Interactive effects of clay and polyacrylamide properties on flocculation of pure and subsoil clays.
Polyacrylamide (PAM) is a synthetic polymer and has been widely used around the world as a soil amendment (Orts et al. 1999). Various PAMs with different charge properties and molecular weights have been used to control soil erosion, reduce clay dispersion, stabilise soil aggregates and soil structure, increase water infiltration, and reduce surface runoff (Shainberg et al. 1990; Ben-Hur et al. 1990; Levy et al. 1992; Barvenik 1994; Lentz and Sojka 1994; Levy and Agassi 1995; Lentz et al. 1996; Green et al. 2000; Vacher et al. 2003; Ben-Hur 2006; Mamedov et al. 2009). PAM has been also used in water treatment and environmental protection, including solid-liquid separation in clarification of waste water, runoff water, and irrigation water (Barvenik 1994; Lentz et al. 2000). For example, PAM was used to control irrigation-caused soil erosion and enhance water infiltration on ~800 000 ha of irrigated lands in USA (Sojka et al. 2007).
Soil erosion is a serious problem in China. About 38% of the territory experiences various degrees of soil erosion. In the past decade, research on using PAMs to reduce soil erosion in China has achieved promising results (Xia et al. 2002; Lei et al. 2003; Tang et al. 2003; Yan and Zhang 2008; Wang et al. 2011). However, the results obtained from those studies are preliminary and have not been applied at a large scale to control soil erosion in China. Further experimental studies are needed to improve understanding of the mechanisms of PAM-clay interactions, and to develop a knowledge base that can be used to optimise decisions on matching the type and rate of PAM with type of soil.
In the process of natural rainfall, raindrops hit the soil surface, causing aggregate breakdown and clay dispersion. The dispersed, fine clay particles can be suspended in surface water for days or even weeks, and can be easily transported over long distances to rivers and lakes downstream. Muddy water is a major threat to aquatic organisms by reducing light penetration, and it considerably degrades surface water quality. PAM has been used in industrial wastewater treatment and sewage sludge clarification (Barvenik 1994), turbidity reduction of flood irrigation water (Lentz and Sojka 1994; Lentz et al. 1996), and flocculation of suspended sediment in streams and storm-water ponds (Mason et al. 2005; Sojka et al. 2007; Bolto and Gregory 2007). However, studies on the use of PAM to reduce turbidity of storm runoff water and to clarify ponds and reservoirs are largely non-existent in China, and need to be conducted to promote and document the potential use of PAM to control surface water turbidity in highly eroded areas.
Soil physicochemical properties are largely determined by the chemical properties of soil clays. Major soil clays comprise kaolinitc, montmorillonite-smectite, vermiculite, mica, and chlorite. Kaolinite and montmorillonite arc the most common clays in natural soils. The former predominates in highly weathered subtropical and tropical soils, whereas the latter prevails in the less weathered soils in the temperate zone. Thus, understanding the interaction between variously charged PAMs and these two common clay minerals, i.e. the interactive effects of PAM and clay properties on clay flocculation and PAM adsorption and desorption, is essential to providing scientific knowledge for use of PAM in different soils or regions to reduce soil erosion and to improve soil physical properties. Soils in different regions often have different soil properties and clay mineralogy, and require different PAM type and application rates for optimal amelioration.
The objectives of this study were (i) to investigate the mechanisms of PAM adsorption and desorption by pure kaolinitc and montmorillonite specimens as well as the ability of PAMs to flocculate the two clays, and (ii) to further test and evaluate the ability of PAMs to clarify and flocculate clay suspensions of four natural subsoils having distinct clay mineralogy.
Materials and methods
Interactions between PAMs and pure clay
Properties of clay specimens and PAMs
Pure kaolinitc and montmorillonite specimens were used for PAM adsorption and desorption studies. The kaolinite (Xinyang, Henan Province) has a surface area of 70 [m.sup.2] [g.sup.-1], an area charge density (CD) of 0.9 [micro][mol.sub.c] [m.sup.-2], and a mass CD of 0.07 [mmol.sub.c] [g.sup.-1]. The montmorillonite (Xixiang, Shaanxi Province) has a surface area of 790 [m.sup.2] [g.sup.-1], an area CD of 1 [miro][mol.sub.c] [m.sup.-2], and a mass CD of 0.8 [mmol.sub.c] [g.sup.-1]. Three PAMs, Magnifloc 494C, Superfloc N300, and Superfloc A110 (Cytec Industries Inc., West Patterson, NJ, USA), with different molecular weights and charge properties were used (Table 1). The cationic Magnifloc 494C has a medium CD with 20% of units being positively charged. The anionic Superfloc A110 has an 18% hydrolysis rate (18% units are hydrolysed and negatively charged). The non-ionic Superfloc N300 is slightly negatively charged.
Pre-treatment of clay specimens
We weighed 10 g of each clay specimen into a 40-mL centrifuge tube, added 25 mL of 1 m NaCl[O.sub.4] solution, and shook it for 24 h. Following shaking, we centrifuged the mixture at 8000 r.p.m. for 30 min, decanted the supernatant, and then added 25 mL of 1 m NaCl[O.sub.4] for the second wash. After the third wash with 1 m NaCl[O.sub.4], the clay was washed with deionised water twice to remove access sodium ions ([Na.sup.+]). Finally, the pretreated, Na-saturated clay was transferred to a beaker, and deionised water was added to obtain a clay colloid concentration of 40 g clay [L.sup.-1] for later use.
