INTERFACIAL CHEMISTRY ASPECTS OF DE-INKING FLOTATION OF MIXED OFFICE PAPER

Preprint 01-9 INTERFACIAL CHEMISTRY ASPECTS OF DE-INKING FLOTATION OF MIXED OFFICE PAPER J. Drelich, J. Pletka, P. Boyd, E. Raber, D. Herron, E. Luhta, H. Walqui, N. Tervo, S. Boston, J. Wieland, J. Morgan, and N. Sabo Michigan Tech Univ. Houghton, MI, USA ABSTRACT Mini-projects studying the flotation chemistry of paper de-inking were incorporated to an educational program at Michigan Technological University with support from the paper recycling industry. Interfacial activity of frothing agents, surfactants and polymers in de-inking flotation systems has been studied through flotation, atomic force microscopy, zeta potential, contact angle, surface tension, and froth stability measurements. These fundamental studies were developed to understand microscopic and molecular effects occurring in flotation of ink particles that may lead to improved designs in deinking flotation processes. Fundamental studies indicate that ethoxylated alcohols are efficient frothing agents in flotation and can be used to clean the recovered paper from hydrophobic ink particles. Next, it was further supported in this research that the fatty acids activated by calcium ions act as collectors in flotation of ink particles. Finally, improvements in de-inking flotation are reported in laboratory tests involving aqueous solutions of polymeric flocculants. INTRODUCTION Paper recycling is becoming active part of the paper industry because of growing environmental concerns and economic reasons. The paper recycling process is a way to wash, remove, and bleach unwanted particles from collected printed paper so that it can be sold as an alternative to virgin paper. In modern paper recycling plants, the process of removing unwanted particles from the pulp can take as much as three times the amount of steps that it does in the mineral processing industry due to the fact that a higher quality product is needed to compete with the virgin paper. Many of these steps change parameters such as ph and temperature. Also, solution chemistry is different at each stage of the paper recycling technology. Therefore, to stay on the cutting edge of technology, testing is always being done to try to improve the paper recycling technology at different stages of screening, washing, bleaching, and flotation operations. Flotation is one of the primary methods used to deink printed-paper pulp. It works by collecting dispersed ink particles on air bubbles and trapping them in a froth layer. The spectrum of chemicals added to the flotation circuits include ink particle collectors (fatty acid or alcohol ethoxycarboxylates), collector activator (calcium ions), frothing agents (nonionic surfactants), and ph regulators (NaOH or Na 2 CO 3 ) (Borchardt 1994; 1997; Somasundaran et al. 1999). In the mineral-processing field today, it is not uncommon to test new chemicals to try to improve the flotation process. The same holds true for the paper recycling process, which uses similar flotation techniques. There are however, significant differences between materials to be processed in mineral processing technology and de-inking of paper (Table 1) that pose some differences in flotation approaches. As listed in Table 1, ink particles are of substantially lower concentration in the paper pulp than valuable minerals are in mineral slurry. Furthermore, ink particles are often flaky in their shape and have smoother surfaces as compared to mineral particles. Also, ink is a substantially lighter material than most minerals. The above-mentioned characteristics of paper flotation systems, plus others listed in Table 1, are primarily responsible for relatively slow kinetics and poor efficiency in de-inking flotation. As the result, a challenge for technological innovation in the area of paper de-inking using flotation separation is still open. Many current research activities around the world concentrate on more efficient flotation 1 Copyright 2001 by SME

