Ultrasound-initiated synthesis of cationic polyacrylamide for oily wastewater treatment: Enhanced interaction between the flocculant and contaminants

Accepted Manuscript Ultrasound-initiated synthesis of cationic polyacrylamide for oily wastewater treatment: Enhanced interaction between the flocculant and contaminants Chuanliang Zhao, Huaili Zheng, Baoyu Gao, Yongzhi Liu, Jun Zhai, Shixin Zhang, Bincheng Xu PII: DOI: Reference:

S1350-4177(17)30510-2 https://doi.org/10.1016/j.ultsonch.2017.11.006 ULTSON 3943

To appear in:

Ultrasonics Sonochemistry

Received Date: Revised Date: Accepted Date:

19 September 2017 3 November 2017 3 November 2017

Please cite this article as: C. Zhao, H. Zheng, B. Gao, Y. Liu, J. Zhai, S. Zhang, B. Xu, Ultrasound-initiated synthesis of cationic polyacrylamide for oily wastewater treatment: Enhanced interaction between the flocculant and contaminants, Ultrasonics Sonochemistry (2017), doi: https://doi.org/10.1016/j.ultsonch.2017.11.006

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Ultrasound-initiated synthesis of cationic polyacrylamide for oily wastewater treatment: Enhanced interaction between the flocculant and contaminants Chuanliang Zhao 1,2, Huaili Zheng 1,2,*, Baoyu Gao 3, Yongzhi Liu 1,2, Jun Zhai 1,2, Shixin Zhang 1,2 and Bincheng Xu 1,2 1

Key laboratory of the Three Gorges Reservoir Region's Eco-Environment, State Ministry of Education, Chongqing University, Chongqing 400045, P.R. China

2

National Centre for International Research of Low-carbon and Green Buildings, Chongqing University, Chongqing 400045, P.R. China

3

Shandong Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science and Engineering, Shandong University, Ji’nan 250100, China

* Correspondence: [email protected]; Tel.: +86 23 65120827

Graphic abstract

1

Abstract: Weak interaction between flocculants and oil is a main bottleneck in the treatment of oil-containing wastewater. To solve this problem, a novel flocculant PAB with cationic micro-block structure and hydrophobic groups of benzene rings was synthesized by ultrasound initiated polymerization technique and applied to remove turbidity and oil from water. To avoid unnecessary addition of reagents in traditional template and micellar copolymerization, surface-active monomer benzyl(methacryloyloxyethyl)dimethylammonium chloride (BMDAC) with self-assembly ability in aqueous solution was employed to synthesize flocculants. The critical association concentration of BMDAC measured by conductivity and surface tension methods was 0.014 mol·L-1. The results of reactivity ratio, statistical analysis of sequence-length distribution and 1H NMR provided evidence for the synthesis of copolymer with cationic micro-block. In addition, the apparent viscosity measurement indicated that PAB had an obvious hydrophobic association property. Finally, flocculation tests demonstrated that flocculation performance was greatly improved by adding PAB and the removal rate of oil and turbidity both reached the maximum (87.5% and 92%) at dosage of 40mg·L-1 and pH of 7.0. Flocculation mechanism investigation demonstrated that the cooperation of charge neutralization, adsorption bridging, and hydrophobic association effect played an important role. The formed flocs by PAB was large, compact, difficult to break, and easy to regrow because of the enhanced interaction between flocculants and oil. In summary, this study can provide important reference in the design of organic flocculants in oily wastewater treatment applications. Keywords: Cationic polyacrylamide; Ultrasound; Micro-block structure; Hydrophobic association; Oily wastewater; Surface-active monomer

2

1. Introduction Nowadays, the amount of industrial oily wastewater discharge and leakage have increased dramatically with the rapid development of oil industries in China, such as leather, transportation, food, steel and petrochemical [1]. As one of the volatile contaminants, the oily wastewater not only polluted groundwater and surface water but also the air [2]. The improper disposal of oily wastewater would lead to ecological and environmental problems. However, the management of the oil-containing wastewater is confronted with crucial challenge due to the complex substances, usually high concentrations of dispersed oil, grease, and suspended particles. Different water treatment technologies, such as adsorption [3], air flotation [4], membrane filtration [5], biological treatment [6], flocculation–coagulation [7], and electrochemical treatment [8], have been applied to remove oil from oil-containing wastewater. Among the above methods, the flocculation technology, which is widely used in water treatment, showed unique advantages in oily wastewater treatment because of its no phase transition, easy operation, low cost, and good treatment efficiency. The properties of the flocculants was the crucial factors of flocculation performance. Therefore, the choice and design of flocculants with effective and efficient flocculation performance have been attracting increasing attention for oily wastewater treatment. Cationic polyacrylamide (CPAM) is the most common flocculant during flocculation process. It can simultaneously display charge neutralization and adsorption bridging functions because of its large molecular weight, high charge density, and many functional groups. However, satisfactory results cannot be obtained if the traditional CPAM is used to treat oily wastewater. There are two reasons, one is that cationic monomers are random distributed along the molecular chain. The 3

cationic charge has not been fully utilized, and thus the role of charge neutralization is discounted. In recent years, many researchers synthesized CPAM with concentrated cationic charge distribution through template polymerization methods [9-11]. However, not only does this method require a huge amount of templates, which influences the purity of the polymer, but also the addition of template reduces the CPAM molecular weight. The other reason is that the interaction between hydrophobic oily colloids and hydrophilic CPAM molecular chains was weak. It is necessary to introduce hydrophobic groups to flocculants backbones. The most commonly used method to introduce hydrophobic groups is micellar copolymerization in which the external conventional surfactant as the micelles to solubilize the hydrophobic monomer is usually employed [12]. However, the external surfactants could cause some negative impacts, such as chain transfer effects, complex post-treatment processes, and undesirable toxicity. To overcome the two problems, we introduced a new functional monomer to synthesize flocculants with cationic micro-block structures and hydrophobic groups. Surface-active monomers (surfmers) containing amphiphilic structure and polymerizable vinyl double bonds display unique physicochemical properties [13]. They can form organized molecular assemblies by self-assembly in aqueous solution, such as micelle, vesicle and lyotropic liquid crystal [14]. In the process of surfmers and acrylamides (AM) copolymerization, the pre-assembly of surfmers will improve the ordered nature of the reaction system. Once the polymerization is initiated by an initiator, organized molecular assemblies will generate block structure in polymers. Meanwhile, the hydrophilic groups can ionize to give cationic sites along the chain and the hydrophobic groups can enhance interactions with hydrophobic oily colloids. Therefore, the process and reagents can be greatly simplified compared with

