing with PVPh or by adsorbing on the latex. To test this possibility, flocculation experiments with PEO/PVPh-30 and PEO/PFR were conducted in the presence of low (2000 and 5000) molecular weight poly(ethylene glycol). The results, summarized in Figure 10, show that when the poly(ethylene glycol) concentration was higher than 10 mg/L, τR was increased from 0.4 to 0.6 for the PEO/PVPh30 system, which indicates some interference. Perhaps this demonstrates the adsorption of poly(ethylene glycol) onto the PVPh/water interface, inhibiting the adsorption of high molecular weight PEO. By contrast, the PFR showed no evidence of interference. This corroborates previous work in which fluorescence dye techniques revealed no direct evidence of complex formation between poly(ethylene glycol) and PFR.9 In the preliminary stages of this work, potential nonphenolic cofactors were evaluated. A striking example
PVPh is a linear, well-defined homopolymer which is soluble in water at pH values greater than 9. For most of the flocculation experiments, PVPh was present as colloidally dispersed particles. By contrast, phenolic resin PFR-C271 was water soluble. In spite of this difference, the behavior of PVPh as a flocculation cofactor was similar to that of PFR. Common features include the following: (1) Better flocculation was obtained at low pH. This suggested that hydrogen bonding was involved in the flocculation processes. (2) There existed an optimum weight ratio of about 1:3 between PEO and PVPh for flocculation. (3) Flocculation was not a strong function of the PVPh molecular weight. (4) Flocculation was not sensitive to the addition order of the cofactor and PEO whereas flocculation was poor when latex was added last. (5) Flocculation was complete a few seconds after the second polymer component was added. (6) The presence of wood fibers helped flocculation. Nevertheless, subtle differences were observed. These included the following: PVPh was partially deactivated by the presence of low molecular weight poly(ethylene glycol) whereas PFR was not; and PFR showed optimum performance over a narrower concentration range than did PVPh. The dynamics of the PEO/PVPh system is complicated. Two colloidal species, PVPh and latex, both interact with PEO to form precipitates. On the other hand, there is no evidence of flocculation in the absence of PEO. The role of the fibers is less clear. Flocculation as judged by supernatant turbidity was better with fibers. Replacing the fibers with 1 mm silica beads or with mechanical stirring also gave good flocculation in the PEO/PFR system.14 Therefore we speculate that the fibers help mixing and form a substrate for the removal of flocculated latex. Indeed, the rapid flocculation kinetics also suggests that hydrodynamic effects, although not well defined in our experiments, were important. Molecular Mechanics Calculations It has been postulated for many years that polyphenolic cofactors form hydrogen-bonded complexes with PEO. An important feature of H bonding, compared with electrostatic or hydrophobic interactions, is that specific oxygen and hydrogen atoms are involved. Furthermore, these atoms must be very close (i.e.