machines and cells, and improvements in the design of solution chemistry that promotes agglomeration of ink particles and/or reinforces changes in the shape of ink particles (Doshi and Dyer 1997; 1998). To increase the efficiency of flotation separation, researchers are looking for new reagents and technological approaches. The innovation in this area can be reached, to a certain degree, through understanding the microscopic mechanisms of flotation and molecular interfacial effects in this separation process. The effects of reagents on deinking flotation were examined through flotation experiments and fundamental surface chemistry measurements. In this paper, we briefly review the selected results of the mini-projects that were carried out by the undergraduate students in collaboration with Great Lake Pulp and Fiber Mill located in Menominee, MI. Table 1.General Characteristics of Pulp Components of Flotation Systems Used In Processing of Minerals and De-inking of Paper Parameter Flotation of Minerals De-inking Flotation Components of pulp mineral and gangue particles cellulose fibers, ink, and mineral filler particles Skimmed fraction Concentration of pulp Concentration of solids to be floated Hydrophobicity of particles Typical size of particles Shape of particles Density of particles usually product (valuable mineral) 10-30 wt% solids 1wt% usually hydrophilic; collectors are used to improve hydrophobicity of valuable mineral 100-300 µm relatively uniform shape with aspect ratio of ~ 1:1; usually no substantial difference between the shape of valuable mineral and gangue material; particles have rough surface with sharp irregularities 1.6->4.0 g/cm 3 always reject (ink and mineral filler particles) 1.0-1.2 wt% solids << 0.1 wt% most of the ink particles are hydrophobic ink particles: 20-200 µm; cellulose fibers: a few millimeters often flat-like ink particles with the aspect ratio of >2:1; substantial difference in size and shape of ink particles and cellulose fibers; ink particles have smooth surface with rounded irregularities 1.3-1.4 g/cm 3 HYDROPHOBICITY OF INK AND ITS SIGNIFICANCE Flotation reagents used in paper recycling are generally termed surfactants or surfactant blends, but usually can be broken down into two main components: frother and collector. Because most of the inks, particularly oil-based (polymer-based) inks present in mixed office paper, are hydrophobic, the use of a frother without any collector is the first guess for the mineral process engineer. For example, Table 2 shows the results of advancing contact angle values measured on the toner substrate for either pure water or surfactant blend solutions (additional examples of contact angle measurement results are reported in refs (Drelich et al. 1996; 2000)). High hydrophobicity of ink substrate expressed by the advancing contact angle of 79-88 degrees (Table 2) is a result of the polymeric composition of the ink, made of polystyrene/ polyacrylate copolymer (note that roughness of the ink surface has also a small effect on contact angles reported in Table 2 as the advancing contact angle measured for water on a smooth surface of the polystyrene/polyacrylate toner is about 80 degrees (Azevedo et al. 1999a)). The high hydrophobicity of ink is comparable to, or even larger, than hydrophobicity of mineral specimens covered with hydrophobic collector layers. However, it is now a well recognized fact in the flotation of minerals that the high hydrophobicity of floated particles is not always sufficient condition for 2 Copyright 2001 by SME

the effective particle-bubble attachment to occur. Repulsive electrical double layer forces can make an energetic barrier between the particle and gas bubble and as the result, significantly reduce a probability of the particle-bubble attachment process (the kinetics of flotation to be reduced). This aspect of particle flotation process was analyzed in our research activities. Figure 1 illustrates the interfacial forces measured between polyethylene particle (polyethylene served as a substitute of air bubble in the atomic force microscopy studies) and ink substrate. Table 2.Advancing Contact Angles Measured On Melted Xerox 5052/1050 Ink Substrate (Pletka et al. 2000) Solution Surface Tension [mn/m] Water 72.0 0.042 g/l Lionsurf 792L blend *) (ph 7) 53.4 0.042 g/l Lionsurf 792L blend (ph 9) 55.7 0.042 g/l Lionsurf 792L blend plus 50:M CaCl 2 (ph 7) 65.3 0.042 g/l Lionsurf 792L blend plus 50:M CaCl 2 (ph 9) 63.9 *) Lionsurf blend is a mixture of polyalkylene oxide surfactant and fatty acid. Contact Angle [deg] 87 79 83 88 87 Figure 1. Interfacial Forces Versus Distance Measured Between 19 µm Polyethylene Particle and Xerox 5052/1050 Ink Substrate In Water (ph 7) Using Atomic Force Microscope (Pletka et al. 2000) As shown in Figure 1, although strong attractive forces were measured in deionized water at distances smaller than about 10 nm, the repulsive forces were noted at larger distances. The attractive forces are attributed to hydrophobic effects as both polyethylene and ink are hydrophobic materials; advancing contact angle of about 100 and 80 degrees, respectively. The van der Waals forces also contribute to attraction between ink and polyethylene but at distances much below 10 nm and they are not of primary concern in de-inking flotation systems (Drelich et al. 2000). The repulsive interaction between ink and polyethylene in deionized water (Figure 1) is a result of negative surface potentials at both materials. Such strong repulsive forces were not observed between polyethylene and ink in tap water (Figure 1) due to a compression of the electrical double layer by the inorganic ions dissolved in tap water. Therefore, hydrophobicity-origin attractive forces seem to dominate over the repulsive electrical double layer forces in tap water. The results of interfacial force measurements such as shown in Figure 1 indicate that the repulsive forces between gas bubbles and hydrophobic ink particles are probably dominated by attractive forces in daily mill operations, where process water with increased amount of dissolved ions (at even higher level than in tap water used in the fundamental studies) is always used (Pletka et al. 2000). Thus, due to high hydrophobicity of many inks and insignificant repulsive electrical double layer forces between air bubble and ink particles, the de-inking flotation of hydrophobic ink particles can be successful with the use of frother alone. The role of frother is to build a stable froth layer that allows the separated hydrophobic ink particles to be skimmed from the top of the flotation cell. This however, not necessary leads to elimination of collectors as discussed in the further part of the paper. ETHOXYLATED ALCOHOLS AS FROTHERS Ethoxylated alcohols (C n H 2n+1 E m ) and their derivatives were introduced to the paper recycling plant quite recently as dispersants and frothing agents (Borchardt 1992; 1993). The interfacial activity of polyoxyethylene lauryl ethers (C 12 H 25 E m ; m=4,8,10,23) was also studied in de-inking flotation systems in one of the MTU mini-projects. For example, it was found that a gain of 4-8 points in the paper pulp product brightness can be obtained for the mixed office waste with the use this frother (polyoxyethylene lauryl ethers) alone (Figure 2). The results in Figure 2 clearly support the abovediscussed approach that the hydrophobic ink particles do not require collectors to be effectively floated from the paper pulp. It was reported in our laboratory however, that the paper pulp product from the flotation with a commercial (Lionsurf) frothercollector surfactant blend (Lionsurf is currently used by many paper recycling mill) was usually brighter by 1-3 points from that separated with the 3 Copyright 2001 by SME

polyoxyethylene lauryl ether frother (not shown). Figure 3 shows selected results of flotation with commercial surfactant blend at different temperatures (based on Pletka et al. 2000). Direct comparison of results in Figure 2 and 3 cannot be made due to different paper pulp feed used in both experiments. Figure 2. Effect of ethoxy group number in the chemical structure of polyoxyethylene lauryl ether (C 12H 25E m; m=4,8,10,23), used as a frother in deinking of mixed office waste, on brightness of paper pulp product. The tests were carried out in a 2L Denver flotation cell (1400 rpm; 1wt% solids; 3 min) using paper pulp feed (57-58% brightness) received from the Great Lake Pulp and Fiber. The concentration of frother was 12 µm. The flotation tests were done at a room temperature. hydrophobic surfaces of ink and polyethylene and its effect on wettability of these surfaces is discussed in more details in another contribution (Drelich et al. 2001). Although the reduction of hydrophobicity is undesirable in de-inking flotation, a few degrees decrease in contact angle, such as shown in Figure 4, should have a negligible effect on de-inking flotation. Stability of the froth layer might be a more important factor to be taken into consideration during the selection of frother and its dosage. The effect of froth stability on the removal of hydrophobic