4

traditional methods. In addition, ultrasonic initiated polymerization has drawn wide attention due to the characteristics of faster polymerization rate, narrower molecular weight distribution, higher monomer conversion and smaller amount of initiator than other initiation systems [15]. Furthermore, more stable uniform monomer assemblies get formed in ultrasound-initiated polymerization because of cavitational activity near the interface between aqueous phase and hydrophobic pocket by micro jets , which is conducive to the introduction of the cationic micro-block structure [16]. To the best of our knowledge, this is the first investigation of an ultrasound initiated CPAM with cationic micro-block structure and hydrophobic groups for the removal of oil from water. This technology of ultrasound initiated copolymerization with surface-active monomers is promising in the field of structural design and optimization of new flocculants, because of the use of less hazardous chemicals, a simplified preparation process and an increased product performance. In this study, the surfmer, benzyl(methacryloyloxyethyl)dimethylammonium chloride (BMDAC) with self-assembly ability in aqueous solution was adopted to prepare the P (AM-BMDAC) (PAB) initiated by ultrasound with potassium persulfate as initiator. Another monomer, methacrylatoethyl trimethyl ammonium chloride (DMC), was used to prepare the copolymer (P (AM-DMC) (PAD)) under the same synthetic conditions in that the DMC is similar to BMDAC in structure but without self-assembly and hydrophobic properties. In order to elucidate the effect of surfmer, a comparison between PAB and PAD in this study was conducted. The reactivity ratios of monomers were calculated, and the sequence distribution of the molecules was statistically analyzed. The properties of polymers were characterized by apparent viscosity and 1HNMR spectroscopy. Moreover, flocculation performance was optimized by investigating the effect of wastewater initial pH, dosage on turbidity and

5

oil removals. Meanwhile, the flocculation mechanism was discussed by a detailed investigation of the zeta potentials and flocs morphological properties. 2. Materials and methods 2.1 Materials AM, BMDAC, and DMC used in this experiment were all of technical grade and used without further purification. The monomer AM (98.5%, w/w) was obtained from Jiangxi Changjiu Biochemical Industry Co., Ltd. (Nanchang, China); BMDAC (60 wt% in water) and DMC (60 wt % in water) was procured from Zibo Wonderful Chemical Co., Ltd. (Zibo, China). Analytical grade urea [CO(NH2)2], hydrochloric acid (HCl), sodium hydroxide (NaOH) and potassium persulfate (K2S2O8) were purchased from Aladdin Co. Ltd. (Shanghai, China). All aqueous and standard solutions were prepared with deionized water supplied by a Milli-Q Plus water purification system (Millipore). 2.2 Preparation of copolymer The copolymer synthesis by AM and BMDAC was carried out in a 100ml glass reactor which was put into a 5 L water tank. A standard sonic wave probe (tip diameter: 13 mm, length: 136 mm) mounted right above water tank was immersed into the liquid level. An ultrasonic generator (Sonics Vibra-cell, VCX 500) controlled by standard power source (net power output: 500 W, frequency: 20 kHz) was used as an ultrasonic source to initiate the polymerization. The monomer solution was prepared by adding predetermined amount of AM (155.93 mmol) and BDMDAC (17.33 mmol) in distilled water to reach a monomer mass ratio of 40%, and it was then transferred to the reactor. Then, urea (2.0 wt % of total monomers mass) was added to increase the solubility and the pH of the reaction solution was adjusted to 4 (±0.2) by HCl and NaOH solution. 6

Prior to the addition of the initiator (3.0 wt % of total monomers mass), the reaction vessel was purged with nitrogen gas (99.99%) at the rate of 40 mL·min-1. The ultrasound power and irradiation time were set to 450 W and 30 min, and the optimization results were presented in Supplementary Text S1. After irradiation, a milk-white semitransparent solid was obtained which indicated the formation of copolymer. The products were purified with acetone and ethanol for several times and then dried in a vacuum oven at 60°C for 2 days. PAD was prepared by the same synthesis process. The possible reaction mechanisms of PAB and PAD are shown in Supplementary Scheme S1. 2.3 Measurement of critical association concentration (CAC) The pre-assembly of BMDAC will improve the ordered nature of the reaction system. The change of aqueous solution properties can reflect the behavior of self-assembly. Therefore, the conductivity and surface tension was measured by a DDS-307 conductivity meter (INESA Scientific Instrument Co., Ltd, Shanghai) and BCZ-600 surface tension meter (Benchuang Analytical Instrument Co., Ltd, Zibo), respectively. Based on the results, curves were drawn and the intersection of changed two sides was CAC. 2.4 Determination of the monomer reactivity ratio Monomer reactivity ratio (r), namely, the ratio of chain radical homopolymerization rate constant to the copolymerization rate constant reflects the relative polymerization activity [17]. It was defined based on Eq. (1):  (1)  where k11 and k12 are the rate constants of homopolymerization and copolymerization,  =

respectively. When r is more than 1, the monomer tends to homopolymerize with the 7

same monomers; otherwise, the monomer tends to copolymerize with other monomers. Besides, the larger the value is, the higher the relative activity will be. Monomer reactivity ratio is a crucial parameter to control polymer composition and structure in binary copolymerization [18]. In this experiment, two series of polymer products, each containing eight PAB or PAD samples, were prepared under the same condition described in section 2.2 except that the monomer conversions of all samples were controlled to less than 10%. Prior to synthesis, the reaction molar ratios of AM and cationic monomers were controlled at 9:1, 8:2, 7:3, 6:4, 5:5, 4:6, 3:7, 2:8 and 1:9. Afterward, the cationic degree of polymer was measured by colloid titration. The monomer reactivity ratios were calculated with Fineman-Ross, Kelen–Tüdö and Yezrielev−Brokhina−Roskin (Y−B−R) methods of which the detailed procedures were shown in Supplementary Text S2-S4 and Fig. S4-S7 [19]. 2.5 Sequence-length distributions of polymers Sequence-length distribution is referred to the monomer unit amount in its continuous chain segment and is important to analyze the block structure of the polymer. In order to elucidate the influence of monomer self-assembly on the polymer microstructure, four molar feed ratios (fm) between AM and BMDAC or DMC (8:2, 6:4, 4:6, 2:8) were adopted for statistical analysis. The details of the calculating equations are exhibited in Supplementary Text S5. 2.6 Characterizations Intrinsic viscosity of copolymer was measured by an Ubbelohde viscosity meter (Shanghai, China) equipped with a thermostatic water container (DKB-501S, Shanghai Jing Hong Laboratory Instrument Co., Ltd.). The 1H NMR spectra were obtained in 8

deuterium oxide (D2O) with an AVANCE 500 NMR spectrometer (BRUKER Company, Germany). Rheological characteristic was measured by Anton Paar MCR302 rheometer (Anton Paar GmbH, Austria). 2.7 Flocculation tests 2.7.1 The characteristic of the wastewater samples

