Cationic Starches as Substitute for Synthetic Cationic Flocculants in

J. Phys. Chem. B 2007, 111, 8641-8648

8641

Cationic Starches as Substitute for Synthetic Cationic Flocculants in Solid-Liquid Separation of Harbor Sludge† M. Shirzad-Semsar, S. Scholz, and W.-M. Kulicke* Institute for Technical and Macromolecular Chemistry, UniVersity of Hamburg Bundesstr. 45, D-20146 Hamburg, Germany ReceiVed: January 12, 2007; In Final Form: April 3, 2007

Harbor sludge (about 25% total solid) has to be dredged to keep the waterways free. Thus, annually 1.2 million m3 of dredged material has to be cleaned. For this process, three different synthetic flocculants with optimal molar masses, ionogenities, and concentrations are added in order to get a good dewatering efficiency and shear strength of the flocs. But as synthetic flocculants bring about unwanted fish toxicity and insufficient biodegradability, this study intends to check whether these flocculants can successively be substituted by cationic starches which have already been proven to be less toxic than synthetic flocculants. Five different starch derivatives with an average degree of substitution higher than 0.5 were characterized, especially in terms of the molar mass and coil size distribution, and flocculation tests, zeta potential measurements and filtrate turbidity tests were carried out in order to create optimum flocculation conditions. The flocculation and dewatering measurements clearly show that the synthetic cationic flocculant PA (0.2 kg/tTS) can be best substituted by cationic starch KS 2 (c ) 0.1 kg/tTS, Mw ) 1.1 e+08). For substitution of PTAC (c ) 0.3 kg/tTS) by cationic starches, we observed that a maximal dewatering efficiency is reached with an approximately 3-fold dose of KS 1 (1 kg/tTS, Mw ) 8.1 e+07).

1. Introduction Wastewater purification has been a problem not only with sewage sludge, but also in the paper, the cosmetic, and the pharmaceutical industries for a long time. Here it is well-known that the particles in the nanometer and micrometer range settle only slowly and the rate of sedimentation depends on particle geometry and on the flow behavior in the sedimentation tank and additionally becomes more complicated by equally charged particles. As is generally known, in modern wastewater treatment, synthetic flocculants are used to obtain clarity of the filtrate.1,2 A special problem becomes obvious in harbors, where large quantities of dredged material must be coped with. The dredged material is flocculated and incorporated in the seas worldwide. At present in Hamburg harbor an amount of 1.2 million m3 of dredged material is treated annually using additional process water. Since here the flocculated harbor sludge is deposited on silt disposal sites, the flocs need a high shear stability (>30 kN/m2). The ecological accommodation of the flocculated dewatered harbor sludge in mud hills amounts to another 10 years. For the cleaning of the resulting wastewater and the safe disposal of the classified harbor sediments of the Hamburg harbor, the plant METHA (MEchanical Treatment and Dewatering of Harbour Sediments) was brought on line in 1993. For a good dewatering process, sufficient shear strength and time for maturation of the flocs, four synthetic flocculants with different molar masses and in different concentrations were used (double dual procedure).3 Improvements have led to a modified flocculation concept.4 Now three different synthetic flocculants † Part of the special issue “International Symposium on Polyelectrolytes (2006)”. * Corresponding author. E-mail: [email protected].

are used during the flocculation process. Table 1 shows the different synthetic flocculants used in the plant METHA. The total amount of added polyelectrolytes in METHA is about ctotal ) 1.5 kg/tTS. Commercial forms of synthetic flocculants5,6 may contain toxic monomers from the synthesis and additives. In addition, they also possess a bad biodegradability.7-9 In Germany, the disposal of flocculated sludge with polyacrylamide derivatives on agrarian surfaces has been limited and will be strictly prohibited at the end of 2013.10 Furthermore, new alternatives to synthetic polymers must be developed due to rising oil prices. Flocculants on the basis of renewable primary products, especially cationic starches besides chitosans,11 are alternatives to synthetic flocculants. They are already used in the treatment of both waste and drinking water.12 It has been shown by the so-called HEST-test (hen’s fertile egg screening test) that cationic starch derivatives with a degree of substitution (DS) up to 0.6 are less toxic than the usual synthetic flocculants like poly(acrylamide-co-N,N,N,-trimetyl-ammonium-ethylacrylate)chloride (PTAC) and poly(diallyl-dimethyl-ammonium)-chloride (PolyDADMAC).13 The aim is now to successively substitute synthetic cationic flocculants (see Table 1) by cationic starches and to obtain at least the same dewatering efficiency and flocs of a high shear stability. The high shear stability is necessary because of the flocs-disposal on dumps where bulldozers are used. It has been shown that besides the degree of functionalization, especially the molar mass and the coil size distribution have to be taken into account for optimal flocculation conditions.14,15 In order to find a tailor-made solution for the flocculation of harbor sludge, different starch derivatives were characterized using 13C NMR spectroscopy, polyelectrolyte titration, viscometry, and asymmetrical flow field-flow fractionation (aFFFF) coupled with a multi-angle laser light scattering (MALLS) and a refractive

10.1021/jp0702705 CCC: $37.00 © 2007 American Chemical Society Published on Web 06/07/2007

8642 J. Phys. Chem. B, Vol. 111, No. 29, 2007

Shirzad-Semsar et al.