For each clay and PAM type, we set up seven 10-mL test tubes, and added 0.15mL of the clay colloid suspension (40 g [L.sup.-1]) to each tube with a micropipette. Then we added predetermined amounts of the stock solution of PAM (1000 mg [L.sup.-1]) to each tube and brought the volume to 10 mL with deionised water. The final clay concentration in each tube was ~0.6 g [L.sup.-1], and the PAM concentrations were 0, 20, 40, 60, 80, and 100 mg [L.sup.-1] for anionic and non-ionic PAM, and 0, 10, 20, 30, 40, 50, and 60 mg [L.sup.-1] for cationic PAM. Each tube was shaken by hand for ~1 min and allowed to stand for 6 h. Three mL of suspension was taken near the surface, with a pipette, from each tube, and light absorption was measured at 600 nm using a UV-visible spectrophotometer (model UV-754; Jinhua Scientific and Technological Instrument, Shanghai). A standard linear equation between light absorbance and clay colloid concentration was established for each clay by measuring absorbance of clay colloids of known concentrations. This linear equation was then used to calculate clay colloid concentration using measured light absorbance.
PAM adsorption isotherm
We transferred 5mL of the 40-g[L.sup.-1] clay suspension to a 40-mL centrifuge tube, added a predetermined amount of the 1000-mg [L.sup.-1] PAM stock solution, and brought the volume to 25 mL with deionised water. No replication was made. The clay concentration in the final mixture was ~8 g [L.sup.-1], and PAM concentration ranged from 0 to ~700 mg [L.sup.-1]. Intervals of PAM concentrations were smaller at low concentrations (<100 mg [L.sup.-1]) and larger at high concentrations (>100 mg [L.sup.-1]). The mixture was shaken for ~24 h at room temperature, and then centrifuged at 8000 r.p.m. for 30 min. Three mL of the supernatant was taken, and light absorbance was measured at 190 nm using a UV-visible spectrophotometer (model UV-754). The measured absorbance was used to calculate the PAM concentration with a standard linear equation established between absorbance and PAM solution of known concentrations. The PAM adsorption, [M.sub.s], (mg PAM [g.sup.-1] clay) was calculated as:
[M.sub.s] = V([C.sub.0] - C)/W
where [C.sub.0] is the initial concentration of PAM (mg [L.sup.-1), C is the equilibrium PAM concentration following adsorption (mg [L.sup.-1]), V is the suspension volume (L), and W is the clay mass in suspension (g).
Following the absorption test, one concentration from each of the three PAMs was arbitrarily selected for a further desorption test. The initial PAM concentrations of 200, 400, and 400 mg [L.sup.-1] were selected for anionic, cationic, and non-ionic PAM, respectively. The supernatant of each adsorption solution was decanted, and ~25mL of deionised water was added to each centrifuge tube. The tube was shaken for 24 h, and then centrifuged at 8000 r.p.m. for 30 min. The supernatant was collected, and the washing was repeated four times. The supernatant was measured for light absorbance at 190nm with the spectrophotometer. Total PAM desorption was calculated by multiplying total wash volume by the PAM concentration of the wash solution. The desorption rate was computed as the percentage of the total desorbed PAM over the total adsorbed amount.
Interactions between PAM and soil clay
Properties of soil and PAM
Four subsoils with low organic carbon content were collected from Shandong and Jiangxi provinces to represent montmorillonite- and kaolinite-dominated soils, respectively. This is to extend the results of pure montmorillonite-PAM or kaolinite-PAM interactions to whole soils with similar clay minerals and to compare the differences in clay flocculation behaviours in response to PAM addition in the two distinct systems. Subsoils were used to minimise the interference of natural organic matter with the adsorption and desorption of the synthetic organic PAMs. The collected subsoils were air-dried and passed through a 2-mm sieve. Particle size distribution was measured with a hydrometer, and clay mineral composition with an X-ray diffractometer (D/Max2550VB+/PC; Rigaku Corporation, Tokyo).
Physical and chemical properties of the four subsoil samples are shown in Table 2. Samples 1 and 2 were sand soils with [less than or equal to] 4.6% clay and [greater than or equal to] 90% sand. Soil pH in 1:1 soil: [H.sub.2]O suspension was slightly basic, ranging from 7.4 to 8.0. Organic carbon was very low (<0.2%). The exchangeable calcium ([Ca.sup.2+]) was relatively high (4.3-9.0 [cmol.sub.c] [kg.sup.-1]). Sample 3 was clay soil with 50% clay and 28.5% sand. Sample 4 was clay loam with 32% clay and 37.5% sand. Both samples 3 and 4 had low organic carbon (<0.35%), low exchangeable [Ca.sup.2+] and magnesium ([Mg.sup.2+]) (<0.9 [cmol.sub.c][kg.sup.-1]), and low soil pH (~5).
Four major types of clay minerals were found in all four subsoil samples (Table 3). The four clay minerals were kaolinite, montmorillonite, vermiculitc, and mica. The clay minerals in samples 1 and 2 were predominantly montmorillonite, whereas those in samples 3 and 4 were primarily kaolinite, accounting for >80% of the clay minerals. Samples 1 and 2 had relatively high exchangeable [Ca.sup.2+].
Environmental toxicity is a concern when using PAMs to ameliorate soils and clarify surface water bodies. Anionic and non-ionic PAMs are reportedly environmentally friendly, whereas cationic PAMs have the potential to damage aquatic life (Barvcnik 1994). Considering the potential toxicity of cationic PAMs to aquatic organisms, as well as the fact that the natural organic matters are negatively charged, we tested only neutral and anionic PAMs for potential use in clarifying runoff water, farm ponds, and reservoirs. The molecular weight and CD of the six non-ionic and anionic PAMs used are shown in Table 1. They included Superfloc products (N300, A110, A100, 1606; Cytec Industries), Soilfix Polybead (Ciba Speciality Chemicals, Suffolk, VA, USA), and Chemtall 923 VHM (Chemtall, Riceboro, GA, USA). The WGZ-2000P turbidimeter (Shanghai Xinrui Instrument Co. Ltd) was used to measure turbidity of clay suspensions.