The simulated oil-containing wastewater was prepared with 0# diesel and octyl phenol polyoxyethylene ether (OP-10) as emulgator. The raw water was characterized by measuring the oil, turbidity and zeta potential. Chemical analysis of the wastewater showed that the concentrations of oil, turbidity, pH values, and zeta potential were 500 mg/L, 212 NTU, 7.0, and -6.43 mV, respectively. The oil content of the wastewater was measured by an UV–visible spectrophotometer (TU-1901, Beijing Purkinje General Instrument Co., Ltd., China). The absorbance was detected at the wavelength of 261 nm. A calibration curve was measured in advance. The oil removal efficiency was calculated as follows:   × 100% (2)  where C0 and C were the oil concentrations in the solution before and after Oil removal (%) = 1 −

flocculation, respectively. 2.7.2 Flocculation experiments The flocculation experiment was performed using a programmed jar test apparatus (ZR4-6, Zhongrun Water Industry Technology Development Co. Ltd., China). Oily wastewater (700 mL) was transferred into a 1.0 L beaker, and the initial pH of the wastewater was adjusted to the set value using 0.5 mol/L HCl and 0.5 mol/L NaOH. A

9

series of flocculants (shown in Table 1) were chosen to comparatively analyze the flocculation efficiency. Before each test, the stock solution of flocculants was always freshly prepared by dissolving 0.1 g of flocculants into 100 mL of water. A measured amount of stock solution was pipetted into the wastewater sample. Then, the wastewater samples were mixed rapidly at 200 rpm for 60 s after dosing, followed by slow stirring at 40 rpm for 15 min and sedimentation for 30 min. After sedimentation, supernatant samples were collected from 5 cm below the water surface of the wastewater samples to measure oil content and turbidity (2100Q turbidimeter, HACH, USA). The zeta potential of wastewater after flocculation was immediately performed on the Zetasizer Nano ZS90 (Malvern Instruments Ltd., Malvern, UK). Table 1. Details of used flocculants in flocculation experiments. Intrinsic viscosity Average molecular Cationic degree (dL·g-1) weight (10 6 Da) (%) PAB 6.94 20 2.86 PAD 7.18 2.98 20 CCPAM 7.22 3.00 60 P (DMDAAC) 3.70 100 1.30 1 CCPAM: commercial cationic polyacrylamide; P (DMDAAC): homopolymer of dimethyldiallylammonium chloride. Flocculants 1

2.7.3. Characterization of morphological properties of flocs Floc size was measured by a laser diffraction analyzer (Mastersizer2000, Malvern, UK). Dynamic sludge flocs size was monitored during the whole formation, breakage and re-flocculation process. d50 (50% of the flocs were of the sizes in the range 0-d50) was adopted to represent the average floc size. Floc strength (SF) and recovery factors (RF) were widely used to compare the degree of breakage and re-flocculation and were calculated as follows [20]:  =

 (3) 

10

! =

 − (4) −

where A, B and C are the average flocs sizes in the steady phase before breakage, after the breakage period and after the re-growth to new steady phase, respectively. The photos of oily flocs formed by flocculants were collected by a digital camera, and the fractal dimension (Df) of flocs calculated through image analysis method on the basis of photos was used to evaluate the surface morphology and fractal structure of flocs. The image-pro plus 6.0 software was employed to analyze the floc photos formed by flocculants with optimal dosage. A double logarithmic relationship of projected perimeter (L) and projected area (A) of selected region from floc photo was given for two dimensional fractal dimension and was shown in Eqs.(5) and (6) [21].

 = # × $%&

(5)

ln() = () × ln($) + ln(#)

(6)

where K is a constant. In Eq. (6), ln(L) and ln(A) are the independent variable and dependent variable, respectively. Finally, the linear regression between ln(L) and ln(A) was established accordingly, and the obtained line slope was the fractal dimension of flocs. 3. Results and discussion 3.1 Self-assembly performance of PAB The high-performance flocculants with cationic micro-block structures and hydrophobic groups was designed on the basis of the self-assembly ability of BMDAC. It is necessary to analysis the aggregation behavior. BMDAC was free distribution in solution under low concentration. On the contrary, as monomer concentration increased, a large number of surfmer molecules associated with each other forming

11

assemblies. The minimum concentration of association was called CAC [22]. The aqueous solution properties, such as the conductivity and surface tension changed obviously around the CAC. As shown in Fig. S2, when the concentration of BMDAC reached CAC, the increased trend of conductivity slowed down. This was due to that the movement of counter-ions was restricted. When surfmer molecules associated with each other forming assemblies, the counter-ions was fixed on the surface of assemblies. Therefore, their contribution to conductivity dropped off. As shown in Fig. S3, the surface tension tended to a constant value after the concentration reached CAC. The assemblies were easy to concentrate on the surface resulting in molecular oriented arrangement, and then the surface tension reduced [23]. According to the measurement of conductivity and surface tension, the CAC of BMDAC was 0.014 mol·L-1. In the following synthesis, the concentration of BMDAC must be kept higher than 0.014 mol·L-1 to develop the micro-block structure. 3.2 Monomer reactivity ratios of the polymers Table 2. Monomer reactivity ratios of the polymers. Methods Fineman-Ross Method Kelen–Tüdö Method Y−B−R Method Average