TABLE 1: Synthetic Flocculants Used in the Plant METHA flocculant

charge (%)

Mw (g/mol)

c(kg/t)

1. cationic 2. anionic 3. cationic

100 41 51

6.4 × 104 3.2 × 106 6.1 × 106

0.03 0.35 0.9-1.2

TABLE 2: Overview of Parameters for the Harbor Sludge Used total solid (TS) (%)

ignition lost (%)

pH value

zeta potential ζ (mV)

average diameter (µm)

5-10

10-15

7.2

-17

20-60

TABLE 3: Manufacturing Data sample

note

KS 1 KS 2 KS 3

liquid, conserved with mergal V 698 K 1 liquid, conserved with mergal V 698 K 1 solid, content of amylose approximately 50% solid, crosslinked with epichlorohydrin solid, crosslinked with epichlorohydrin

KS 4 KS 5

degree of substitution (DS) approximately 1 approximately 0.7 approximately 0.7 approximately 0.7 approximately 0.7

index (RI) detector, in order to establish structure property relationships. In addition, flocculation tests were carried out, using a patented flocculation and dewatering apparatus (FDA).16 For the investigation of the electrical stabilization of harbor sediments, zeta potential measurements and turbidity tests of the filtrates were carried out. 2. Materials and Methods 2.1. Materials. 2.1.1. Harbor Sludge. The flocculation tests were conducted with separated fine harbor sludge (i.e., without flocculants) of the Hamburg plant METHA. Table 2 gives an overview of the parameters for that sludge. 2.1.2. Cationic Starches. Five cationic starches (KS 1-KS 5) on the basis of potato starch have been investigated. KS 1 and KS 2 are noncross-linked cationic starch solutions. KS 3 is a noncross-linked cationic starch with a 50% amylose content. KS 4 and KS 5 are cationic starches which are cross-linked with epichlorohydrin. The starch solutions (KS 1, KS 2) are stabilized with 0.2% mergal V 698 K 1. All samples were supplied by Emslandsta¨rke GmbH (Emlichheim, Germany). 2.1.3. Synthetic Flocculants. The synthetic polyanion and polycation based on polyacrylamide used at METHA were employed as components of the dual system, namely poly(acrylamide-co-Na-acrylate) (PAAm-AA) as polyanion and a polyamine made of (chloromethyl) oxirane and N- methylmethanamine (PA) and poly(acrylamide-co-N,N,N,-trimethylammonium-ethylacrylate) chloride (PTAC) as polycation (see Table 1). 2.2. Methods. 2.2.1. Sample Preparation. An exact characterization of the starch derivatives KS 1 and KS 2 is not possible, since the pure solid content is not exactly known. Since it could not be excluded that the stabilizer affects the viscometry and the light scattering measurements, the starch derivatives were precipitated with ethanol from the solution and afterward dried in a drying chamber. Table 3 gives an overview of the manufacture data. The insoluble parts were determined gravimetrically by filtration and centrifugation to determine the exact polymer concentration in solution. In all cases, 0.4 wt % solution was prepared by adding the cationic starch samples. The cationic starch was diluted with tap water, and was conditioned under

stirring with a so-called dissolver-stirrer at a stirring speed of 500 rpm for 24 h at room temperature. 2.2.2. 13C NMR Spectroscopy. The 13C NMR spectra (IGATED) were recorded on a Bruker MSL 400 spectrometer (Bruker, Karlsruhe, Germany) using a 10 mm 13C-1H-dualsample-head with 2H-Lock and a control computer (type Avance 400, Bruker) at a measuring frequency of 100 MHz and a temperature of 80 °C. 2.2.3. Ultrasonic Degradation. The starch ethers have been degraded by ultrasonic degradation to improve the signal-tonoise ratio in 13C NMR spectroscopy. A sonifier W-450 ultrasonic degradation device (Branson Schallkraft GmbH, Heusenstamm, Germany) with 3/ 4′′ titanium resonator was used for the polymer degradation. The sound frequency of the device was 20 kHz, the maximum output was 400 W, and the density of the ultrasonic output was approximately 80 W/cm2. After the degradation process, the sample solution was centrifuged for 1 h at 13 000 rpm to remove metal swarfs from the ultrasonic resonator. Afterward the solutions were lyophilized (Betta 1-16, Osterode, Germany), and the sample solutions for NMR were prepared. 2.2.4. Polyelectrolyte Titration with Particle Charge Detector. The charge quantity of the cationic starches was determined with a particle charge detector PCD 03 pH (Mu¨tek analytics GmbH, Herrsching, Germany). The charge detector was coupled with a DL21 automatic titrator (Mettler-Toledo GmbH, Giessen, Germany). The cationic starches (0.05 wt %) were titrated against a 0.001 N PES-Na (polyethylene sulfonate) solution. 2.2.5. Viscometry. The intrinsic viscosity [η] of the samples was determined with an Ubbelohde capillary viscometer (type Ic, Schott, Hofheim, Germany) at 25 °C ( 0.1 °C. The measuring data were evaluated according to the Huggins equation:

