Effect of molecular weight and charge density on the performance of polyacrylamide in low-grade oil sand ore processing.

Li, Haihong ; Long, Jun ; Xu, Zhenghe 等

INTRODUCTION

Polyacrylamide polymers have been widely used as process aids in the mineral, coal, and oil sand industries to enhance solid-solid or solid-liquid separation (Moddy, 1992; Cymerman et al., 1999; Chen et al., 2003; Pearse, 2005). Their performance strongly depends on their physicochemical properties such as integrated functional groups, charge density, and molecular weight (MW). Common polyacrylamide molecules have only one basic functional group of amide (-CON[H.sub.2]), and their ability to selectively adsorb on solid surfaces is greatly enhanced with the integration of other functional groups such as carboxyl (-CO[O.sup.-]), hydroxyl (-OH), ether (-C-O-C-), amine (-RN[H.sub.2]), ammonium (-[N.sup.+][R.sub.3]), etc. (Chen et al., 2003). For polymers with the same functional groups, such as the family of hydrolyzed polyacrylamide (HPAM), charge density and MW are critical in determining their functionality. It was found that HPAM flocculants with a low charge density performed the best for calcite flocculation in terms of both settling rate and water clarity (Seyrankaya et al., 2000), and the optimum charge density was about 31% (Nishkov and Marinov, 2003). In the flocculation of oil sand tailings, a HPAM polymer of 27% anionicity achieved the best settling of fine particles (Cymerman et al., 1999). Effect of MW was even more drastic. At a low MW(<1 million Daltons), HPAM polymers were reported to act as dispersants in phosphate beneficiation (Nagaraj et al., 1987), alumina particle dispersion (Baklouti et al., 2003), and coal dispersion (Pawlik, 2005). At a high MW (>1 million Daltons), they were flocculants and widely used in fine clay flocculation (Xiao et al., 1999; Ovenden and Xiao, 2002; Yoon and Deng, 2004), fine alumina flocculation (Fan et al., 2000; Glover et al., 2004), dewatering of fine coal tailings (Sabah et al., 2004), and dewatering of oil sand tailings (Cymerman et al., 1999; Long et al., 2006a).

In a recent study (Li et al., 2005), a HPAM polymer with a MW of 17.5 million Daltons and an anionicity of 22% (a commercial polymer flocculant with a trade name of Percol 727) was successfully used as a process aid to recover bitumen from a low-grade oil sand ore. The bitumen extraction was carried out using a water-based process. Presently, such water-based processes are widely used in industry to recover bitumen from oil sands ores. In these processes, water is added to oil sands to form a slurry from which bitumen is liberated from the sand grains and recovered by flotation. In general, bitumen recovery decreases with increasing fines content in the oil sand ores. Here fines are defined by the oil sands industry as the mineral solids smaller than 44 microns. Test results have shown a close correlation between ore processability and content of fines (Liu et al., 2003). Slime coating, defined as a layer of fine particles coating on bitumen droplets, has been recognized to have a profound impact on bitumen aeration. It is speculated that these fines, depending on their wettability, on bitumen set up a steric barrier retarding bitumen droplets to contact with air bubbles, consequently resulting in a lower bitumen recovery. In addition, with an increasing content of fines in the ores, more fine particles could be trapped and brought by the air bubbles and/or bitumen droplets to the bitumen froth, leading to a deteriorated froth quality. In order to increase bitumen recovery from low-grade ores, chemical process aids are often used in bitumen extraction. However, the use of conventional process aids, such as caustic, results in difficulties in tailings treatment because they make the fine particles in a well-dispersed state in the tailings slurry. To overcome such problems, as aforementioned, we used Percol 727 to process a low-grade ore. It was found that the addition of this polymer in the bitumen extraction process not only improved the bitumen recovery but also enhanced the tailings settling.