PAD rAM 0.64 0.63 0.64 0.64

rDMC 0.42 0.39 0.39 0.40

PAB rAM 0.55 0.57 0.57 0.56

rBMDAC 1.47 1.52 1.56 1.52

The reactivity ratios of AM and cationic monomers calculated according to three methods were shown in Table 2 and the results were close to previous study [24,25]. In PAD copolymerization, rAM and rDMC were both lower than 1 which indicated monomers tend to copolymerization rather than homopolymerization [26]. Therefore, monomers inclined to copolymerize disorderly forming shorter chain segments of same units in polymer molecule chain. As opposed to above scenario, rBMDAC was more than 12

1 in PAB system, which was attributed to stronger self-assembly ability of BMDAC. Before polymerization reaction, BMDAC could form micellar aggregation in aqueous phase whereas DMC just assumed dispersed state. The concentration of surfmers in micellar aggregation was similar to bulk polymerization, facilitating homopolymerization. When encountering growing AM macroradical headgroup, they were embedded in polymer in the form of aggregation, of which the structure was regarded as micro block. Accordingly, it was deduced that there were more and longer cationic blocks in the molecular chain of PAB than that in PAD. Whereas, owing to that AM reactivity ratio was less than 1, the PAB copolymer was composed of long BMDAC segments embed into short AM chain segments [27]. In addition, more stable uniform monomer assemblies get formed in ultrasound initiated polymerization because of cavitational activity at near interface between aqueous phase and hydrophobic pocket by micro jets, which is conducive to the introduction of cationic micro-block structure. The above procedure and experimental setup were depicted in Fig. 1.

Fig. 1 Schematic experimental setup and mechanism of the copolymerization. 3.3 Sequence distributions of the polymers The sequence distributions of the monomer segments were in terms of microscopic perspective to evaluate molecular arrangements in comparison with macroscopical polymer composition equations. The results were summarized in Fig. 2 and the detailed experimental data were listed in the Supplementary Tables S3-S6. As shown in Fig. 2, the AM segments with the length of 1, 2, 3 in PAD (PAB) under fm=8:2 accounted for 28.22% (30.71%), 20.26% (21.28%), 14.54% (14.74%), respectively.

13

Correspondingly, the DMC (BMDAC) segments with the length of 1, 2, 3 in PAD (PAB) accounted for 90.97% (72.49%), 8.21% (19.94%), 0.7% (5.49%), respectively. It illustrated that the sequence distribution of cationic monomers was narrower with shorter chain segments, whereas AM sequence distribution was wider and longer under this condition. When the initial molar ratio of reactants increased to 6:4 or 4:6, monomers in the two polymers tended to an alternate distribution along the molecular chain. Sequentially increased to 2:8, the situation of the AM and DMC (BMDAC) segments was just the contrary to fm = 8:2. Our results suggested that monomers sequence distributions were mainly dependent on fm, namely, feed ratio. In addition, there were more long cationic chain segments in PAB than those in PAD under the same feed ratio. Besides, the length advantage became more and more obvious with the increase of feed ratio. This observation may be explained by that BMDAC monomers are inclined to form bigger micelle aggregates at higher dosage which contributed to embedding longer cationic chain segments when encountered with growing macroradical. The results are in agreement with the analyses of reactivity ratios. In a word, PAB had more and longer cationic blocks in molecular chain than PAD.

Fig. 2 Sequence distributions of BMDAC and AM in the polymers under (a) fm=8:2, (b) fm=6:4, (c) fm=4:6 and (d) fm=2:8. 3.4 Characterizations of the polymers 3.4.1 Apparent viscosity of PAB and PAD Generally, apparent viscosity is employed as a convenient and reliable method for investigating the associative performance of copolymers in water solution [28]. In this study, apparent viscosity was determined as a function of polymer concentration. In 14

order to exclude the effect of molecular weight and cationic degree, the parameters of the two copolymers was similar which were listed in Table 1. As shown in Fig. 3, the apparent viscosity of the two copolymers increased with the increasing concentration. There is a saltation at around 0.3 g·dL-1 in the curve of PAB, while the curve of PAD was relatively flat. In addition, it is worth noting that the apparent viscosity of PAB and PAD are almost the same in the beginning. However, while the continuously increasing concentration to above 0.3 g·dL-1, the apparent viscosity of PAB increase remarkably whereas that of PAD maintains the linear increasing as before. This phenomenon led to the apparent viscosity of PAB higher than that of PAD above 0.3 g·dL-1, which was attributed to two reasons. First, the introduction of the hydrophobic benzyl group on the polymer chain is beneficial for hydrophobic association. Intramolecular association was dominant in solvent when the polymer concentration was below the CAC. Such intramolecular association will lead to the curly configuration of the polymer chain to reduce the hydrodynamic volume [29]. Thus, the apparent viscosity of PAB is low in the beginning. On the contrary, when polymer concentration is above CAC, intermolecular association dominated forming a physical network structure of polymer chains with giant hydrodynamic volume [30]. Thus, high apparent viscosity is observed in the curve of PAB. Second, BMDAC was inclined to form micro-block structure in the PAB, while DMC was random distribution in PAD. The micro-block was conducive to the occurring of intermolecular association. Thus, PAB presented higher apparent viscosity. The apparent viscosity of PAB and PAD achieve 46.4 mPa·s-1 and 34.2 mPa·s-1 at 0.8 g·dL-1, respectively.

Fig.3 Apparent viscosity as a function of polymer concentration for PAB and PAD

15

3.4.2 1H NMR spectra Fig. 4 demonstrated the 1H NMR spectra of copolymers and cationic monomers homopolymers. It could be seen from Fig. 4 (a) that the asymmetric peaks at δ = 1.69 ppm and δ = 2.23 ppm were associated with the protons of the backbone methylene (a1) and methine groups (b 1), respectively. Protons in the two sequential methylene groups (c1, d1) connected to the ammonium group of BMDAC were observed at δ = 4.64 and 3.79 ppm. The sharp peak at δ = 3.15 ppm was attributed to the protons of the two equivalent methyl (e1) at quaternary ammonium groups. The peaks at δ = 7.60 and 4.61ppm were assigned to protons of the phenyl group (g1) and methylene (f1) connected to the phenyl group. The comparison between P(BMDAC) and PAB revealed that most protons of cationic monomer appeared in PAB, which confirmed the successful copolymerization of BMDAC and AM. Moreover, the chemical shifts of adsorption peaks were very slight, indicating that the chemical environments of cationic monomer in PAB and homopolymer were almost the same. On the other hand, there are many split peaks marked with stars and arrows in PAD compared with P(DMC) shown in Fig. 4 (b). The result can be explained with the stereochemistry of the copolymers. PAD was a randomly distributed copolymer of AM and DMC, thus the adsorption peaks of cationic monomers were remarkably interfered with by adjacent protons generating more diverse chemical environments [31]. Nevertheless, the phenomena did not occur in PAB, which manifested the BMDAC was distributed in the form of block structure along copolymer molecular chain. Fig. 4 1H NMR spectra of the polymers: (a) P(BMDAC) and PAB; (b) P(DMC) and PAD. 3.5 Flocculation properties 3.5.1 Effect of polymer dosage 16