ηsp ) [η] + kH[η]2c + ... c

(eq 1)

From this dependency the intrinsic viscosity [η] and the Huggins constant kH, which is a measure for the solvent quality, were determined. Furthermore, the critical concentration c*[η], the concentration at which the solution volume is completely filled with polymer coils, was evaluated. It is defined according to the following equation:

c*[η] )

mpolymer 2.5 ) Vsolution [η]

(eq 2)

The experiments were conducted within the range of ηr ) 1.2 and 2.5. The solvent used for these measurements was a 0.1 M NaNO3 solution with 3 mM NaN3. 2.2.6. Determination of the Molar Mass and Size Distribution by Flow Field Flow Fractionation Hyphenated to a Multi-Angle Laser Light Scattering Detector and a RefractiVe Index Detector. In this work we used an asymmetrical channel with a height of 250 µm (Wyatt Technology Corp., Santa Barbara, CA). The channel was made of a stainless steel bottom and a transparent acrylic glass cover. There is a semipermeable membrane between cover, bottom and spacer made of regenerated cellulose with a cutoff of 10 kDa (Wyatt). The channel flow and the crossflow were established and regulated by an Eclipse F (Wyatt). The eluent used for these measurements was an aqueous solution with 0.1 M NaNO3 and 3 mM NaN3. For the determination of the molar mass and size distribution of every fraction, the channel was hyphenated to a DAWN EOS light-

Solid-Liquid Separation of Harbor Sludge

J. Phys. Chem. B, Vol. 111, No. 29, 2007 8643

scattering photometer (Wyatt) and an interferometric refractometer Optilab DSP (Wyatt). The values for Kc/Rθ are plotted in a one-dimensional ZimmPlot against sin2 θ according to the following equation:

KCRΘ )

16π2n02 2 1 + R sin 2(Θ/ 2) Mw 3Mwλo G

(eq 3)

Extrapolation to the angle θ ) 0 yields the reciprocal of the molar mass, Mw, as the intercept of the ordinates and RG from the slope for each fraction. In this way the molar mass and the radius of gyration can be determined for every sample fraction. 2.2.7. Pressure Filtration. The flocculation efficiency of flocculants was determined by means of a patented flocculation and dewatering apparatus (FDA).16 With this apparatus, sludges were flocculated by addition of flocculants and dewatered by means of compressed air (2 bar, 180 s). By integration of filtrate/ time curves, the dimensionless dewatering index (IE) is obtained,17 IE being a measure for the effectiveness of the flocculants. Flocculant’s conditioning was done by stirring at 500 rpm for 1 min. 2.2.8. Turbidity Test. The turbidity of the filtrates was determined with a turbidity measuring instrument 2100AN (Hach, Loveland, USA). The turbidity values of the filtrates are indicated in nephelometric turbidity units (NTU). 2.2.9. Laser Doppler Anemometry (Zeta Potential Determination). The Zeta potential determinations of mud particles were carried out with a ZetaSizer 3000 of the company Malvern of Instruments Ltd. (Malvern, UK). The measurements were taken at 25 °C and a field intensity of 29.5 Vcm-1 for each sample, using the software PC for Windows (version 1.32, Malvern of Instruments Ltd.). 8-10 values were recorded and afterward the arithmetic average was calculated. 2.2.10. Thermal GraVimetry. The dry content of the cationic starches was determined by thermal gravimetry. A pan containing 20 mg of sample is put into the oven and is heated at 5 K/min. The measurement stops at approximately 660 °C. As the water evaporates during the heating, the balance arm moves and the weight difference between the pan with the sample and the counter weight is determined. After the sample humidity has evaporated, the heating process proceeds. Thus, not only the humidity content, but also a possible disintegration process can be determined. Although the diagram shows a nitrogen valve, the system was not purged with nitrogen, because the samples are not sensitive to air: (1) balance arm (2) pan (3) oven (heating rate: 5 K/min) (4) temperature of sample (5) counter weight (6) flowmeter for cooling water (7) flowmeter for nitrogen (8) lifting mechanism for the oven 3. Results and Discussion 3.1. Characterization of Cationic Starches. 3.1.1. Determination of Solid and Salt Content. First the solid content and the salt content of the five starch derivatives were determined. The solid content (moisture content) was determined by means of thermal gravimetry. In Figure 2, thermograms for all samples are shown. The thermograms show a decrease of mass in dependence of the temperature. For the evaluation of the solid content, the curve indicates the decrease of mass. The difference between the

Figure 1. Schematic setup of a thermal gravimetry.