To understand the role of this polymer in both bitumen extraction and tailings settling, Long et al. (2006b) employed the technique of single molecule force spectroscopy to measure the adhesion forces of single Percol 727 molecules on the surfaces of various oil sand components, such as bitumen, sand, and clay, using an atomic force microscope (AFM). The measured adhesion forces together with the zeta potential values of these surfaces indicate that the polymer would preferentially adsorb onto clay surface than onto bitumen surface. When the polymer was used as a process aid in the extraction process, the polymer-induced formation of large flocs of fine particles reduced the number of individual fine particles in the oil sands slurry. As a result, the chance for slime coating to occur was reduced. This would benefit attachment of air bubbles to bitumen droplets and thus improve the flotation efficiency and consequently bitumen recovery. The formation of large floccules also increased the settling rate of fine solids in the tailings. It is the selective adsorption of HPAM that benefits both bitumen recovery and tailings settling when the polymer was added directly to the bitumen extraction process at an appropriate dosage.

As discussed earlier, both charge density and MW of HPAM polymers are critical in determining their functionality and performance. On the basis of the successful attempt of using Percol 727 to process low-grade oil sand ores, in the current study, the effect of charge density and MW of HPAM polymers on their performance as process aids in low-grade oil sand ore processing was investigated. HPAM polymers with a wide variety of charge densities and MWs were tested. The effect of charge density and MW on bitumen recovery, froth quality, and tailings settling was evaluated. An AFM was used to directly measure the bitumen-solid and solid-solid interaction forces so as to understand the role of MW in tuning these forces. The ultimate goal of this study is for the selection/design of a most effective polymer process aid for low-grade oil sand ore processing.

EXPERIMENTAL

Materials

The oil sand ore used was a transition ore provided by Syncrude Canada Ltd. It contained 8.8, 8.7, and 82.5 wt% of bitumen, water, and solids, respectively. In the solids, 25.9% were fines (defined as mineral particles smaller than 44 microns).

The water used to recover bitumen from the oil sand ore was an industrial process water from the Aurora plant of Syncrude Canada Ltd. It contained about 40 ppm calcium and 17 ppm magnesium and its pH was ~8.2. In the current study, we call this water the Aurora process water.

The polymers chosen are listed in Table 1. They are all HPAM polymers. The MW of these polymers was in the range of 1-40 millions Daltons. To test the effect of charge density, the polymer with a MW of 15 million Daltons was selected and its charge density varied from 10, 30, to 50%. All other polymers had an anionicity of ~30%.

Bitumen Extraction

Bitumen extraction tests were conducted with the Laboratory Hydrotransport Extraction System (LHES). Details about the LHES were described elsewhere (Li et al., 2005). In each test, 1.5 kilograms of the ore sample and 5 L of Aurora process water were used. A polymer solution was prepared at a concentration of 0.05 wt% using deionized water one d prior to its use in the extraction test. The process water was heated to 35[degrees]C and added to the LHES and a predetermined volume of the polymer solution together with the ore sample was then added. The oil sand slurry was conditioned for 5 min in the LHES at a circulating rate of 3 m/s. The temperature was maintained at 35[degrees]C. Then, air was supplied at 200 mL/min and a timer was immediately started to indicate the flotation time. The test proceeded for 1 h with froth collections in 3, 10, 20, 30, 40, and 60 min. The collected bitumen froths were analyzed using the Dean Stark method to determine the contents of bitumen, solids, and water.

Tailing Settling

Tailings slurry samples were directly taken from the bitumen extraction experiments. Settling tests were conducted in 50 mL graduated cylinders. After 50 mL of tailings sample were poured into a cylinder, the cylinder was sealed by a glass stopper, and then it was gently turned upside down several times to mix the slurry. As soon as the cylinder was placed on a bench surface, the settling test was started and no further disturbances were allowed. The descent of the solids/liquid interface (mud line) was carefully observed, recorded, and plotted as a function of settling time. The slope of the settling curve at time zero was obtained as the initial settling rate. Also recorded was the final sediment volume.