The above characterization results of copolymers corroborate that there is a blocky distribution of functional units in PAB and a random distribution in PAD. Therefore, to accurately evaluate the impact of cationic sequence distribution and hydrophobic groups on oil removal efficiency, four flocculants detailed in Table 1 were chosen for comparison. Fig. 5 presented the results of flocculation under different dosages. Obviously, the trend of oil removal rate is similar to that of turbidity removal rate, which was due to the fact that the turbidity of an emulsion is related to the concentration of droplets. But turbidity also depends on the size of oil droplets. As the droplets tend to coalesce, the turbidity may decrease due to a lower number of droplets or an increase in their size. Therefore, there exist a difference between them. It can be seen that oil and turbidity removal rate increased with increasing flocculants and reached the highest point at the optimal dosage, then decreased when continuing increase dosages. From Fig. 5 (a) and (b) it also can be seen that the optimal oil removal rate (former in parentheses) and turbidity removal rate (latter in parentheses) of each flocculant were PAB (87.5%, 92%), CCPAM (82.1%, 87.2%), PAD (81.1%, 85.9%), and P (DMDAAC) (74.2%, 82.8%). This result was closely related to the molecular structure and performance parameters of four flocculants. As shown in Fig. 5 (c), the zeta potential of CCPAM was highest among these four flocculants, indicating the strongest charge neutralization ability. But the removal rate was lower than that of PAB. The optimal removal rates of oil and turbidity were 87.5% and 92%, respectively, which obtained at the PAB dosage of 40 mg/L. Compared with CCPAM, the better flocculation performance of PAB may be ascribed to the introduction of hydrophobic groups, which enhanced the interaction between flocculants and oily drops. Another phenomenon worth noting is that the zeta potential of PAD was lower than that of PAB in the investigated dosage range. As shown in

17

Table 1, the two flocculants have the same cationic degree. Therefore, the discrepancy of zeta potential was due to cationic sequence distribution. The blocky distribution could concentrate cationic charges in polymer molecular chain, which contributed to more sufficiently neutralizing the surface negative charge and compressing electrical double layer of oily drops [31]. In addition, the flocculation performance of P (DMDAAC) was worst in which there were only cationic monomers. The polymerization activity of cationic monomers was lower than AM, thus the molecular weight cannot match up to the copolymers. Homopolymer could not perform well in flocculation for the weakly bridge effect.

Fig. 5 Effects of the dosage on the flocculation performance and zeta potential. 3.5.2 Effect of initial wastewater pH value The surface charge of the colloidal particles and the ionization degree of functional groups in the flocculants were closely linked to pH value [33]. Hence, determining the optimum pH value was necessary for practical wastewater treatment. The flocculation performance was investigated in the pH range of 3.0–9.0 at their optimal dosage and the results are shown in Fig. 6. Obviously, the removal rates of oil and turbidity increased in pH from 3.0 to 5.0. Then, the increasing trend became slow from 5.0 to 7.0. When the pH value was greater than 8, the flocculation performance significantly deteriorated. The optimal removal rates of turbidity and oil for four flocculants were all attained at pH 7.0. The difference was that PAB presented the highest removal efficiency at almost all investigated pH. Furthermore, it had wider pH scope for application. From Fig. 6 （c） ）, the zeta potential of colloidal system declined continuously as pH rose. Meanwhile, the absolute value of zeta potential was greater below pH 5.0 or

18

above pH 8.0, indicating that oil drops took more positive or negative charges. Surplus charges resulted in strong electrostatic repulsion among oil drops. Therefore, the oily particles cannot be captured together by polymer chains effectively, lowering the efficiency of the bridging. In the pH range of 5.0−8.0, the zeta potential of the system gradually transitioned from positive charge to negative charge regions. There are only a few charges on the colloidal surface. Thus, the adsorption bridging effect among colloidal particles and polymer chain was strengthened. The PAB displayed excellent flocculation performance, even in an acidic environment. Other than charge neutrality, the hydrophobic association effect could also be responsible for the flocculation behavior.

Fig. 6 Effect of initial wastewater pH value on the flocculation performance and zeta potential. 3.5.3 Flocs size and Fractal dimension In the practical flocculation process, large and compact flocs always rapidly separate from water. Since the fractal theory was introduced to understand flocculation mechanism, many studies have shown that flocs could exhibit excellent fractal characteristics [34-36]. The fractal dimension as a critical parameter was commonly employed to describe the density and compactness of flocs. Thus, two-dimensional fractal dimensions (Df) of flocs at optimal dosages were calculated through the image-pro plus 6.0 software. The size and Df of flocs were compared and analyzed in present research. The results are depicted in Figs. 7 and 8. As shown in Fig. 8 (a–d), the Df of flocs was 1.50, 1.46, 1.39 and 1.31 for PAB, CCPAM, PAD and P (DMDAAC), respectively. Many researchers have reported that flocs have a more compact structure when the Df value is greater [37,38]. Fig. 7 (a-d) 19

also demonstrated that obvious larger and more intensive flocs formed after the flocculation by PAB. These results also showed that the introduction of surfmers significantly improved the flocs size and fractal dimension. Previous study corroborated cationic micro-block structure had a negative effect on flocs size [39]. The strengthened charge neutralization ability effectively reduced charge repulsion and compressed the electric double layer of colloids forming smaller and more compact flocs. On the other hand, hydrophobic groups can enhance the interaction between oily drops and polymer chains facilitating the adsorption bridging. Meanwhile, the intermolecular hydrophobic association may contribute to forming complex physical network which trapped many small flocs. Hence, it can be speculated that both hydrophobic association and stronger charge neutralization together took effect to generate large and compact flocs for the wastewater flocculated by PAB. Compared with PAB, flocs by PAD and P (DMDAAC) were smaller and looser, resulting in worse flocculation performance. To obtain similar results, high cationic degree polymer like CCPAM with 60% must be used. Nevertheless, owing to the high prices of cationic monomers (shown in Supplementary Table S7), it is inevitably that the economic cost would be increased sharply when high cationic degree polymer was used in coagulation-flocculation.