ink particles from paper pulp was analyzed in this research. It was found that de-inking flotation relies, at least to a certain extend, on the stability of froth generated in the flotation cell. For example, Figure 5 shows the correlation plotted between froth stability data and data on the brightness of the paper pulp recovered in de-inking flotation tests. The same 12 µm polyoxyethylene lauryl ether (C 12 H 25 E m ; m=4,8,10,23) solutions were used in de-inking flotation and froth stability tests. Note that both flotation and froth stability tests were done independently. It is clear that in spite of reduced hydrophobicity of ink surface (Figure 4), the increase in efficiency of the de-inking flotation was recorded (Figure 2). This effect is attributed to stability of froth layer in the flotation cell as illustrated by the froth height vs. brightness correlation in Figure 5. Figure 3. Effect of temperature on brightness of paper pulp product in flotation tests. The tests were carried out in a 2L Denver flotation cell (1400 rpm; 1wt% solids; 5 min) using paper pulp feed (brightness 54%) received from the Great Lake Pulp and Fiber. The concentration of Lionsurf 792L surfactant blend (polyalkylene oxide plus fatty acid) was 0.028 g/l. Furthermore, Figure 4 shows the results of contact angle measurements for 12 :M C 12 H 25 E m solutions on a Xerox ink substrate. As experimental data demonstrates, hydrophobicity of the ink surface (expressed in relaxed contact angles) was reduced in contact with polyoxyethylene lauryl ether solutions. This reduction increased with increasing number of ethoxy group (m) in the chemical structure of polyoxyethylene lauryl ether, from m=4 to m=10. The adsorption of ethoxylated alcohols on Figure 4. Effect of ethoxy group number in the chemical structure of polyoxyethylene lauryl ether (C 12H 25E m; m=4,8,10,23) on sessile-drop contact angle measured on the surface of Xerox 5052/1050 ink substrate. Concentration of C 12H 25E m was 12 µm and a temperature was 20 o C. The contact angle was measured after sessile drop of solution was equilibrated on the ink surface for 8.5 min. Result for zero ethoxy group number represent data obtained with water. 4 Copyright 2001 by SME

between ink particles in formation of ink-ink aggregates, and probably between ink particle and gas bubble during flotation. Ink-ink enhance flotation because of the increased inertia of the formed aggregates. Additionally, a change in the shape of floating particles can be another factor affecting the de-inking flotation. However, these fundamental aspects of de-inking flotation are still under debate. Figure 5. Correlation between froth stability data and paper pulp product brightness results for the flotation experiments. The brightness results are those from Figure 2 (caption of figure provide details of flotation experiments). The froth stability is represented by the froth layer height generated in a glass column by purging air through a porous grid placed in a 12µM polyoxyethylene lauryl ether (C 12H 25E m; m=4,8,10,23) solution. Time on the graph represents the lifetime of the froth in the tests. THE ROLE OF FATTY ACID AND CALCIUM IONS IN DE-INKING FLOTATION Fatty acids have been successfully implemented in the paper recycling as a de-inking flotation collector. Fatty acids alone are ineffective in improving the attractive forces between bubble and hydrophobic ink particles. It is a well known fact in the de-inking flotation that calcium ions activate the action of fatty acids (Beneventi and Carre 2000). This was also proved in our fundamental research on contact angle measurements (Pletka et al. 2000) and atomic force microscopy (AFM) studies (Drelich et al. 2000). First, the attractive forces between a polyethylene particle (model of gas bubble) and ink significantly increase when a surfactant blend, containing fatty acids, is assisted by calcium ions (Pletka et al. 2000). The same tendency was observed in AFM measurements between polyethylene and ink in solutions of sodium oleate (Drelich et al. 2000). Second, contact angle results indicate that the hydrophobicity of Xerox ink increases in the presence of both fatty acids and calcium ions (Table 2). The role of fatty acid-calcium soaps in de-inking flotation was discussed by several authors in the past and this discussion was recently reviewed and summarized by