Figure 2. Thermogram of the cationic starches.

TABLE 4: Analytical and Preparation Data of Starch Derivativesa sample

TS (%)

DSN

salt content Cl- (%)

nonsoluble part (%)

sample preparation

KS 1 KS 2 KS 3 KS 4

28.2b 24.4c 100 98.9

0.82 0.61 0.65 0.64

17.6 13.3 17.1 17.4

0.2 0.1 0.1 70.4

KS 5

99.2

0.65

17.3

62.2

11 µm, centr.c 11 µm, centr.2) 11 µm, centr.c 60 µm, 20 µm, 11 µm, centr.c 60 µm, 20 µm, 11 µm, centr.c

a centr. ) centrifugated. TS ) total solid. DSN ) degree of substitution, determined by elemental analysis. b Data for solution c 45 min at 13.000 rpm

baseline at 100% mass and the first inflection point of the curve at 123 °C (using the example of KS 2) express the moisture loss. The ongoing decrease of mass shows decomposition processes. For the sample KS 2, a solid content of 24.4% was computed. In Table 4 the solid content, salt content, DS, nonsoluble parts, and the sample preparation with centrifuged parts are listed. The nonsoluble parts were filtrated by a 11 µm nylon membrane filter (Whatman). As salt content, the content of chloride was determined argentometrically according to the Scho¨ninger decomposition.18 All samples show a high salt content of more than 17%, except for KS 2 which has a chloride content of 13.3%. Also the nitrogen content was determined by elemental analysis for the determination of the DS. 3.1.2. 13C NMR Spectroscopy. The structure of the five flocculants on the basis of potato starch was recorded by 13C NMR spectroscopy. Figure 3 shows a 13C NMR spectrum of KS 1. The signals of the unmodified cationic starch KS 1 appear for C1 at 100.6 ppm and for C2-C6 between 61.9 and 81.4 ppm.



Figure 3. 13C NMR spectrum of KS 1 measured in D2O/TSP at 80 °C, IGATED. (Bruker MSL 400), (R ) H or cationic group).

Figure 4. Concentration ratio of the specific viscosity to concentration for KS 1, KS 2, and KS 3. The dotted lines indicate the measuring region between ηrel ) 1.2 and 2.5.

TABLE 5: Charge Contents in eq/g Determined by Titration against Sodium Polyethylene Sulfonate

TABLE 6: Intrinsic Viscosity, kH, and Critical Concentration of the Cationic Starches

cationic starches

charge content (eq/g)

KS 1 KS 2 KS 3 KS 4 KS 5

0.00091 0.00100 0.00106 0.00029 0.00032

Introduction of 2-hydroxypropyltrimethylammoniumchloride elicits three signals at 97.5 ppm (C1s, signal for a C1 atom in the neighborhood of a modified carbon atom in position 2), 81.4 ppm (C2s, signal for a modified C2-Atom) and 69.5 ppm (C6s, signal for a modified C6 atom). The signals of the 2-hydroxypropyltrimethylammoniumchloride group appear between 70 and 75 ppm (C8), at 69.5 ppm (C7), at 66.3 ppm (C9) and for C10 an intensive signal is shown at 55.6 ppm. In addition to this, an attempt was made to determine the DS besides the elemental analysis also by means of 13C NMR spectroscopy. The 13C-spectra of the cationic starches show a strong bandwidth. As a result, the determination of the area below the signal peak was prone to error and the DS were incorrect. 3.1.3. Polyelectrolyte Titration for the Determination of Charge Content. The charge quantities of five flocculants on the basis of potato starch were examined by polyelectrolyte titration (Table 5). The determined charge contents were in the range of 0.000 29-0.001 06 eq/ g. The charge contents do not correlate with the DS. Using polyelectrolyte titration, predominantly the surface charges are obtained. In addition, short chain polymers can enter into the pore system so that the results are to be interpreted carefully.19 KS 1 should have exhibited the highest charge content, since it possesses the highest DS among all samples. Also KS 2 to KS 5 are expected to show comparable charge contents based on the fact that they have approximately the same DS. The low charge contents of KS 4 and KS 5 are attributed to their synthesis but probably also to the low solubility of these crosslinked cationic starches. The nonsoluble parts were separated by a 11 µm nylon membrane filter (Whatman).We will try to improve the solubility of these samples. 3.1.4. Viscometry20. One of the most widely used and easiest methods of finding the volume demand of polymers in solution is viscometry. The starch derivatives KS 1, KS 2, and KS 3