We weighed 5g of subsoil into a 100-mL beaker, added 100mL of deionised water, and stirred the mixture with a glass rod. We added a predetermined amount of the 1000-mg [L.sup.-1] PAM stock solution with a micropipette, stirred it for 10 s, let it stand for 30 s, and then measured turbidity of the suspension with the WGZ-2000P turbidimeter with the probe between 1 and 4 cm depth. Three replicates were made for each PAM concentration, and the mean turbidity was used in the analysis. The initial PAM concentrations in the flocculation series ranged from 0 to 10 mg [L.sup.-1], based on a preliminary test. Turkey's multiple range comparison was used to test the turbidity differences at each PAM concentration for all four subsoils (SAS version 8, 2000; SAS Institute, Cary, NC).
Results and discussion
Interactions between PAM and pure clay
Effects of PAM charge type on clay flocculation
Figure 1a showed that there was no clay flocculation for Na-saturated clay colloids when anionic PAM was present. This indicates that the anionic PAM, in the absence of metal cations in the aqueous solution, acted as a dispersant that stabilised the already dispersed clay colloids. Without a cation bridge, montmorillonite and kaolinite particles, both having permanent negative charge, would repel the negatively charged PAM molecules and thus form a more stable colloidal suspension.
When cationic PAM was added to the dispersed clay colloids, clay flocculation occurred immediately. For the montmorillonite suspension, clay flocculation increased (i.e. turbidity decreased) gradually as cationic PAM concentration increased up to 40 mg [L.sup.-1] and then levelled off afterwards, indicating that the optimal flocculation concentration was 40 mg [L.sup.-1] (Fig. 1b). For the kaolinite suspension, clay flocculation occurred rapidly as the PAM concentration increased to 10 mg [L.sup.-1]; however, the trend was reversed when the PAM concentration was >10 mg [L.sup.-1], showing that more clay particles were dispersed as the PAM concentration exceeded 10 mg [L.sup.-1]. It was clear that the optimal flocculation concentration was 10 mg [L.sup.-1] for this colloidal suspension (Fig. 1b), indicating that kaolinite was easier than montmorillonite to flocculate. The reverse trend above 10 mg [L.sup.-1] concentration might have been caused by gradual charge reversal (Theng 1982; Seybold 1994). Kaolinite had fewer permanent negative charges than montmorillonite. As the PAM concentration increased, more PAM molecules would be available for adsorption on clay particles, and the adsorbed cationic PAM would neutralise the small negative charge of the kaolinite, eventually resulting in charge reversal. Positively charged clay-PAM complexes would repel each other and form stable colloidal suspension. Another possible reason for more dispersion at higher PAM concentrations may be steric repulsion (Gregory 1989). When excess PAM is adsorbed to clay particles at the train segments or the adsorption points, tails and loops of the linear PAM chains are formed and extend into the aqueous solution (Theng 1982). This thick layer of PAM tails and loops keeps the separations between clay particles large enough to prevent van der Waals attraction from operating and inducing flocculation (Gregory 1989).
The optimal flocculation concentration of the non-ionic PAM was ~20 mg [L.sup.-1] for both montmorillonite and kaolinite suspensions (Fig. 1c). For the montmorillonite suspension, the flocculation level remained largely unchanged when PAM concentration was >20 mg [L.sup.-1]. Flowevcr, for the kaolinite suspension, flocculation declined when PAM concentration was >20mgL and diminished to near zero at 40 mg [L.sup.-1], where the turbidity was similar to that at 0 mg [L.sup-1] concentration. The driving forces for non-ionic PAM adsorption were the van der Waals force (Seybold 1994), entropy change (Theng 1982), and hydrogen bonds between clay particles and PAM molecules (Laird 1997). The different behaviours between montmorillonite and kaolinite were likely caused by differences in their physicochemical properties. Compared with montmorillonite, kaolinite normally has larger platelet, less specific surface area, less permanent negative charge, higher edge: face area ratio, and more positive edge charge due to broken ionic bonds. The minuscule negative charge of the non-ionic PAM would be attracted to positive charge at the edge surface to neutralise the positive charge. More PAM molecules might have been attracted to the positive charge sites at higher PAM concentrations, resulting in overall negatively charged clay-PAM complexes and thus forming stable colloidal suspension (a phenomenon of charge reversal). Overall, this study showed the general order of flocculation power was cationic > non-ionic > anionic. This result was consistent with the order reported by Helalia and Lctey (1988) who studied PAM adsorption by three soils.
PAM adsorption isotherm by pure clay
The adsorption levels of anionic PAM by Na-saturated montmorillonite and kaolinite colloids were both very low, and the maximum adsorptions were <10 mg [g.sup.-1] (Fig. 2a). The low level of adsorption was due to a lack of divalent cations in the suspension. Without divalent cation bridges between negatively charged clay particles and PAM molecules, mutual electrostatic repulsion would keep clay particles and PAM molecules apart, and the separation would be much greater than the threshold distance at which the van der Waals attractive force begins to operate. As a result, greater adsorption did not occur.