Colloidal Force Measurements

A Nanoscope E AFM with a vendor-supplied fluid cell (Digital Instruments, Santa Barbara, CA) was used for the surface force measurement. Gold-coated silicon nitride cantilevers also from Digital Instruments were chosen. Clay particles with a pseudo spherical shape were used as the probe for the force measurements by attaching them onto the apex of a cantilever (lever type 100 [micro]m wide) with a spring constant of 0.58 N/m. The clay particles were chosen under an optical microscope from a great number of particles, which were directly obtained from the tailings slurry of a bitumen extraction experiment without chemical addition. Prior to each set of force measurements, the prepared clay probes were thoroughly rinsed with deionized water and ethanol, followed by blow-drying with ultrapure-grade nitrogen. The probes were then exposed to an ultraviolet light for more than 5 h to remove any possible organic contaminants. A photograph of a prepared clay probe was provided elsewhere (Li et al., 2005). The details of using AFM for colloidal force measurements can be easily found in the literature (Liu et al., 2003). Briefly, in AFM force mode, a triangular waveform is applied to the AFM Z piezo tube. As a result, the sample surface attached to the piezotube moves towards and away from the cantilever tip (the colloid probe) by the extension and retraction of the piezotube. The force acting between the probe and the surface is determined from the deflection of the cantilever using Hooke's law. Each force plot represents a complete extension-retraction cycle of the piezo. When a sample surface approaches a probe, the long-range interaction force between the two surfaces is measured, while the adhesion (or pull-off) force can be obtained during the retraction process. For quantitative comparison, the measured long-range interaction force (F) and adhesion force (pull-off force) were normalized by probe radius (R).

Force measurements were performed in a fluid cell where clay probes interacted with bitumen or silica surface in aqueous solutions that were directly taken from the supernatant of tailing slurry of bitumen extraction. The bitumen surface was prepared by coating a thin layer (~100 nm) of bitumen onto 10x10 mm2 silica wafers (NANOFAB, University of Alberta, Canada) using a spin-coater. The silica wafers had an oxidized surface layer of ~0.6 [micro]m. A detailed description on the preparation of the bitumen surface and the characteristics of the prepared bitumen surface can be found elsewhere (Liu et al., 2003). All force measurements were conducted after an incubation time of 60 min. Preliminary experiments showed that a 60-min period was sufficient for the two surfaces immersed in the aqueous medium to equilibrate. As the surface of the clay probes was quite irregular, each force measurement was performed several times with different clay probes to obtain representative results. All force measurements were conducted at room temperature of 22 [+ or -] 1[degrees]C.

EXPERIMENTAL RESULTS AND DISCUSSION

Effect of Polymer Charge Density

HPAM molecules contain two functional groups, that is, amide (-CON[H.sub.2]) and carboxylic (-CO[O.sup.-]) groups. Charge density or anionicity represents the percentage of the carboxylic groups in a HPAM molecule. An increased number of carboxylic groups in the polymer structure leads to an increase in the polymer anionicity. As discussed in the introduction section, charge density has a significant impact on the performance of a polymer.

To determine the effect of charge density of HPAM polymers on their performance in oil sand processing, three HPAM polymers with the same MW of 15x[10.sup.6] Daltons but different charge densities of 10, 30, and 50% were tested in the current study. Figure 1 presents the experimental results of bitumen recovery obtained for the case of no polymer addition and those using the three polymers as a process aid at a dosage of 20 ppm. In this paper, polymer dosage in ppm refers to mg of polymer in per litre of oil sand slurry. (To convert this slurry volume-based ppm to mg of polymer per kilogram of oil sand, a multiplying factor of 3.8 should be used. For example, 20 ppm on the basis of oil sand slurry volume, mg/L, is equivalent to 76 ppm on the basis of oil sand mass, mg/kg.) For the transition ore processed, without the addition of any chemical aids, the bitumen recovery was only about 52.5%. The use of the HPAM polymer with a 10% anionicity slightly decreased the recovery to about 49%. The recovery was restored to about 52% when the polymer of 30% anionicity was used. If the anionicity of the polymer used was up to 50%, the recovery was significantly dropped to approximately 36%. These results indicate that polymer charge density is critical to their performance.