Fig. 7 Floc size distribution of wastewater flocculated by (a) PAB, (b) PAD, (c) CCPAM and (d) P (DMDAAC) at optimal dosages. Fig. 8 Fractal dimension comparison of flocs formed in (a) PAB, (b) PAD, (c) CCPAM and (d) P (DMDAAC) flocculation at optimal dosages. 3.6 Breakage and re-flocculation of flocs

Fig. 9 Effects of different flocculants on breakage and re-flocculation. 20

Table 3. Floc parameters of different flocculants under the same condition.

Flocculants PAB PAD CCPAM P (DMDAAC)

SF(%) 62.5 47.2 53.1 38.5

RF(%) 34.3 17.1 34.7 45.5

The flocculation suspensions are always exposed to high shear force in the real industrial process. The flocs would break into pieces and regenernate to a extent when the shear force disappeared. As shown in previous research [40], the flocs strengh and re-flocculation ability have an important influence on flocculation efficiency. When polyelectrolytes were used to induce aggregation, the change in floc morphology depend on flocculation mechanism [41]. The variations of average floc size (d 50) with different flocculants can be found in Fig. 9. Obviously, the floc size increased rapidly and reached to the equilibrium during the slow stirring. A sudden increase in stirring speed resulted in breakage of the flocs and then they gradually increased when the shear force returned to previous level. The strength and recovery factors of flocs by different flocculants were calculated and presented in Table 3. It was noted that the strength of PAB flocs was the highest, which may related to the enhanced interaction between flocculants and oily drops involving hydrophobic association and charge neutralization. On the other hand, our previous research showed that blocky distribution of cationic charges induced extension adsoption on the colloid surface, which is benefitcal for bridging. In general, when bridging effect was more intensive, the flocs are strongger [42]. The P (DMDAAC) flocs were the most fragile since the primary mechenism was electrostatic patching. On the contrary, the re-flocculation ability of P (DMDAAC) flocs was the highest whereas RF of PAD was only 17.1%. Exposed to high shear force, the tails and loops of PAD molecules were broken. Therefore, the flocculation dominated by bridging is difficult to recover to previous

21

level. But in the case of electrostatic patching, shear force has low effect on the polymers with low molecular weight and high charge density [43]. Particularly, the re-flocculation ability of PAB (RF= 34.3%) was still high. The phenomenon may be caused by cationic micro-blocks remaining on the surface of oily drops which could act as “cationic islands” inducing electrostatic patching re-flocculation. 3.7 Flocculation mechanism On the basis of the above results and discussion, the flocculation mechanism of PAB was schematically depicted in Fig. 10. During PAB removing oil process, charge neutralization, adsorption bridging, and hydrophobic association would cooperate with each other to play an important role. Specifically, cationic micro-block in macromolecular chains effectively neutralized the surface negative charge of emulsifying oil drops making oil drops aggregation and destabilization. Long polymer chains increase the chance of the collision with colloid particles by adsorption bridging effect, which made destabilized oil drops form flocs. Under hydrophobic groups interaction, the flocculants and oily drops are able to be tightly bound together. Furthermore, the intermolecular hydrophobic association would contribute to forming bigger flocs, and then the oil was removed from water. The coordination of three effects tremendously enhanced the removing performance. More importantly, the problems of weak interaction between flocculants and oil were solved by cationic micro-block and the introduction of hydrophobic groups, which may give valuable guidance in optimization of new flocculants to remove targeted contaminants.

Fig. 10 Possible flocculation mechanism of PAB. 4. Conclusion

22

In this study, a novel flocculant PAB with cationic micro-block structure and hydrophobic groups of benzene rings was successfully synthesized by ultrasound initiated polymerization technique. In comparison with other flocculants, PAB displayed prominent flocculation performance in the removal of turbidity and oil. The optimal removal rates of oil and turbidity were 87.5% and 92%, respectively, which obtained at the dosage of 40 mg/L, pH of 7.0. Flocculation mechanism investigation demonstrated that the cooperation of charge neutralization, adsorption bridging, and hydrophobic association effect played an important role. The formed flocs by PAB were large, compact, difficult to break, and easy to regrow because of the enhanced interaction between flocculants and oil. This excellent flocculant benefitted from the successful combination of surfmers and ultrasound initiated polymerization technology. These results had notable practical significance for structural design and optimization of new flocculants, because of the use of less hazardous chemicals, a simplified preparation process and an increased product performance.

Acknowledgments: This research was supported by the National Natural Science Foundation of China (Project No. 21677020 and 21477010).