The adsorption levels of cationic PAM by montmorillonite were much greater than by kaolinite (Fig. 2b). This is because montmorillonite had more permanent negative charge than kaolinite, and more negative charge would attract more positively charged PAM. For kaolinite, the adsorption amounts increased as the PAM concentration increased, reached a maximum near the PAM concentration of 50 mg [L.sup.-1], and then levelled off. For montmorillonite, the PAM adsorption amounts increased rapidly as PAM concentration increased to up to 30 mg [L.sup.-1], and the rate of the increase declined afterwards, reaching maximum capacity near 120 mg [L.sup.-1]. Strangely, the adsorption amounts decreased gradually at PAM concentrations >120 mg [L.sup.-1]; the reason for this is unclear. There are two possible explanations for this rare phenomenon. First, because of the great flocculating power of cationic PAM, high concentrations of cationic PAM can cause rapid flocculation and form large floes. The large floes may limit access to some adsorption sites inside floe cavities and therefore reduce total PAM adsorption (Ben-Hur et al. 1992). Second, cationic PAM molecules are more stretched and extended at lower concentrations, whereas they are normally coiled up at higher concentrations. More extended molecules can reach more adsorption sites of clay particles and therefore increase PAM adsorption. On the other hand, the coiled PAM molecules surrounding clay particles would limit close contacts because of steric hindrance between PAM molecules and adsorption sites of clay particles, and thus reduce total adsorption (Misra 1996).
The adsorption of the non-ionic PAM on montmorillonite and kaolinite was much greater than that of the cationic and anionic PAM (Fig. 2c). The non-ionic PAM had a negligible amount of negative charge, and there was no electrostatic repulsion between clay particles and PAM molecules. As a result, PAM molecules could get very close to the exterior surfaces of clay particles. The close association between clay particles and PAM molecules allowed the attractive van der Waals force to dominate, resulting in massive adsorption. This explanation is corroborated by the fact that the PAM adsorption was much greater by montmorillonite than by kaolinite, because the surface area of the montmorillonite was 10 times that of the kaolinite. There are two other possible explanations why large amounts of non-ionic PAM were adsorbed on both clay colloids. The lack of electrostatic repulsion allowed close contacts between clay particles and PAM molecules, and the close contacts allowed (i) the formation of H-bonds between the two (Laird 1997), and (ii) replacement of the neatly ordered water molecules at the clay surfaces by PAM molecules (Theng 1982), which is a spontaneous process driven by an increase in entropy of the system.
The isothermal adsorption curves in Fig. 2 showed that PAM adsorption was mainly determined by PAM charge properties as well as by the physicochemical properties of the clay minerals. The adsorption capacities of both kaolinite and montmorillonite in the absence of divalent cation bridges followed the order non-ionic PAM > cationic PAM > anion PAM. Clay charge type and density, as well as its specific surface area, considerably affected the quantity of PAM adsorption. This result seemingly deviates from that of Aly and Letey (1988), who studied PAM adsorption by Wyoming Na-montmorillonite and reported that the general adsorption order was cationic > non-ionic > anionic. The maximum PAM concentration used in their study was 100 mg [L.sup.-1]. The same adsorption order would be concluded from the present study if the adsorption rates below the concentration of 100 mg [L.sup.-1] were the only ones considered.
PAM desorption by pure clay
Figure 3 shows that PAM desorption from clay particles was essentially irreversible. Less than 3% of the total adsorbed PAM was desorbed after four washings with deionised water for both montmorillonite and kaolinite. These results were in accord with those in the literature (Theng 1982; Letey 1994), and clearly indicated that PAM molecules, once adsorbed by clay particles, were unlikely to desorb regardless of the charge properties of PAM and the clay type. There are three possible explanations for the irreversible adsorption. First, once adsorbed, intermolecular H-bonds can be formed between PAM molecules and clay surfaces. Second, the attractive van der Waals force becomes predominant once the PAM and clay are in close contact. Third, the long-chain PAM molecules may have multiple train segments over a clay particle or even multiple clay particles. It seems impossible to detach simultaneously at the multiple train segments (Theng 1982; Letey 1994).
Effectiveness of PAM in flocculating soil clay
For the sand soil of sample 1, the nephelometric turbidity unit (NTU) decreased only slightly as the concentrations of all four PAMs increased (Fig. 4), indicating that all four PAMs were not good flocculants for the soil. When the PAM concentration was <0.1 mg [L.sup.-1], there were no differences in turbidity among four PAMs (P>0.05). When the PAM concentrations were in the range 0.1-0.8 mg [L.sup.-1], turbidity was least for the PAM with 7% CD, intermediate for the PAM with 18% and 30% CD, and greatest for the non-ionic PAM. However, when PAM concentrations were near 2 mg [L.sup.-1], the turbidity followed the order non-ionic <7% CD<18% CD <30% CD, indicating that the flocculating power was inversely related to the CD at high concentrations. Clearly, the PAM flocculation power varied not only with PAM concentration but also with PAM charge property. The dynamic effects of CD and PAM concentration on PAM flocculation power revealed complex interactions between PAM and clay particles, which were influenced by chemical properties of both PAM and clay minerals. As PAM concentration increased, more PAM molecules would be adsorbed to clay particles. The total negative charge of the clay-PAM complex, being equal to the sum of the negative charges of the clay and PAM, would be proportional to the total amounts of PAM adsorption. As a result, the greater the CD of the PAM, the more total negative charge the clay-PAM would have. The negatively charged complexes would repel each other and thus reduce the chance of flocculation. This might explain why the flocculation power followed the order nonionic > 7% CD > 18% CD > 30% CD at the relatively high PAM concentration of 2 mg [L.sup.-1]. Overall, the low flocculation rate of this particular soil might be an integrated result of the following three factors. First, montmorillonite platelets were very small (generally much smaller than kaolinite). Even though the flocculation occurred, the floes were unlikely to be big enough to settle to below the depth of 4 cm within 30 s. Second, the silt fraction of the soil was only 5.4%. The fewer silt-sized particles in suspension would largely eliminate chances of forming large floes with clay particles. Third, there was a lack of divalent cation bridges in the mixture. The exchangeable [Ca.sup.2+] (4.3 [cmol.sub.c] [kg.sup.-1]) and [Mg.sup.2+] (0.2 [cmol.sub.c] [kg.sup.-1]) were relatively low in this soil.