sample

[η] (mL/g)

c*[η] (%)

kH

KS 1 KS 2 KS 3

189 245 96

1.32 1.02 2.60

0.995 0.966 0.637

were investigated, because the amount of insoluble components in KS 4 and KS 5 is too high. In Figure 4, the concentration ratio of the specific viscosity to concentration, which was evaluated according to the Huggins equation, is displayed. Figure 4 and Table 6 show that the intrinsic viscosity and therefore the volume demand, which you can obtain from the axis intercept, decreases from KS 2 over KS 1 to KS 3. KS 1 and KS 2 show nearly the same kH of 0.995 and 0.966, respectively. KS 3 shows the lowest kH value in this series (0.637). KH is a measure for the solvent quality. If the solvent quality decreases, the coefficient kH increases until the value of 0.5 for a θ-solvent is achieved. Values above 0.5 indicate a contraction of the macromolecules. Values above 0.5 are reached for KS 1 and KS 2 and also for KS 3 indicating that the polymer coils are contracted. Another explanation for these high kH values is that KS 1, KS 2 and also KS 3 have a compact, highly branched structure.21 It is noteworthy that KS 3 has about 50% amylose, as a result of this, the kH value is lower. 3.1.5. Determination of the Molar Mass and Size Distribution by Asymmetrical Flow Field Flow Fractionation Hyphenated to a MALLS and RI Detector. Multi-angle laser light scattering (MALLS) is the method of choice, if you want to determine the molar mass and the radius of gyration absolutely. If additional knowledge about the molar mass/coil size distributions is required, a fractionation unit is hyphenated to the MALLS and RI detector. Two methods are widely used for the fractionation of polymers: size-exclusion chromatography and flow field-flow fractionation (which is available in the symmetrical and asymmetrical version). In this work, we decided to use asymmetrical flow field-flow fractionation (aFFFF) which has been coupled to the MALLS and RI detector. The refractive index increment used for the evaluation of the light scattering data is 0.147.22 It was determined after dialysis of the cationic starch samples, because it is necessary to determine dn/dc under



Figure 5. Elution diagrams of KS 1 (upper left), KS 3 (upper right), and KS 2 (lower left). Only the molar mass distributions are displayed.

TABLE 7: Weight-Averaged Molar Mass (Mw), Radius of Gyration (RG), and Recovery Rate (RR) of the Cationic Starches sample

Mw (g/mol)

RG (nm)

RR (%)

KS 1 KS 2 KS 3

8.1 × 107 1.1 × 108 3.4 × 107

194 213 135

70 100 60

conditions of a constant chemical potential.23 The results for KS 1, KS 2, and KS 3 obtained by this method are displayed in Table 7. Table 7 shows the same tendencies as the viscosity data from Table 6. The radius of gyration increases from KS 3 above to KS 1 to KS 2. The same holds true for the tendency concerning the weight averaged molar mass. All elution diagrams (Figure 5), obtained by aFFFF hyphenated to a MALLS and RI detector, show a monomodal size distribution for the cationic starches. The sample peak of KS 1 (Figure 5, upper left) starts at approximately 4 mL and ends at 30 mL. The specimen contains molecules between 3 × 107 and 2 × 108 g/mol. The sample peak of KS 1 shows a steep left flank and a strong tailing. The sample peak of KS 2 (Figure 5, lower left) starts at approximately 10 mL and ends at an elution volume of 50 mL. KS 2 contains molar masses between 4 × 107 and 2 × 108 g/mol. The MALLS signal shows a broad elution peak where two maxima and thus two fractions might be anticipated. The sample peak of KS 3 (Figure 5, upper right) starts at approximately 6 mL and ends at 20 mL. The sample contains polymers of molar masses between 1 × 107 and 1 × 108 g/mol. The sample peak shows, as in the case of KS 1, an asymmetrical distribution with a steep left flank and a strong tailing. In contrast to the elution diagrams of, e.g., bovine serum albumine (BSA), polystyrene lattices and pullulane which

Figure 6. Diagram of the flocculation and substitution procedure.

showed a quite smooth elution signal, in Figure 5 the elution diagrams of the cationic starches are not smooth but jagged.24-26 We have been able to exclude that these jagged signals are due to detector noises. The jagged signals in all elution diagrams of the cationic starches are potentially due to adsorption phenomena on the semipermeable membrane. To improve the characterization of the cationic starches, further research work is being carried out currently in our lab concerning the optimization of the semipermeable membrane and the sample preparation. 3.2. Flocculation Experiments. 3.2.1. Flocculation Experiments with Cationic Starches as Substitute for the Synthetic Cationic Flocculants PA and PTAC. During the flocculation process, mechanical loads are exercised on the flocs, which have an effect on the floc size. Therefore, investigations of the dewatering efficiency are more exact with pressure filtration processes (FDA) in contrast to sedimentation processes (imhoff cone). Figure 6 shows the diagram of the flocculation and substitution procedure and explains at which place we would like to substitute synthetic flocculants by cationic starches. With cationic starches of different concentrations and 0.07 kg/tTS of the anionic poly(acrylamide-co-Na-acrylate) (PAAmAA), dual flocculation investigations with harbor sludge were carried out. The thus thickened harbor sludge was flocculated


Figure 7. Dual flocculation investigation with harbor sludge; comparison of the dewatering indexes for cationic starches (KS 1-KS 5) and synthetic cationic PA with different concentrations.