In the bitumen extraction tests, the oil sand slurry had a pH value of about 8.4 (this is also the pH used in the industrial bitumen extraction). Under such a condition, the surfaces of bitumen, sand grains, and clays are all negatively charged. When an anionic polymer is added as a process aid, there are electrostatic repulsive forces between the polymer and these negatively charged components in the oil sand slurry. At a lower charge density (e.g. 10%), a weak repulsion between the HPAM molecules and the surfaces of bitumen and solids allows the polymer to adsorb on these surfaces by such interactions as hydrogen bonding and van der Waals forces. As a result, the addition of such polymers in the extraction process could induce hetero-coagulation between bitumen and solids, thus reducing the bitumen recovery. When the polymer anionicity is increased to a certain value (e.g. 30%), the repulsion between the polymer and bitumen could be sufficiently strong to prevent the polymer adsorption on the bitumen surface. Thus, bitumen-solid hetero-coagulation would not occur and as a result, bitumen recovery is restored. Furthermore, if the repulsion between the polymer and solids under such circumstance remains weak to allow the polymer adsorption on the surface of solids, the flocculation of fine solids could reduce the number of fine solid particles in the oil sand slurry and consequently lead to an improvement in bitumen recovery. When the polymer anionicity is too high (e.g. 50%), the repulsions between the polymer and other components, including bitumen and solids, in the oil sand slurry are very strong. Thus, it becomes difficult for the polymer to adsorb on the surfaces of these components. As a result, the polymer would stay in the water phase. Instead of flocculating, it may disperse these particles. The bitumen recoverywould then significantly deteriorate. Comparing the three bitumen recoveries obtained with the use of the three polymers of different charge densities, one finds that the polymer of 30% charge density achieved the highest bitumen recovery. Other study (Cymerman et al., 1999) also showed that HPAM polymers of ~30% anionicity performed the best in flocculation. Therefore, 30% was regarded as the optimal charge density for these polymers to achieve the best performance while used as process aids, and in the following study on the effect of MW, only those polymers with a charge density of 30% were chosen.

[FIGURE 1 OMITTED]

In addition to the effect on bitumen recovery, the polymer charge density also affected the quality of the bitumen froth. Figure 2 shows the results of the bitumen-to-solids (B/S) and bitumen-to-water (B/W) ratios in the resulting bitumen froth. A decreasing trend of the B/S ratio with increasing charge density is clearly shown in Figure 2a. A similar trend is also observed in Figure 2b for the B/W ratio. For the cases of charge densities of 10% and 30%, the B/W ratios are higher than that for the case of no polymer addition. The quality of bitumen froth was affected by many factors in the separation, aeration and flotation process. The details are discussed in the later section of AFM force results.

Effect of Polymer MW on Bitumen Extraction

To investigate the effect of polymer MW on bitumen recovery, bitumen extraction tests were conducted using six HPAM polymers with the same charge density of 30% but different MWs (0.01, 1, 15, 17.5, 20, and 40 million Daltons, Table 1). As previous studies have indicated that a dosage of 20-30 ppm often resulted in the best performance of HPAM polymers in oil sand processing (Li et al., 2005) and tailings treatment (Cymerman et al., 1999), in the current study, the same polymer dosage of 20 ppm was used in all the extraction tests. The results of bitumen recovery as a function of polymer MW are shown in Figure 3. With the increase of polymer MW, the variation of bitumen recovery can be divided into three regimes. Starting from the case of no polymer addition up to the use of the polymer with a MW of 1 million Daltons (regime I), the bitumen recovery decreases from about 50 to 45%. With a continual increase of the polymer MW to 17.5 million Daltons (regime II), the bitumen recovery increases and reaches a peak value of 70% at a MW of 17.5 million Daltons. At a further higher polymer MW(regime III), the bitumen recovery is reduced again to 40% at a MW of 40 million Daltons. These results clearly indicate that the polymer MW is critical in determining their functionality and performance.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Figure 4 shows the effect of the polymer MW on the quality of bitumen froths. The B/S and B/W ratios as a function of the polymer MW as shown in Figures 4a and 4b, respectively, indicate the presence of two distinguishable regimes. At a lower MW ([less than or equal to] 1 million Daltons), the use of a polymer aid improved the froth quality by reducing the content of both solids and water in the froths. However, at a higher MW ([greater than or equal to] 15 million Daltons), the froth quality deteriorated.