23

References: [1] T. Yang, B. Qiao, G. Li, Q. Yang, Improving performance of dynamic membrane assisted by electrocoagulation for treatment of oily wastewater: Effect of electrolytic conditions, DESALINATION 363 (2015) 134-143. [2] Y. Sun, C. Zhu, H. Zheng, W. Sun, Y. Xu, X. Xiao, Z. You, C. Liu, Characterization and coagulation behavior of polymeric aluminum ferric silicate for high-concentration oily wastewater treatment, Chem. Eng. Res. Des. 119 (2017) 23-32. [3] X. Liu, L. Ge, W. Li, X. Wang, F. Li, Layered Double Hydroxide Functionalized Textile for Effective Oil/Water Separation and Selective Oil Adsorption, ACS Appl. Mater. Inter. 7 (2015) 791-800. [4] P. Painmanakul, P. Sastaravet, S. Lersjintanakarn, S. Khaodhiar, Effect of bubble hydrodynamic and chemical dosage on treatment of oily wastewater by Induced Air Flotation (IAF) process, Chem. Eng. Res. Des. 88 (2010) 693-702. [5] X. Zhu, A. Dudchenko, X. Gu, D. Jassby, Surfactant-stabilized oil separation from water using ultrafiltration and nanofiltration, J. Membrane Sci. 529 (2017) 159-169. [6] N. Sarac, A. Ugur, A green alternative for oily wastewater treatment: lipase from Acinetobacter haemolyticus NS02-30, Desalin. Water Treat. 57 (2016) 19750-19759. [7] Z.L. Yang, B.Y. Gao, C.X. Li, Q.Y. Yue, B. Liu, Synthesis and characterization of hydrophobically associating cationic polyacrylamide, Chem. Eng. J. 161 (2010) 27-33. [8] C. An, G. Huang, Y. Yao, S. Zhao, Emerging usage of electrocoagulation technology for oil removal from wastewater: A review, Sci. Total Environ. 579 (2017) 537-556. [9] L. Feng, H. Zheng, Y. Wang, S. Zhang, B. Xu, Ultrasonic-template technology inducing and regulating cationic micro-blocks in CPAM: characterization, mechanism and sludge flocculation performance, RSC Adv. 7 (2017) 23444-23456. [10] L. Feng, H. Zheng, B. Gao, S. Zhang, C. Zhao, Y. Zhou, B. Xu, Fabricating an anionic polyacrylamide (APAM) with an anionic block structure for high turbidity water separation and purification, RSC Adv. 7 (2017) 28918-28930. [11] L. Feng, H. Zheng, B. Gao, C. Zhao, S. Zhang, N. Chen, Enhancement of textile-dyeing sludge dewaterability using a novel cationic polyacrylamide: role of cationic block structures, RSC Adv. 7 (2017) 11626-11635. [12] H. Zheng, Y. Sun, C. Zhu, J. Guo, C. Zhao, Y. Liao, Q. Guan, UV-initiated polymerization of hydrophobically associating cationic flocculants: Synthesis, characterization, and dewatering properties, Chem. Eng. J. 234 (2013) 318-326. [13] C. Zhao, H. Zheng, L. Feng, Y. Wang, Y. Liu, B. Liu, B.Z. Djibrine, Improvement of Sludge Dewaterability by Ultrasound-Initiated Cationic Polyacrylamide with Micro-block Structure: The Role of Surface-Active Monomers, MATERIALS 10 (2017). [14] M. Summers, J. Eastoe, Applications of polymerizable surfactants, Adv. Colloid Interfac. 100 (2003) 137-152. [15] K. Sankar, V. Rajendran, Ultrasound assisted free radical polymerization of glycidyl methacrylate by a new disite phase-transfer catalyst system: A kinetic study, Ultrason. Sonochem. 19 (2012) 1205-1212. [16] B.A. Bhanvase, D.V. Pinjari, S.H. Sonawane, P.R. Gogate, A.B. Pandit, Analysis of semibatch emulsion polymerization: Role of ultrasound and initiator, Ultrason. Sonochem. 19 (2012) 97-103. 24

[17] C. Erbil, S. Ozdemir, N. Uyanik, Determination of the monomer reactivity ratios for copolymerization of itaconic acid and acrylamide by conductometric titration method, POLYMER 41 (2000) 1391-1394. [18] S. Polowinski, Template polymerisation and co-polymerisation, Prog. Polym. Sci. 27 (2002) 537-577. [19] P. Bera, S.K. Saha, Water-soluble copolymers of acrylamide with diacetone - acrylamide and N-t-butylacrylamide on aqueous montmorillonite surface: synthesis and characterisation, Eur. Polym. J. 36 (2000) 411-419. [20] P. Jarvis, B. Jefferson, S.A. Parsons, Breakage, regrowth, and fractal mature of natural organic matter flocs, Environ. Sci. Technol. 39 (2005) 2307-2314. [21] J. Ma, K. Fu, L. Jiang, L. Ding, Q. Guan, S. Zhang, H. Zhang, J. Shi, X. Fu, Flocculation performance of cationic polyacrylamide with high cationic degree in humic acid synthetic water treatment and effect of kaolin particles, Sep. Purif. Technol. 181 (2017) 201-212. [22] C.L. MCCORMICK, J.C. MIDDLETON, D.F. CUMMINS, WATER-SOLUBLE COPOLYMERS .37. SYNTHESIS AND CHARACTERIZATION OF RESPONSIVE HYDROPHOBICALLY-MODIFIED POLYELECTROLYTES, MACROMOLECULES 25 (1992) 1201-1206. [23] S.R. Kline, Polymerization of rodlike micelles, LANGMUIR 15 (1999) 2726-2732. [24] H. TANAKA, COPOLYMERIZATION OF CATIONIC MONOMERS WITH ACRYLAMIDE IN AN AQUEOUS-SOLUTION, J. Polym. Sci. Pol. Chem. 24 (1986) 29-36. [25] C. VU, J. CABESTANY, CHARACTERIZATION OF CATIONIC WATER-SOLUBLE POLYACRYLAMIDES, J. Appl. Polym. Sci. 42 (1991) 2857-2869. [26] I. Erol, B. Sahin, Functional styrenic copolymer based on 2-(dimethylamino)ethyl methacrylate: Reactivity ratios, biological activity thermal properties and semi-conducting properties, J. Fluorine Chem. 178 (2015) 154-164. [27] T.V. Plisko, A.V. Bildyukevich, Debundling of multiwalled carbon nanotubes in N, N-dimethylacetamide by polymers, Colloid Polym. Sci. 292 (2014) 2571-2580. [28] Y. Liao, H. Zheng, L. Qian, Y. Sun, L. Dai, W. Xue, UV-Initiated Polymerization of Hydrophobically Associating Cationic Polyacrylamide Modified by a Surface-Active Monomer: A Comparative Study of Synthesis, Characterization, and Sludge Dewatering Performance, Ind. Eng. Chem. Res. 53 (2014) 11193-11203. [29] L. Jia, L. Yu, R. Li, X. Yan, Z. Zhang, Synthesis and solution behavior of hydrophobically associating polyacrylamide containing capsaicin-like moieties, J. Appl. Polym. Sci. 130 (2013) 1794-1804. [30] Q. Yang, C. Song, Q. Chen, P. Zhang, P. Wang, Synthesis and Aqueous Solution Properties of Hydrophobically Modified Anionic Acrylamide Copolymers, J. Polym. Sci. Pol. Phys. 46 (2008) 2465-2474. [31] Z. Abdollahi, M. Frounchi, S. Dadbin, Synthesis, characterization and comparison of PAM, cationic PDMC and P(AM-co-DMC) based on solution polymerization, J. Ind. Eng. Chem. 17 (2011) 580-586. [32] W. Chen, H. Zheng, Q. Guan, H. Teng, C. Zhao, C. Zhao, Fabricating a Flocculant with Controllable Cationic Micro-block Structure: Characterization and Sludge Conditioning Behavior Evaluation, Ind. Eng. Chem. Res. 55 (2016) 2892-2902. [33] S. Jia, Z. Yang, W. Yang, T. Zhang, S. Zhang, X. Yang, Y. Dong, J. Wu, Y. Wang, Removal of 25