Sample 2 was more responsive than sample 1 to the PAM addition (Figs 4 and 5). The suspension turbidity decreased in a power function as PAM concentration increased for all four PAMs (note that the x-axis is a logarithmic scale) (Fig. 5). The differential responses between samples 1 and 2 were attributable to the difference in concentrations of exchangeable [Ca.sup.2+] on the exchange sites of the clay minerals. Sample 2 had the greatest content of exchangeable [Ca.sup.2+], of 9.0 [cmol.sub.c] [kg.sup.-1], on the exchange sites, which formed cation bridges between negatively charged PAM and clay particles and resulted in rapid flocculation. Apart from the difference in exchangeable [Ca.sup.2+] between samples 1 and 2, the other physicochemical properties of the two soils were similar. Both had similar clay minerals, being predominantly smectitic. Both had sand fractions of >90%, and silt and clay fractions of <10%. Soil pH, organic carbon, and exchangeable [Mg.sup.2+] were almost identical between the two samples. The contrasting results of samples 1 and 2 strongly indicated that the presence of divalent cation bridges was critical for adsorption of anionic PAM on clay (Theng 1982; Letey 1994; Laird 1997).
The relative flocculation powers of four PAMs varied with PAM concentration (Fig. 5). When the PAM concentration was <0.08 mg [L.sup.-1], the PAM with 7% CD was more effective than the other three in flocculating the dispersed clay particles. When the PAM concentration was 0.08-0.8 mg [L.sup.-1], the PAM with 18% CD was the most effective and the non-ionic PAM the least, with the PAMs of 7% and 20% CD intermediate. When the PAM concentration increased from 2 to 4 mg [L.sup.-1], the flocculation power seemed to increase slightly for the non-ionic PAM but decreased slightly for the PAM with 18% CD, with 7% and 20% CD PAMs being largely unchanged. When the PAM concentration was 4mgL the flocculation power followed the order 0% [greater than or equal to] 7% CD >18% CD >20% CD. This sequence was the same as that of sample 1 at the concentration of 2 mg [L.sup.-1]. This result indicated that when PAM concentration was above a critical concentration, PAMs with greater CD would be less effective in flocculating clay. This could be explained by charge reversal of the clay-PAM complex (Theng 1982; Seybold 1994). PAMs with higher CD could neutralise more positive charges of clay, including broken bonds at the platelet edge surface and cations on the exchange sites, and would result in PAM-clay complexes that had net negative charge. The negatively charged complexes would tend to repel each other and thus reduce chance of coagulation.
The turbidity of sample 3 decreased rapidly from 1800 to <50 as PAM concentration increased from 0 to 2 mg [L.sup.-1] (Fig. 6) (note the logarithmic scale in x-axis). There were no significant differences in terms of the flocculation power among the four PAMs at nearly all concentrations (P=0.05), indicating that the CD and molecular weight of the PAM had little effect on the flocculation power in this particular soil. The rapid reduction in turbidity upon PAM addition was largely attributable to the effects of the physicochemical property of this particular soil. This soil had 21.5% silt and 50% clay, of which kaolinite accounted for >80%. Soil pH was ~5.1. ICaolinite platelets are generally much larger than those of montmorillonite, and normally have larger edge surface area and more variable charge sites at the edge surfaces than montmorillonite. The edge surfaces of kaolinite arc positively charged at acid pH (e.g. pH 5) because of protonation (Miller et al. 1990). The positively charged edges of clay platelets would electrostatically adsorb negatively charged PAM molecules, leading to rapid flocculation. This result is consistent with the findings of Peng and Di (1994), who reported that the optimal pH for flocculating kaolinite was 5-6. In addition, larger platelets of kaolinite would form larger floes that would settle out faster. High silt content would also enhance the chance of forming large floes with flocculating clay particles. In addition, high clay content in the suspension increases the probability of clay particle collision and therefore improves the chance of coalescence.
The flocculation behaviour of sample 4 (Fig. 7) was similar to that of sample 3 (Fig. 6), indicating that the PAM flocculation power was similar for both soils. This was because both soils had similar physicochemical properties (Table 2). Both soils had relatively high clay contents, low organic carbon, low exchangeable [Ca.sup.2+] and [Mg.sup.2+], and low pH (near 5.0). Both soils had a similar mixture of clay minerals, being overwhelmingly kaolinitic (Table 3). When the PAM concentration was <0.1 mg [L.sup.-1], there were no statistical differences in flocculation power among the four PAMs (Fig. 7). When the PAM concentration was 0.1-1 mg [L.sup.-1], the PAM with 30% CD and 28 x [10.sup.6] g [mol.sup.-1] molecular weight was statistically more effective in flocculating clay than the PAM with 18% CD. Interestingly, for the two PAMs with 30% CD, the Superfloc 1606 with higher molecular weight (28 x [10.sup.6] g [mol.sup.-1]) appeared more effective than the Soilfix Polybead with lower molecular weight (16 x [10.sup.6] g [mol.sup.-1]), but the difference was not statistically significant (P>0.05). When the PAM concentration was 2 mg [L.sup.-1], all four PAMs tested had similar flocculation ability, with the turbidity being reduced to near zero.