Figure 5 shows the effect of the polymerMWon tailings settling. The tailing samples used for the settling tests were directly taken from the tailings slurries produced in the bitumen extraction tests where the polymers with different MW were used as a process aid. As shown in the inset of Figure 5a, the sediment height (or the mud-line height, h) was recorded as a function of settling time (t) during each settling test. The plot of h versus t as shown by the open circles in Figure 5a represents a typical settling curve obtained. The slope of the initial portion of this settling curve, as indicated by the short-dash line, is defined as the initial settling rate. Figure 5b shows the initial settling rates as a function of the polymer MW. The initial settling rate for the case of no polymer addition is slightly below 2 cm/min, indicating a slow settling of the fine solid particles in the tailings slurry. When a HPAM polymer with a low MW of 0.01 or 1 million Daltons was used in the bitumen extraction process, the initial settling rate became even lower and decreased with increasing MW. This is because the low-MW polymers acted as a dispersant that kept the fine solid particles in a stable dispersed state in the tailings slurry.

[FIGURE 5 OMITTED]

In contrast to low-MW polymers, polymers with a high MW ([greater than or equal to] 15 million Daltons) induced much faster tailings settling. The initial settling rates for these cases were about 12 cm/min, nearly six-time higher than that without polymer addition. With a high MW, the polymers acted as a flocculant. The bridging effect between the fine solid particles caused by the polymers resulted in the formation of large flocs and thus in a fast settling rate.

To clearly appreciate the effect of the polymer MW on tailings settling, photographs showing the state of tailings samples after a settling period of 24 h are provided in Figure 6. For the case of no polymer addition or the cases with the addition of a low MW polymer (MW [less than or equal to] 1 million Daltons) in the bitumen extraction process, no clear supernatant layer was observed after 24 h of settling, indicating that the fine particles were in a dispersed state and the tailings samples settled very slowly. However, when a high MW polymer (MW [greater than or equal to] 15 million Daltons) was used in the extraction process, a clear supernatant layer was clearly observed. Particularly, the clarity of the supernatant layers was improved with increasing MW from 17.5 to 20 and 40 million Daltons.

[FIGURE 6 OMITTED]

From the results presented in Figures 3 to 5, one can conclude that the role of HPAM polymers with a higher MW (=15 million Daltons) is different from that of HPAM polymers with a lower MW ([less than or equal to]1 million Daltons) in low-grade oil sand ore processing. For low-MW polymers, they acted as dispersants, leading to improved froth quality but deteriorated bitumen recovery and tailings settling. In direct contrast, high-MW polymers acted as flocculants, resulting in deteriorated froth quality but improved bitumen recovery and tailings settling. To achieve a higher bitumen recovery, the HPAM polymer used must have a higher MW. In this study, the optimal MW was about 17.5 million Daltons.

Results of AFM Force Measurements

Clay-bitumen interactions

To understand the effect of HPAM MW on bitumen recovery, the long-range interaction and adhesion forces between bitumen and clay fines in aqueous solutions were measured by AFM. The aqueous solutions used in the force measurements were the supernatant of tailings slurries produced from the bitumen extraction tests.

Figure 7 shows the measured forces between clay and bitumen. The long-range interaction forces (Figure 7a), which were measured during the approaching process of the clay probe towards the bitumen surface, were purely repulsive. It should be noted that because the surface of the clay probes used in the force measurements was rough and irregular, as would be anticipated, the data obtained are highly scattered. Therefore, for each condition, 2-3 representative force profiles are plotted to show a general trend. All the force profiles in Figure 7a show little difference among them. However, a careful comparison shows that when a low-MW polymer (MW = 0.01 or 1 million Daltons) was used in the extraction process, the force profiles (open circles and squares) are slightly above those obtained under the condition of no polymer addition (filled triangles), indicating a slightly stronger repulsion between clay and bitumen. In contrast, a slightly weaker repulsion was obtained when a high-MW MW polymer (MW>15 million Daltons) was used, for example, the inverted triangles for the case of MW=20 million Daltons. As the long-range forces are mainly determined by the electrostatic interactions between clay and bitumen, the repulsive nature of the force profiles obtained indicate that the surfaces of both bitumen and clay were negatively charged. The use of HPAM polymers with varying MW did not impose a significant impact on the surface charges of both clay and bitumen.