Cu(II) and tetracycline using an aromatic rings-functionalized chitosan-based flocculant: Enhanced interaction between the flocculant and the antibiotic, Chem. Eng. J. 283 (2016) 495-503. [34] P. Zhao, S. Ge, Z. Chen, X. Li, Study on pore characteristics of flocs and sludge dewaterability based on fractal methods (pore characteristics of flocs and sludge dewatering), Appl. Therm. Eng. 58 (2013) 217-223. [35] M. Huang, Y. Wang, J. Cai, J. Bai, H. Yang, A. Li, Preparation of dual-function starch-based flocculants for the simultaneous removal of turbidity and inhibition of Escherichia coli in water, Water Res. 98 (2016) 128-137. [36] W.L. Yan, Y.L. Wang, Y.J. Chen, Effect of conditioning by PAM polymers with different charges on the structural and characteristic evolutions of water treatment residuals, Water Res. 47 (2013) 6445-6456. [37] Q. Chen, Y. Wang, Influence of single- and dual-flocculant conditioning on the geometric morphology and internal structure of activated sludge, Powder Technol. 270 (2015) 1-9. [38] R.K. Chakraborti, J.F. Atkinson, J.E. Van Benschoten, Characterization of alum floc by image analysis, Environ. Sci. Technol. 34 (2000) 3969-3976. [39] Y. Zhou, H. Zheng, B. Gao, Y. Gu, X. Li, B. Liu, A.M. Jimenez, Waste activated sludge (WAS) dewatering properties of an original hydrophobically modified polyacrylamide containing a cationic micro-block structure, RSC Adv. 7 (2017) 28733-28745. [40] M.G. Rasteiro, F.A. Garcia, D. Hunkeler, I. Pinheiro, Evaluation of the Performance of Dual Polyelectrolyte Systems on the Re-Flocculation Ability of Calcium Carbonate Aggregates in Turbulent Environment, Polymers-Basel 8 (2016). [41] M.G. Rasteiro, I. Pinheiro, F.A.P. Garcia, P. Ferreira, D. Hunkeler, Using Light Scattering to Screen Polyelectrolytes (PEL) Performance in Flocculation, Polymers-Basel 3 (2011) 915-927. [42] X. Huang, Y. Zhao, B. Gao, S. Sun, Y. Wang, Q. Li, Q. Yue, Polyacrylamide as coagulant aid with polytitanium sulfate in humic acid-kaolin water treatment: Effect of dosage and dose method, J. Taiwan Inst. Chem. E. 64 (2016) 173-179. [43] S.Y. Yoon, Y.L. Deng, Flocculation and reflocculation of clay suspension by different polymer systems under turbulent conditions, J. Colloid Interf. Sci. 278 (2004) 139-145.

26

Figure captions: Fig. 1 Schematic experimental setup and mechanism of the copolymerization. Fig. 2 Sequence distributions of BMDAC and AM in the polymers under (a) fm=8:2, (b) fm=6:4, (c) fm=4:6 and (d) fm=2:8. Fig. 3 Apparent viscosity as a function of polymer concentration for PAB and PAD. Fig. 4 1H NMR spectra of the polymers: (a) P(BMDAC) and PAB; (b) P(DMC) and PAD. Fig. 5 Effects of the dosage on the flocculation performance and zeta potential. Fig. 6 Effect of initial wastewater pH value on the flocculation performance and zeta potential. Fig. 7 Floc size distribution of wastewater flocculated by (a) PAB, (b) PAD, (c) CCPAM and (d) P (DMDAAC) at optimal dosages. Fig. 8 Fractal dimension comparison of flocs formed in (a) PAB, (b) PAD, (c) CCPAM and (d) P (DMDAAC) flocculation at optimal dosages. Fig. 9 Effects of different flocculants on breakage and re-flocculation. Fig. 10 Possible flocculation mechanism of PAB.

1

Fig. 1 Schematic experimental setup and mechanism of the copolymerization.

Fig. 2 Sequence distributions of BMDAC and AM in the polymers under (a) fm=8:2, (b) fm=6:4, (c) fm=4:6 and (d) fm=2:8.

2

Fig. 3 Apparent viscosity as a function of polymer concentration for PAB and PAD.

Fig. 4 1H NMR spectra of the polymers: (a) P(BMDAC) and PAB; (b) P(DMC) and PAD.

3

Fig. 5 Effects of the dosage on the flocculation performance and zeta potential.

4

Fig. 6 Effect of initial wastewater pH value on the flocculation performance and zeta potential.

5

Fig. 7 Floc size distribution of wastewater flocculated by (a) PAB, (b) PAD, (c) CCPAM and (d) P (DMDAAC) at optimal dosages.

6

Fig. 8 Fractal dimension comparison of flocs formed in (a) PAB, (b) PAD, (c) CCPAM and (d) P (DMDAAC) flocculation at optimal dosages.

Fig. 9 Effects of different flocculants on breakage and re-flocculation.

7

Fig. 10 Possible flocculation mechanism of PAB.

8

Graphic abstract

Highlights 1. A novel cationic polyacrylamide with cationic micro-block structure and hydrophobic groups of benzene rings was synthesized by ultrasound initiated polymerization technique. 2. The ultrasonic-initiated copolymer showed an excellent performance in the oily wastewater treatment. 3. The formed flocs by PAB was large, compact, difficult to break, and easy to regrow because of the enhanced interaction between flocculants and oil. 4. The cooperation of charge neutralization, adsorption bridging, and hydrophobic association effect played an important role in flocculation process.

27