Beneventi and Carre (2000). The details of molecular and submicroscopic mechanisms governing improvement of de-inking flotation in the presence of fatty acid-calcium complexes are not clear yet, although a promoting role of precipitated colloids of fatty acid-calcium soap in a coagulation process has been well documented (for review see Beneventi and Carre, 2000). For example, the contact angle results in Table 2 indicate that hydrophobic fatty acid-calcium precipitates adsorb to the hydrophobic surface of ink. The adsorbed soaps also serve as bridges SELECTIVE FLOCCULANTS AS NEW ADDITIVES IN DE-INKING FLOTATION If the particle size is too small, the particle s inertia is negligible and tends to follow the streamline around the air bubble. This leads to a smaller probability of collision with an air bubble and therefore the particle has less of a chance to be captured. Usually, ink particles present in the paper pulp are of very broad size distribution. For example, the size of ink particles ranges from about 10 :m to 600 :m (Azevedo et al. 1999b). Fine ink particles with a diameter smaller than 30 :m are particularly resistant to flotation (Borchardt 1997). Therefore, any coagulation or flocculation process initiated in the flotation cell might improve the efficiency of de-inking flotation. This has been observed for many years by using fatty acids and calcium ions (see section above). Flocculating polymers are not used by paper recycling mills in de-inking flotation yet. Polymers might initiate flocculation of both cellulose fibers and ink particles reducing the selectivity of flotation and causing a loss of cellulose fibers. Both components of the paper pulp, cellulose fibers and ink particles, have surfaces that are negatively charged in water. Thus, a selection of polyelectrolytes based on zeta potential measurements cannot be recommended here. However, both ink and cellulose differ significantly in hydrophobicity; cellulose fibers are hydrophilic whereas oil-based ink particles are hydrophobic. Therefore, it seems to be logical to use selective organic flocculants, which selectively adsorb to hydrophobic ink particle surfaces. The selective adsorption would promote flocculation of ink particles leaving hydrophilic cellulose fibers outside the ink-ink aggregates. In one study, de-inking flotation experiments using random selected flocculants were performed. The results from a series of experiments using a Percol 727 flocculant and using a common surfactant blend designed for paper pulp used in our experiments are shown in Figure 6. A brightness of 63.0+0.5 % was about the highest value that could be reached with the surfactant blend for this particular paper pulp sample. As shown in Figure 6, the use of Percol 727 at a concentration of 0.15 to 0.5 mg/l led to an improvement in de-inking flotation by 2-3 points in 5 Copyright 2001 by SME

the brightness of the resulting paper product. Also, analysis of the dirt count in the paper products from this series of experiments indicated that the dirt count numbers were reduced from 350 ppm to 180 ppm for all dirt sizes, and from 240 ppm to 70 ppm for dirt count sizes smaller than 0.04 mm 2. Similar results were noted in tests with other flocculants; Percol 7651 and Superfloc 210+. Figure 6. Effect of Percol 727 flocculant on the efficiency of deinking flotation. The tests were carried out in a 2L Denver flotation cell (1400 rpm; 1wt% solids; 3 min) using paper pulp feed (brightness 56-58%) received from the Great Lake Pulp and Fiber. The Lionsurf 792L surfactant blend (polyalkylene oxide plus fatty acid)(squares) or C 4E 10 (diamonds) was added at a concentration of about 0.03 g/l and 12µM, respectively, to generate a froth. The results of the flotation tests with flocculants shown in Figure 6 are encouraging from both practical and fundamental points of view. The role of flocculant in the de-inking flotation system is still unclear, particularly regarding the interaction of flocculant with all reagents and particulates present in the paper pulp. For example, it was noted that the addition of the Percol 727 flocculant has effect on the surface hydrophobicity of ink. The contact angle data in Table 3 illustrate this effect. As shown in Table 3, an addition of the Lionsurf surfactant blend to the solution reduces the contact angles measured on the ink surface. This is in accordance