[FIGURE 7 OMITTED]

Figure 7b shows the adhesion forces between clay and bitumen obtained during the retraction process of the probe from the bitumen surface after contact was made. The measured clay-bitumen adhesion forces display three variation trends (as indicated by the three regimes) with increasing polymer MW, which are very similar to those of bitumen recovery as shown in Figure 3. In regime I, the clay-bitumen adhesion became slightly stronger from the case of no polymer addition to the cases with a low-MW polymer addition (MW=0.01 and 1 million Daltons). An increasing adhesion between clay and bitumen increases the chance of fine particles to attach to bitumen surface, leading to the occurrence of slime coating, and consequently resulting in a decrease in bitumen recovery. In regime II, the measured adhesion force decreases with increasing MW until the lowest adhesion was reached when the polymer with a MW of 17.5 million Daltons was used in the extraction process. The decreasing adhesion reduces the probability of slime coating and thus increases the floatability of bitumen droplets, leading to an improved bitumen recovery. Corresponding to the lowest adhesion at a MW of 17.5 million Daltons, the highest bitumen recovery (70%) was achieved (Figure 3). With a further increase in the MW (regime III), the clay-bitumen adhesion forces increases again, resulting in a decrease in bitumen recovery.

Clay-silica interactions

In addition to its effect on bitumen recovery, the polymer MW also has a prominent influence on tailings settling as was shown in Figure 5. To understand such an effect, the long-range interaction and adhesion forces between a clay probe and a silica surface in aqueous solutions were also measured by AFM. Figure 8 shows typical results of measured forces. For each condition, 2-3 representative force profiles are plotted in this figure. The long-range interaction forces for the case of no polymer addition as shown by the filled triangles in Figure 8a were purely repulsive. The repulsion became marginally stronger when the polymer with a lower MW of 0.01 million Daltons was used (circles). There is no essential difference among the force profiles (squares and diamonds) at a MW of 1 and 15 million Daltons. A further increase of the MW from 17.5 to 20 and 40 million Daltons caused a noticeable decrease in the repulsion (hexagons to inverted triangles) or even resulted in an attractive force between clay and silica (inverted triangles and crosses). The attractive forces appeared within the separation distance of about 10 nm. In such a short distance, the extruded polymer chains adsorbed on one surface could attach to the other surface by such interactions as hydrogen bonding and van der Waals forces during the approach process in the force measurements. The presence of such attractive forces would cause an easy flocculation between fine particles, leading to a fast tailings settling.

Figure 8b shows the clay-silica adhesion forces as a function of MW. Clearly, the effect of polymer MW on the clay-silica adhesion forces can be simply divided into two categories. At a lower MW ([less than or equal to] 1 million Daltons), the polymers caused a decrease in the solid-solid adhesion. For the case of no polymer addition and for the cases with the addition of a low-MW polymer (MW=0.01 and 1 million Daltons), the solid-solid long-range interaction forces were purely repulsive (Figure 8a) and the adhesion forces were very small. Under such circumstance, the fine solid particles in the tailings slurries would remain in a well-dispersed state. Thereby, the tailings settling would be very slow. With the addition of a higher MW polymer ([greater than or equal to] 15 million Daltons), strong solid-solid adhesion forces were measured (Figure 8b). The corresponding long-range forces were either weakly repulsive or even attractive in the presence of these high-MW polymers (Figure 8a). Thus, flocculation between the fine particle would easily occur, leading to fast tailings settling. These anticipations based on the measured forces are consistent with or confirmed by the results of tailings settling, as indicated by the photographs of Figure 6. The left three photographs in Figure 6 show a well-dispersed state of the fine particles after a settling period of 24 h when no polymer or a low-MW polymer was used in the bitumen extraction process. In direct contrast, a clear supernatant layer is clearly shown in the right four photographs when a high-MW polymer was used. In particular, the supernatant layers for the cases of MW at 20 and 40 million Daltons are much clearer than those for MW at 15 and 17.5 million Daltons. Although these high-MW polymers all caused strong solid-solid adhesion forces, the long-range forces for the cases of MW at 15 and 17.5 million Daltons were repulsive albeit weak. Such weak repulsive forces could still disperse very fine particles. For the cases of MW at 20 and 40 million Daltons, however, the long-range forces became attractive. Thus, it is not surprising that the supernatant layers for these cases were much clearer.