The flocculation power of PAM is strongly influenced by the physicochemical properties of clay minerals. Each clay type responds to each PAM type differently in terms of the rate and magnitude of the clay flocculation. Furthermore, a mixture of clay minerals often behaves differently from each individual clay mineral alone. The effectiveness of PAM on clay flocculation in any soil is different from that in any pure clay mineral. Since soils often contain a variety of clay minerals, any flocculation result obtained using pure clay cannot be directly applied to any soil with mixed clay minerals, as is demonstrated here. For example, Goldberg and Glaubig (1987) reported that mixing kaolinite mineral with a small amount of montmorillonite would considerably increase the PAM's critical flocculation concentration. This might be because the smaller negatively changed platelets of montmorillonite could be electrostatically attracted to the positively charged edge surfaces of the larger platelets of kaolinite, resulting in charge neutralisation and therefore reducing the chance of edge-tofacc association among kaolinite particles. Flowever, this study showed that the PAM's critical flocculation concentrations for the whole soils were generally much lower than for the two pure, Na-saturated clay minerals. One likely reason might be the differences in the exchangeable cations. The pure clays were saturated with monovalent Na, whereas the whole soils had fair amounts of divalent cations. It was reported that anionic PAM was much more effective in flocculating Ca-clay than Na-clay (Laird 1997). Another plausible reason would be that the coarse fraction in soils including fine sand and silt particles might facilitate the formation of large floes with small clay floes or even clay particles. Such a scavenging phenomenon would greatly increase the flocculation effectiveness of the PAM in soils compared with pure clays.
Because anionic PAM is harmless to the environment and similar in charge type to the natural soil organic matter (natural macromolecular compound), it is widely used as a soil amendment (Sojka et al. 2007). In the absence of a cationic bridge, anionic PAM may be repelled by negatively charged clay particles and help stabilise clay colloids. On the other hand, anionic PAM can be an effective flocculent in the presence of divalent cations such as [Ca.sup.2+]. Cation type and concentration in soil solution as well as on the exchangeable sites directly affect the ability of an anionic PAM to flocculate clay and to stabilise soil aggregates (Peng and Di 1994; Laird 1997). Therefore, in practical application, in order to ensure the effectiveness of soil improvement, anionic PAM and gypsum are often applied together (Mamedov et al. 2009).
The interactive effects between PAM and clay minerals depended primarily on the physicochemical characteristics of both PAM and clay minerals. Chemical properties such as PAM charge type, CD, and molecular weight, as well as clay type and its exchangeable cations, all affected the magnitudes of PAM adsorption and therefore clay flocculation behaviour. Specifically, the following conclusions were drawn from the study with pure kaolinite and montmorillonite. First, the adsorption amounts of three variously charged PAMs by both clay minerals followed the order non-ionic > cationic > anionic, indicating the roles of charge type and density of PAM in determining adsorption when divalent cations were absent. The adsorption amounts of the three PAMs by Namontmorillonite were consistently greater than by Nakaolinite, showing the effects of charge quantity and specific surface areas of clay minerals on PAM adsorption. Second, despite clay and PAM type, PAM adsorption on clay minerals was irreversible. Third, in the absence of cation bridges, cationic PAM was an effective flocculent, whereas anionic PAM behaved as a colloid stabiliser for both clays. The flocculation power for both clays followed the order cationic > non-ionic > anionic. To increase flocculation ability of anionic PAM, divalent cations such as [Ca.sup.2+] must be present. This is why in practical application, gypsum and anionic PAM are usually applied together to enhance PAM adsorption, clay flocculation, and soil aggregate stability. Although cationic PAM is a better flocculent than anionic and non-ionic PAMs, it should be avoided for clarifying surface water bodies because of its potential harm to aquatic life.
Flocculation series tests with whole subsoils showed that the turbidity readings of the four soils reached their minima when PAM concentrations were 1-3 mg [L.sup.-1], indicating that the optimum flocculation concentration of all PAMs tested was 1-3 mg [L.sup.-1]. except for sample 1. The anionic PAMs tested were more effective in flocculating the clay (sample 3) and clay loam (sample 4) soils, which had predominant kaolinite, than in flocculating the sand soils (samples 1 and 2), which had prevalent montmorillonite, largely because kaolinite had a larger platelet, more positive edge charge, and higher clay and silt concentrations in the suspensions. The CD of the PAMs tested had greater differential effects in flocculating montmorillonite-dominant soils than kaolinite-dominant soils. For sample 1, when PAM concentration was 0.1-1 mg [L.sup.-1], anionic PAM with 7% CD was more effective than non-ionic PAM in flocculating clay; however, at 2 mg [L.sup.-1], flocculation power followed the order 0% [greater than or equal to] 7% CD [greater than or equal to] 18% CD >30% CD. For sample 2, when PAM concentration was <0.08 mg [L.sup.-1], 7% CD PAM was more effective than 18% CD PAM, and when PAM concentration was 0.08-1 mg [L.sup.-1], 18% CD PAM was more effective than non-ionic PAM. However, at the highest concentration of 4 mg [L.sup.-1], flocculation power followed the order 0% [greater than or equal to] 7% CD >18% CD>20% CD. The flocculation series of both samples 1 and 2 (montmorillonite-dominant sand soils) showed that as PAM concentration increased, PAMs with higher CD tended to become less effective than non-ionic PAMs or PAM with lower CD in flocculating montmorillonite-dominant soil clay. This result indicated that the greater the total net negative charge of clay-PAM complex, the harder it is for clay particles to coalesce because of electrostatic repulsion, suggesting that optimum flocculation concentrations for higher CD PAMs would be lower than for lower CD PAMs because of charge reversal. The dependency of the PAM's relative flocculation power on PAM concentration reflected complicated interactions between PAM molecules and clay particles. For samples 3 and 4 (kaolinite-dominant soils), there were essentially no statistical differences in flocculation power for all anionic PAMs with CD ranging from 7% to 30%, except sample 4, for which 30% CD PAM was more effective than 18% CD PAM when PAM concentration was 0.1-1 mg [L.sup.-1]. It is noteworthy that the flocculation series of sample 4 indicated that PAM molecular weight had little effect on clay flocculation.