Figure 8. Dual flocculation investigation with pre-thickened harbor sludge; comparison of the dewatering indexes of cationic starches (KS 1-KS 5) and synthetic cationic PTAC with different concentrations.

TABLE 8: Used Concentration of Cationic Starches Reaching Maximum Dewatering Index as Substitute for PA cationic starches

c (kg/tTS)


0.2 0.1 0.3 0.4 0.8

with 0.3 kg/tTS poly(acrylamide-co-N,N,N,-trimethylammoniumethylacrylate) chloride (PTAC) to find the maximum dewatering index IE (Figure 7). Of all investigated cationic starches, KS 2 showed the highest dewatering index IE at 0.1 kg/tTS. In this case, the synthetic cationic flocculant, polyacrylamine (PA), should be substituted by cationic starches (see Figure 6). An increase of the cationic starch concentration leads to an increase of the dewatering index IE. KS 2 showed the highest dewatering index IE at 0.1 kg/tTS, KS 1 at 0.2 kg/tTS, KS 3 at 0.3 kg/tTS, KS 4 at 0.4 kg/tTS, and KS 5 at 0.8 kg/tTS. From these results it may be concluded that polymers with high molar mass and radius of gyration (see Table 7) lead to a positive dewatering efficiency. The dewatering efficiency decreases with a decrease of molar mass and coil size. Compared with PA, 50% less flocculant would be required using KS 2. Table 8 shows the concentration of cationic starches (KS 1-KS 5) with which the maximum dewatering index IE is reached when substituting PA. The nonsoluble parts from KS 4 and KS 5 were not separated because the separation of nonsoluble particles is not possible at METHA plant. With 0.2 kg/tTS of the cationic polyamine (PA) and 0.07 kg/tTS of the anionic poly(acrylamide-co-Na-acrylate) (PAAmAA), the harbor sludge was flocculated. The thus thickened harbor sludge was flocculated with cationic starches of different concentrations to find the maximum dewatering index IE (Figure 8). In this case, the synthetic cationic flocculant poly(acrylamideco-N,N,N,-trimethylammonium-ethylacrylate) chloride (PTAC), should be substituted by cationic starches (see Figure 6). KS 1 showed the highest dewatering index IE at 1 kg/tTS. Compared with KS 1, 70% less flocculant would be required using PTAC. Table 9 gives the concentration of cationic starches (KS 1-KS 5) with which the maximum dewatering index IE is reached when substituting PTAC. The nonsoluble parts from KS 4 and KS 5 were not separated here either, since a separation of nonsoluble particles is not possible at METHA plant. 3.2.2. Turbidity Filtrate InVestigation with Starch DeriVatiVes Substituting Synthetic Cationic Flocculants PA and PTAC. One

Figure 9. Turbidity of filtrate as a function of flocculant concentration.

TABLE 9: Used Concentration of Cationic Starches Reaching Maximum Dewatering Index as Substitute for PTAC cationic starches

c (kg/tTS)


1.0 1.2 1.8 2 3

aim of the flocculation procedure is to obtain a clear filtrate. Therefore, the influence of the cationic starches on the filtrate cleaning was investigated after flocculation by turbidity measurements. Figure 9 shows the filtrate turbidity investigations of these flocculants. The turbidity curves show that the maximum filtrate cleaning is comparable with the cationic starches as well as with PA. To compare the influence of the five cationic starches by dual flocculation as substitute for PTAC on the filtrate cleaning, we investigated the filtrate after flocculation by turbidity measurements. Figure 10 shows the filtrate turbidity after flocculation. The turbidity curves show that the maximum filtrate cleaning is comparable for all cationic starches. However, an increase of turbidity is obtained using PTAC above 1.5 kg/tTS due to electrostatic restabilization of harbor sediments.14 3.2.3. Adsorption BehaVior of Cationic Starch DeriVatiVes on Harbor Sludge (Zeta Potential Measurements). Zeta potential measurements were carried out on harbor sludge after flocculation with cationic starches of different concentrations and 0.07 kg/tTS of the anionic poly(acrylamide-co-Na-acrylate) (PAAm-


Figure 10. Turbidity of filtrate as a function of flocculant concentration.


Figure 12. Zeta potentials (ζ) of five cationic starch derivatives compared with PTAC.