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

Correlation of measured adhesion forces with bitumen recovery, froth quality, and tailings settling

The above discussion on the results of AFM force measurements indicates that bitumen recovery is by and large related to the adhesion between clay and bitumen. Qualitatively, a higher clay-bitumen adhesion often leads to a lower bitumen recovery. This is because a strong adhesion force between clay and bitumen not only makes the separation of bitumen from solids difficult but also causes the slime coating of bitumen surface by the fine clay particles, thus deteriorating bitumen aeration and flotation. Figure 9a shows the results of bitumen recovery as a function of the clay-bitumen adhesion. An approximately linear relation, as indicated by the dotted line, is shown in this figure. However, it should be noted that clay-bitumen adhesion is not the only factor determining bitumen recovery. Many other factors, such as the interactions between bitumen and air bubbles and between fine particles and air bubbles, also affect bitumen recovery.

Figure 9b shows the results of bitumen froth quality against clay-bitumen adhesion force. Clearly, there is no direct correlation between the bitumen-to-solids (B/S) and bitumen-to-water (B/W) ratios and the clay-bitumen adhesion force. As in the bitumen extraction process, the oil sand slurry is a complicated multiphase system. It contains bitumen, sand grains, fine clays, water, chemical aids, and air bubbles. In addition to the effect of bitumen separation from the solids, the quality of bitumen froth is also related to the bitumen flotation process during which solids and water could be entrapped and brought by bitumen droplets and air bubbles to the bitumen froth. The use of chemical aids, such as the HPAM polymers in the current study, can alter the interactions between the various components in the oil sand slurry, thus changing the dispersion state of the system and consequently influencing the froth quality. In particular, the interactions between fine solids and air bubbles can play an important role in determining the amount of fine solids brought by the air bubbles to the bitumen froth. To better understand how polymer charge density and MW affect the quality of bitumen froth, more systematic studies are needed.

Figure 9c shows the relation between the tailings settling rate and clay-silica adhesion force. As indicated by the dotted line, the relation is approximately linear. A stronger adhesion normally results in a higher initial settling rate. This indicates that the settling of fine particles is controlled by the colloidal interactions between them, in particular, the adhesion forces.

CONCLUSIONS

The effect of charge density and MW of HPAM polymers on their performance as a process aid in low-grade oil sand ore processing was investigated. The following conclusions can be drawn from the results of bitumen extraction tests and AFM force measurements.

Low-molecular weight polymers (MW [less than or equal to] 1 million Daltons) acted as dispersants in the bitumen extraction process, leading to improved bitumen froth quality, but deteriorated bitumen recovery and tailings settling.

In direct contrast, high-molecular weight polymers (MW [greater than or equal to] 15 million Daltons) acted as flocculants, resulting in high bitumen recovery and fast tailings settling but deteriorated bitumen froth quality.

To be used as an effective process aid in the processing of low-grade oil sand ores in terms of achieving higher bitumen recovery and fast tailings settling, a HPAM polymer must have a low to medium charge density and a high MW weight. A charge density of 30% and a MW of 17.5 million Daltons were found to be optimal in the current study.

Bitumen recovery was related to clay-bitumen adhesion force. A stronger clay-bitumen adhesion force normally resulted in a lower bitumen recovery. The settling of tailings was controlled by the solid-solid interactions. The presence of an attractive long-range force and a stronger adhesion force between solid particles led to a fast tailings settling and a clearer supernatant. The role of HPAM polymers was to tune the colloidal interactions.

ACKNOWLEDGEMENTS

The financial support for this work from the NSERC Industrial Research Chair in Oil Sands Engineering (held by JHM) is gratefully acknowledge.

Manuscript received November 20, 2006; revised manuscript received April 20, 2007; accepted for publication June 30, 2007.

REFERENCES

Baklouti, S., M. R. B. Romdhane, S. Boufi, C. Pagnoux, T. Chartier and J. F. Baumard, "Effect of Copolymer Dispersant Structure on the Properties of Alumina Suspensions," J. Euro. Ceramic Soc. 23, 905-911 (2003).

Chen, H. T., R. S. Ravishankar and R. S. Farinato, "Rational Polymer Design for Solid-Liquid Separations in Mineral Processing Applications," Int. J. Miner. Process 72, 75-86 (2003).