Overall, clay flocculation by PAM is highly dynamic and strongly influenced by physicochemical properties of both PAM and soil clay, as well as PAM concentrations. The strong interactive effects between PAM and soil clays indicate that optimal PAM flocculation concentration varies with both PAM and clay properties. Therefore, which PAM to use at which concentration must be tested for each soil in question in order to achieve the maximum flocculation effect. The findings that anionic and non-ionic PAMs were effective in reducing turbidity in soil-water suspensions clearly demonstrate a greater potential for controlling turbidity of runoff water and clarifying farm ponds and reservoirs with PAMs in highly eroded regions in China.
Received 22 April 2014, accepted 23 July 2014, published online 10 October 2014
Aly SM, Letey J (1988) Polymer and water quality effects on flocculation of montmorillonite. Soil Science Society of America Journal 52, 1453-1458. doi: 10.2136/sssaj1988.03615995005200050047x
Barvenik FW (1994) Polyacrylamide characteristics related to soil applications. Soil Science 158. 235-243. doi: 10.1097/00010694199410000-00002
Ben-Hur M (2006) Using synthetic polymer as soil conditioners to control runoff and soil loss in arid and semi-arid regions--a review. Soil Research 44, 191-204. doi: 10.1071 /SR05175
Ben-Flur M, Letey J, Shainberg I (1990) Polymer effects on erosion under laboratory rainfall simulator conditions. Soil Science Society of America Journal 54, 1092-1095. doi: 10.2136/sssaj1990.036159950054 00040028x
Ben-Hur M, Malik M, Letey J, Mingelgrin U (1992) Adsorption of polymers on clays as affected by clay charge and structure, polymer properties, and water quality. Soil Science 153, 349-356. doi: 10.1097/00010694-199205000-00002
Bolto B, Gregory J (2007) Organic polyelectrolytes in water treatment. Water Research 41, 2301-2324. doi:10.1016/j.watres.2007.03.012
Goldberg S, Glaubig RA (1987) Effect of saturating cation, pH, and aluminum and iron oxide on the flocculation of kaolinite and montmorillonite. Clays and Clay Minerals 35, 220-227. doi: 10.1346/CCMN. 1987.0350308
Green SV, Stott DE, Norton LD, Graved JG (2000) Polyacrylamide molecular weight and charge effects on infiltration under simulated rainfall. Soil Science Society of America Journal 64, 1786-1791. doi: 10.2136/sssaj2000.6451786x
Gregory J (1989) Fundamentals of flocculation. Critical Reviews in Environmental Control 19. 185 230. doi: 10.1080/10643388909388365
Helalia AM, Letey J (1988) Polymer type and water quality effects on soil dispersion. Soil Science Society of America Journal 52, 243-246. doi: 10.2136/sssaj1988.03615995005200010042x
Laird DA (1997) Bonding between polyacrylamide and clay mineral surfaces. Soil Science 162, 826-832. doi:10.1097/00010694-199711 000-00006
Lei TW, Tang ZJ, Zhang QW, Zhao J (2003) Effects of polyacrylamide application on infiltration and soil erosion under simulated rainfalls: II erosion control. Acta Pedologica Sinica 40, 401-406. [In Chinese with English abstract]
Lentz RD, Sojka RE (1994) Field results using polyacrylamide to manage furrow erosion and infiltration. Soil Science 158, 274-282. doi: 10.1097/00010694-199410000-00007
Lentz RD, Sojka RE, Carter DL (1996) Furrow irrigation water quality effects on soil loss and infiltration. Soil Science Society of America Journal 60, 238-245. doi: 10.2136/sssaj1996.03615995006000010036x
Lentz RD, Sojka RE, Ross CW (2000) Polymer charge and molecular weight effects on treated irrigation furrow processes. International Journal of Sediment Research 15, 17-30.
Letey J (1994) Adsorption and desorption of polymers on soil. Soil Science 158, 244- 248. doi:10.1097/00010694-199410000-00003
Levy GJ, Agassi M (1995) Polymer molecular-weight and degree of drying effects on infiltration and erosion of three different soils. Australian Journal of Soil Research 33, 1007-1018. doi:10.1071/SR9951007
Levy GJ, Levin J, Gal M, Ben-Hur M, Shainberg I (1992) Polymers' effects on infiltration and soil erosion during consecutive simulated sprinkler irrigations. Soil Science Society of America Journal 56, 902-907. doi: 10.2136/sssaj1992.03615995005600030037x
Mamedov Al, Shainbcrg I, Wagner LE, Warrington DN, Levy GJ (2009) Infiltration and erosion in soils treated with dry PAM, of two molecular weights, and phosphogypsum. Soil Research 47, 788-795. doi:10.1071/SR09027
Mason LB, Amrhein C, Goodson CC, Matsumoto MR, Anderson MA (2005) Reducing sediment and phosphorus in tributary waters with alum and polyacrylamide. Journal of Environmental Quality 34, 1998-2004. doi: 10.2134/jeq2005.0086
Miller WP, Newman K.D, Frenkel H (1990) Flocculation concentration and sodium/calcium exchange of kaolinitic soil clays. Soil Science Society of America Journal 54, 346-351. doi:10.2136/sssaj1990.036159950054 00020008x
Misra DN (1996) Adsorption of polyacrylic acids and their sodium salts on hydroxyapatite: effect of relative molar mass. Journal of Colloid and Interface Science 181, 289-296. doi: 10.1006/jcis. 1996.0380
Orts WJ, Sojka RE, Glenn GM, Gross RA (1999) Preventing soil erosion with polymer additives. Polymer News 24, 406-113.