The maximum dewatering index for the cationic starches, however, is reached at high zeta potential values (see Table 9). At the isoelectrical point, the maximum dewatering index will not be reached. According to the literature,27-29 if the maximum dewatering index is obtained under the isoelectrical point, the flocculation mechanism is a patch mechanism. The observations in this study confirm the bridging mechanism. However, because of differences in the size of the sediments (20-60 µm) both patch and bridging mechanism may occur.

Figure 11. Zeta potentials (ζ) of five cationic starches compared with PA.

AA) and 0.3 kg/tTS polycation (PTAC) to find out the adsorption behavior with dual flocculation. Figure 11 presents zeta potentials (ζ) of the flocculated harbor sludge with the five different cationic starches and PA. The maximum dewatering index IE can first be reached with KS 2 at 0.1 kg/tTS, whereas the zeta potentials (ζ) are over 20 mV. The curves show that a maximum dewatering index IE for all other cationic starches and even for PA will be achieved at high zeta potential and higher flocculant concentration (see Table 8). The plateau of zeta potential of PA is probably lower because of its chemical structure and molar mass. Although the curves of KS 4 and KS 5 show similar characteristics in zeta potential measurements, those differences cannot be correlated to the results from Table 8, because of the differences in chemical structure and molar mass and because by zeta potential measurements only the adsorption behavior is obtained. A further task is to substitute PTAC with cationic starches. Therefore, further zeta potential measurements were carried out on harbor sludge after flocculation with 0.2 kg/tTS of the cationic polyamine (PA) and 0.07 kg/tTS of the anionic poly(acrylamide-co-Na-acrylate) (PAAm-AA) with different concentrations of cationic starches and PTAC. Figure 12 presents zeta potentials (ζ) of the flocculated harbor sludge with the five different cationic starches and PTAC. The maximum dewatering index IE can first be reached with PTAC at 0.3 kg/tTS. The zeta potential (ζ) curve shows that the maximum dewatering index IE for PTAC is reached at the isoelectrical point (ζ ) 0 mVs). At the isoelectrical point of the zeta potentials, the areas with cationic excess charge compensate the negative sectors of the particle surface. Therefore, a maximum flocculation efficiency is to be expected at the isoelectrical point for flocculation processes which work according to a patch mechanism.

4. Conclusion In order to check whether synthetic cationic flocculants, which are used in the industrial plant METHA (MEchanical Treatment and Dewatering of HArbour-sediments), can successively be substituted by cationic starches, five different cationic starches were characterized and flocculation investigations, turbidity investigations as well as zeta potential measurements were conducted. The flocculation and dewatering measurements have convincingly shown that the synthetic cationic flocculant PA can be substituted by cationic starches. Thus, KS 2, the cationic starch with the highest molar mass and radius of gyration, is best in the series. With a decrease of molar mass and radius of gyration, the flocculation efficiency decreases. The filtrate investigation with cationic starches substituting the synthetic cationic flocculant PA shows that the maximum filtrate cleaning is comparable both with all cationic starches and with PA. Concerning the substitution of PTAC by cationic starches, we have been able to show that a maximum dewatering efficiency is reached with an approximately threefold dose (KS 1). The filtrate investigation with cationic starches substituting the synthetic cationic flocculant PTAC shows that the maximum filtrate cleaning is comparable with all cationic starches. Nevertheless, we observe an increase of turbidity for PTAC due to electrostatic restabilization of harbor sediments. Zeta potential measurements show that the maximum dewatering efficiency is reached at a high zeta potential. Since the average sediment sizes are between 20 and 60 µm, both patch and bridging mechanism may occur. Acknowledgment. This work was supported by budgetary funds of the Federal Ministry for Consumer Protection, Fachagentur fu¨r nachwachsende Rohstoffe FNR (No. 22018605). We thank our partners CUTEC-Institut GmbH, Hamburg Port Authority and Emslandsta¨rke GmbH for the positive and constructive co-operation. References and Notes (1) Schuster, C.; Ko¨tz, J.; Kulicke, W-M.; Parker, S; Bo¨hm, N; Jaeger, W. Acta Hydrochim. Hydrobiol. 1997, 25, 27-33.