Cymerman, G., T. Kwong, E. Lord, H. Hamza and Y. Xu, "Thickening and Disposal of Fine Tails from Oil Sand Processing," in "Polymers in Mineral Processing Industry," J. S. Laskowski, Ed., 38th Annual Conference of Metallurgists of CIM, Quebec, Canada (1999), pp. 405-417.

Fan, A., N. J. Turro and P. Somasundaran, "A Study of Dual Polymer Flocculation," Colloids Surface A 162, 141-148 (2000).

Glover, S. M., Y. Yan, G. J. Jameson and S. Biggs, "Dewatering Properties of Dual-Polymer-Flocculated Systems," Intern. J. Mineral Process 73, 145-160 (2004).

Li, H., J. Long, Z. Xu and J. H. Masliyah, "Synergetic Role of Polymer Flocculant in Low-Temperature Bitumen Extraction and Tailings Treatment," Energ. Fuel 19, 936-943 (2005).

Liu, J., Z. Xu and J. H. Masliyah, "Studies on Bitumen-Silica Interaction in Aqueous Solutions by Atomic Force Microscopy," Langmuir, 19, 3911-3920 (2003).

Long, J., H. Li, Z. Xu and J. H. Masliyah, "Role of Colloidal Interactions in Oil Sand Tailings Treatment," AIChE J. 52, 371-383 (2006a).

Long, J., Z. Xu and J. H. Masliyah, "Adhesion of Single Polyelectrolyte Molecules on Silica, Mica, and Bitumen Surfaces," Langmuir, 22, 1652-1659 (2006b).

Moddy, G., "The Use of Polyacrylamides in Mineral Processing," Minerals Eng. 5, 479-492 (1992).

Nagaraj, D. R., A. S. Rothenberg, D. W. Lipp and H. P. Panzer, "Low Molecular Weight Polyacrylamide-Based Polymers as Modifiers in Phosphate Beneficiation," Intern. J. Mineral Process 20, 291-308 (1987).

Nishkov, I. and M. Marinov, "Surface Properties of Talc and Their Effect on the Behaviour of Talc Suspensions," in "Mineral Processing on the Verge of the 21st Century," Exitor Kuzev, Djiev Trade, Sofia (2003), pp. 700-705.

Ovenden, C. and H. Xiao, "Flocculation Behaviour and Mechanisms of Cationic Inorganic Microparticle/Polymer Systems," Colloids Surface A 197, 225-234 (2002).

Pawlik, M., "Polymeric Dispersants for Coal-Water Slurries," Colloids Surface A 266, 82-90 (2005).

Pearse, M. J., "An Overview of the Use of Chemical Reagents in Mineral Processing," Mineral Eng. 18, 139-149 (2005).

Sabah, E., H. Yuzer and M. S. Celik, "Characterization and Dewatering of Fine Coal Tailings by Dual-Flocculant Systems," Intern. J. Mineral Process 74, 303-315 (2004).

Seyrankaya, A., U. Malayoglu and A. Akar, "Flocculation of Marble from Industrial Wastewater and Environmental Consideration," in "Mineral Processing on the Verge of the 21st Century," Ozbayoglu, Ed., Balkema, Rotterdam (2000), pp. 645-652.

Xiao, H., Z. Liu and N. Wiseman, "Synergetic Effect of Cationic Polymer Microparticles and Anionic Polymer on Fine Clay Flocculation," J. Colloid Interface Sci. 216, 409-417 (1999).

Yoon, S. Y. and Y. Deng, "Flocculation and Reflocculation of Clay Suspension by Different Polymer Systems under Turbulent Conditions," J. Colloid Interface Sci. 278, 139-145 (2004).

Haihong Li, Jun Long, Zhenghe Xu and Jacob H. Masliyah *

Department of Chemical and Materials Engineering, University of Alberta, Edmonton, AB, Canada T6G 2G6

* Author to whom correspondence may be addressed.

E-mail address: jacob.masliyah@ualberta.ca

DOI 10.1002/cjce.20029

Table 1. Polymers tested in the current study

MW million Daltons Charge density (%) Provider

 0.01 30 Fisher Scientific
 1 30 Fisher Scientific
15 10 Champion
 30 Technologies
 50
17.5 30 Cytec
20 30 Ciba
40 30 Nalco