Peng FF, Di P (1994) Effect of multivalent salts-calcium and aluminum on the flocculation of kaolin suspension with anionic polyacrylamide. Journal of Colloid and Interface Science 164, 229-237. doi: 10.1006/ jcis. 1994.1161
Scybold CA (1994) Polyacrylamide review: soil conditioning and environmental fate. Communications in Soil Science and Plant Analysis 25, 2171 2185. doi: 10.1080/00103629409369180
Shainberg I, Warrington DN, Rengasamy P (1990) Water quality and PAM interactions in reducing surface sealing. Soil Science 149, 301-307. doi: 10.1097/00010694-199005000-00007
Sojka RE, Bjomeberg DL, Entry JA, Lentz RD, Orts WJ (2007) Polyacrylamide in agriculture and environmental land management. Advances in Agronomy 92, 75-162. doi: 10.1016/S0065-2113(04) 92002-0
Tang ZJ, Lei TW, Zhang QW, Zhao J (2003) Effects of polyacrylamide application on infiltration and soil erosion under simulated rainfalls: 1 infiltration. Acta Pedologica Sinica 40, 178 185. [In Chinese with English abstract]
Theng BKG (1982) Clay polymer interactions: summary and perspectives. Clays and Clay Minerals 30, 1-10. doi:10.1346/CCMN.1982.0300101
Vacher CA, Lock RJ, Raine SR (2003) Effect of polyacrylamide additions on infiltration and erosion of disturbed lands. Australian Journal of Soil Research 41, 1509 1520. doi: 10.1071/SR02114
Wang AP, Li FH, Yang SM (2011) Effect of polyacrylamide application on runoff, erosion, and soil nutrient loss under simulated rainfall. Pedosphere 21, 628-638. doi: 10.1016/S1002-0160(11)60165-3
Xia WS, Lei TW, Liu JG (2002) Development and review of research of preventing soil erosion with polyacrylamide (PAM). Chinese Journal of Soil Science 43, 78-80. [In Chinese with English abstract]
Yan XQ, Zhang XJ (2008) Physical and chemical properties of polyacrylamide and its application to soil amelioration. Agricultural Research in Arid Area 26, 189-192. [In Chinese with English abstract]
Xiaoqian Yan (A, C) and Xunjiang Zhang (B)
(A) Department of Chemical Engineering, Shaanxi Institute of Technology, Xi'an, Shaanxi, China.
(B) Shaanxi Chemical-Engineering Construction Co., Yangling, Shaanxi, China.
(C) Corresponding author. Email: email@example.com
Table 1. Charge density, type, and molecular weights of selected polyacrylamide used in this study Polyacrylamide Charge Charge Molecular weight density type (g [mol.sup.-1]) (%) Magnifloc 494C 20 Positive 6 x [10.sup.6] Superfloc N300 0 Neutral 15 x [10.sup.6] Superfloe A110 18 Negative 15 x [10.sup.6] Superfloc A100 7 Negative 16 x [10.sup.6] Superfloc 1606 30 Negative 28 x [10.sup.6] Soilfix Polybead 30 Negative 16 x [10.sup.6] Chemtall 923VHM 20 Negative 14-17.5 x [10.sup.6] Table 2. Selected soil physicochemical properties of four subsoils Soil Soil Clay Sand Silt pH Organic sample texture (%) carbon (%) Sample 1 Sand 4.6 90.0 5.4 8.1 0.15 Sample 2 Sand 3.8 93.2 3.0 7.4 0.2 Sample 3 Clay 50.0 28.5 21.5 5.1 0.35 Sample 4 Clay loam 32.0 37.5 30.5 5.0 0.16 Soil Exch. cations sample ([cmol.sub.c] [kg.sup.-1]) [Ca.sup.2+] [Mg.sup.2+] Sample 1 4.3 0.2 Sample 2 9.0 0.2 Sample 3 0.9 0.4 Sample 4 0.1 0.1 Table 3. Clay mineral composition and relative percentage in both fine (<1 [micro]m) and coarse (1-2 [micro]m) clay fractions of four subsoils Soil Montmorillonite Kaolinite sample <1 [micro]m 1-2 [micro]m <1 [micro]m 1-2 [micro]m Sample 1 64 13 11 42 Sample 2 90 60 0 11 Sample 3 8 0 80 88 Sample 4 0 0 90 95 Soil Vermiculite Mica sample <1 [micro]m 1-2 [micro]m <1 [micro]m 1-2 [micro]m Sample 1 8 15 17 30 Sample 2 0 3 10 26 Sample 3 11 11 0 1 Sample 4 8 3 2 2
|Printer friendly Cite/link Email Feedback|
|Author:||Yan, Xiaoqian; Zhang, Xunjiang|
|Date:||Oct 1, 2014|
|Previous Article:||How much soil organic carbon sequestration is due to conservation agriculture reducing soil erosion?|
|Next Article:||Opportunities and constraints for biochar technology in Australian agriculture: looking beyond carbon sequestration.|