8648 J. Phys. Chem. B, Vol. 111, No. 29, 2007 (2) Homeyer, A; Krentz, O; Kulicke, W-M. Colloid Polym. Sci. 1999, 277, 637-645. (3) Kulicke, W. M.;Lenk, S.;Detzner, H. D.; Weiss, T. Chem. Eng. Technol. 1993, 5, 541-552. (4) Bode W.; Do¨ring U.; Schmidt-Bonde L.; Schramm W. HANSASchiffahrt-Schiffbau -Hafen 1997, 134 (7), 64-67. (5) Burkert, H.; Srena, C. Ullmann’s Enzykl. Tech. Chem., 4. Aufl., Bd. 11; Verlag Chemie: Weinheim, 1976, 1476. (6) Kulicke, W.-M.; Jacobs, A. Makromol. Chem., Macromol. Symp. 1992, 61, 75. (7) Letterman, R. D.; Pero, R. W. Contaminants in polyelecrolytes used in water treatment. J. Am. Water Works Assoc. 1990, 82, 87-97. (8) Aizawa, T.; Magara, Y.; Musashi, M. Problems with introducing synthetic polyelectrolyte coagulants into the water purification process. Water Supply 1991, 9, 27-35. (9) Heinzmann, B. Einsatz In Organischer Polymere in der Abwasserbehandlung, “Abwasserfa¨llung und Flockung”, Tagungsband der 12. Karlsruher Flockungstage; Hahn, H. H., Wilhelmi, H., Eds.; Oldenbourg Verlag: Munich, 1998; pp 69-94. (10) Verordnung u¨ber das In Verkehr bringen Von Du¨ngemitteln, Bodenhilfsstoffen, Kultursubstraten und Pflanzenhilfsmitteln (Du¨ngemittelVerordnung - Du¨MV) Vom 26. NoVember 2003; Bundesgesetzblatt Jahrgang 2003 Teil I Nr. 57; ausgegeben zu Bonn am 4. Dezember 2003. (11) Roussy, J.; Van Vooren, M.; Guibal, E. J. Dispers. Sci. Technol. 2004, 25/5, 663-677. (12) Kweon, M. R.; Bhirud, P. R.; Sosulski, F. W. Starch/Sta¨rke 1996, 48, 214-220. (13) Krentz, D.-O.; Lohmann, C.; Schwarz, S.; Bratskaya, S.; Liebert, T.; Laube, J.; Heinze, T.; Kulicke, W-M. Starch/Sta¨rke 2006, 58, 161169. (14) Krenz, D.-O. Dissertation, Universita¨t Hamburg, Hamburg, Germany, 2001. (15) Lohman, C. A. Dissertation, Universita¨t Hamburg, Hamburg, Germany, 2006.

Shirzad-Semsar et al. (16) Kulicke, W.-M., et al. German Patent 3.928.463, 1993. (17) Bo¨hm, N.; Kulicke, W-M. Colloid Polym. Sci. 1997, 275, 73-81. (18) Ehrenberger, F. QuantitatiVe Organische Elementaranalyse; Verlag Chemie: Weinheim, Germany, 1991; pp 436-440. (19) Gruber, E.; Grossmann, K.; Schempp, W. Wechselwirkungen Von Synthetischen Kationischen Polymeren mit Fasern und Fu¨llstoffen. Teil 1: Einfluss auf die Adsorption; Wbl. fu¨r Papierfabriaktion 124 Nr.; Deutscher Fachverlag: Frankfurt am Main, 1996; Vol. 1, pp 4-11. (20) Kulicke, W.-M.; Clasen, C. Viscosimetry of Polymers and Polyelectrolytes; Springer-Verlag: Berlin, 2004. (21) Arndt, K.-F; Mu¨ller, G. Polymercharakterisierung; Carl Heiser Verlag: Mu¨nchen, 1996. (22) Bartsch, S. Dissertation, Universita¨t Hamburg, Hamburg, Germany, 1998. (23) Kulicke, W.-M.; Ho¨rl, H. H. Colloid Polym. Sci. 1985, 263, 530540. (24) Thielking, H.; Roessner, D.; Kulicke, W.-M. On-Line Coupling of Flow Field-Flow Fractionation and Multiangle Laser Light Scattering for the Characterization of Polystyrene Particles. Anal. Chem. 1995, 67, 32293233. (25) Klein, J.; Kulicke, W.-M.; Hollmann, J. Chromatographie zur Bestimmung der Molmassen- und Teilchengro¨ssenverteilung von Polymeren. In Analytiker Taschenbuch Band 19; Springer: Berlin, 1998; 19. Auflage, pp 317-349. (26) Kulicke, W.-M.; Lange, S.; Heins, D. In AdVantages of Determining the Molar Mass Distribution of Water-Soluble Polymers and Polyelectrolytes with FFFF-MALLS and SEC-MALLS; Buchbeitrag im ACS Books Symposium Series No. 731/Chromatography of Polymers: Hyphenated and Multidimensional Techniques; Theodor, P., Ed.; John Wiley & Sons: New York, 1999; pp 114-140,. (27) Smellie, R. H; La mer, V. K, J. Colloid Sci. 1958, 13, 589-599. (28) Gregory, J. J. Colloid Interface Sci. 1973, 42, 448-456. (29) Mabire, F.; Audebert, R.; Quivoron, C. J. Colloid Interface Sci. 1984, 97, 448-456.

Cationic Starches as Substitute for Synthetic Cationic Flocculants in

Recommend Documents