with the results in Figure 4, on the effect of ethoxylated alcohols on the hydrophobicity of polyethylene and ink (Drelich et al. 2001). Next, the addition of Percol 727 flocculant to the solution partially restored the hydrophobic properties of ink surface; the contact angles significantly increased (Table 3). This cannot be attributed to the water surface tension as we have not noted such an effect of the Percol 727 on the surface tension of surfactant blend solution in pendant-drop surface tension measurements. It is thus very likely that the increase in contact angles observed for the solutions on the ink surface could be attributed to the structure of adsorbed molecular layers, composed of surfactant blend components and flocculant, on the ink surface. The fundamental research on the molecular interactions and arrangements of alcohols, fatty acids and flocculants on the hydrophobic ink surface is needed to understand the formation of molecular structures that might have impact on kinetics and efficiency of de-inking flotation. Table 3. Sessile-drop contact angles measured for aqueous solutions on Xerox 5025/1020 ink surface Solution Water 0.03 g/l Lionsurf 792L 0.03 g/l Lionsurf 792L + 4 mg/l Percol 727 0.03 g/l Lionsurf 792L + 7 mg/l Percol 727 Contact Angle after 5 min [deg] 88 65 70 72 ACKNOWLEDGMENTS The authors would like to thank Great Lakes Pulp and Fiber (GLPF) for providing paper pulp samples and providing insights into the paper recycling technology. Special thanks are addressed to Ms. Laurie Groleau of GLPF for running the brightness tests. JD acknowledges the National Science Foundation for the partial financial support of the paper recycling project through a subcontract with the University of Utah. JD also appreciates collaboration on the de-inking flotation program with Prof. J.D. Miller and Dr. J. Nalaskowski from the University of Utah and Dr. J.K. Borchardt of Tomah Products. REFERENCES Azevedo, M.A.D., Veeramasuneni, S., Miller, J.D., 1999a, Toner Surface Roughness and Wettability, Progress in Paper Recycling, Vol. 8, No. 2, pp. 20-33. Azevedo, M.A.D., Drelich, J., Miller, J.D., 1999b, The Effect of ph on Pulping and Flotation of Mixed Office Wastepaper, Journal of Pulp and Paper Science, Vol. 25, No. 9, pp. 317-320. Beneventi, D., Carre, B., 2000, The Mechanisms of Flotation Deinking and the Role of Fillers, Progress in Paper Recycling, Vol. 9, No. 2, pp. 77-85. 6 Copyright 2001 by SME

Borchardt, J.K., 1992, Chemical Structure Property Relationships of Deinking Surfactants, Progress in Paper Recycling, Vol. 1, No. 2, pp. 45-60. Borchardt, J.K., 1994, Possible Deinking Mechanisms and Potential Analogies to Laundering, Progress in Paper Recycling, Vol. 2, No. 2, pp. 47-53. Borchardt, J.K., 1994, Mechanistic Insights into De-inking, Colloids and Surfaces Vol. 88, pp. 13-25. Borchardt, J.K., 1997, The Use of Surfactants in De-inking Paper for Paper Recycling, Current Opinion in Colloid and Interface Science, Vol. 2, pp. 402-8. Doshi, M.R., Dyer, J.M. (Eds), 1997, Paper Recycling Challenge II: Deinking and Bleaching, Doshi and Associates Inc., Appleton, WI. Doshi, M.R., Dyer, J.M. (Eds), 1998, Paper Recycling Challenge III: Process Technology, Doshi and Associates Inc., Appleton, WI. Drelich, J., Azevedo, M.A.D., Miller, J.D., Dryden, P., 1996, Hydrophobicity and Elemental Composition of Laser-Printed Toner Films, Progress in paper Recycling, Vol. 5, No. 4, pp. 31-38. Drelich, J., Nalaskowski, J., Gosiewska, A., Beach, E., Miller, J.D., 2000, Long-Range Attractive Forces and Energy Barriers in De-inking Flotation: AFM Studies of Interactions between Polyethylene and Toner, Journal of Adhesion Science and Technology, in press. Drelich, J., Zahn, R., Miller, J.D., Borchardt, J.K., 2001, Contact Angle Relaxation for Ethoxylated Alcohol Solutions on Hydrophobic Surfaces, in Contact Angles, Wettability and Adhesion, K.L. Mittal (Ed.), VSP, in press. Pletka, J., Gosiewska, A., Chee, K.Y., McGuire, J.P., Drelich, J., Groleau, L., 2000, Interfacial Effects of a Polyalkylene Oxide/Fatty Acid Surfactant Blend in Flotation Deinking of Mixed Office Papers, Progress in Paper Recycling, Vol. 9, No. 2, pp. 40-48. Somasundaran, P., Zhang, L., Krishnakumar, S., Slepetys, R., 1999, Flotation Deinking A Review of the Principles and Techniques, Progress in Paper Recycling, Vol. 8, No. 3, pp. 22-36. 7 Copyright 2001 by SME