Novel pathway to produce high molecular weight kraft lignin–acrylic acid polymers in acidic suspension systems

Kraft lignin (KL) produced in kraft pulping process has a low molecular weight and solubility, which limits its application in industry. For the first time, KL was polymerized with acrylic acid (AA) in an acidic aqueous suspension system to produce a water soluble lignin–AA polymer with a high molecular weight in this work. The polymerization reaction was carried out using K2S2O8 as an initiator, and the influence of reaction conditions on the carboxylate group content and molecular weight of resultant lignin polymers was systematically investigated. The mechanism of polymerization of KL and AA was discussed fundamentally. The resulting lignin–AA polymer was characterized by Fourier Transform Infrared spectrophotometry (FTIR), proton nuclear magnetic resonance (1H-NMR) and elemental analyses. The results showed that the phenolic hydroxyl group (Ph-OH) content of KL promoted the polymerization under an acidic environment. Under the conditions of 1.5 wt% of initiator, 3.5 of pH, 10.0 of AA/lignin molar ratio, 0.15 mol L−1 of lignin concentration, 3 h and 80 °C, the carboxylate group content and the molecular weight of the polymer were 7.37 mmol g−1 and 7.4 × 105 g mol−1, respectively. The lignin–AA polymer was water soluble at a 10 g L−1 concentration and a pH higher than 4.5. Furthermore, the flocculation performance of lignin–AA polymer in an aluminium oxide suspension was evaluated. Compared with polyAA, the lignin–AA polymer was a more efficient flocculant for aluminium oxide suspension, which shows its potential to be used as a green flocculant in industry.


Introduction
Lignin is a natural biomacromolecule, found in wood and vascular plants.Over 60 million tons of lignin is produced in the pulp and paper industry annually in the world. 1 Among technical lignins produced, kra lignin is the most dominant one, but is mainly incinerated as a low cost fuel in the pulping industry leading to the waste of resources and growing environmental problems. 2,3With the depletion of fossil fuel and the improvement of environmental awareness, greater endeavors have been made on developing lignin-based materials.However, kra lignin has not yet been utilized effectively. 4,5ignin is a highly stable and complex compound with a three-dimensional aromatic structure formed from three phenylpropanoid monomer units of guaiacyl, syringyl and phydroxyphenyl, connected by ether and carbon-carbon bond in an irregular form. 6Various modication techniques were carried out in the past to produce new lignin-based products for benecial purposes. 7One of these modications is the polymerization of lignin with functional monomers, which can increase both the molecular weight and number of functional groups on lignin structure.Chen et al. 8 produced a polymer by reacting lignosulfonate with 1-ethenylbenzene to enhance thermal stability and molecular weight of the polymer.The polymerization of lignosulfonate with vinyl monomers, i.e., acrylic acid, acrylonitrile, and methyl methacrylate, was also studied in aqueous or organic solvents in the past via chemical radical starters [9][10][11][12] or chemo-enzymatic starters, 13 in order to produce high molecular weight products with high hydrophilic or hydrophobic properties.Meister et al. 14,15 synthesized acrylamide with kra lignin in dioxane solution and used it as a drilling mud additive.In another report, 1-phenylethylene-kra lignin polymer was produced in dimethyl sulfoxide solution and used as an oil recovery agent.However, organic solvents, such as dioxane and dimethyl sulfoxide, were generally used for facilitating the homogeneous polymerization of kra lignin and other monomers.However, these solvents are usually toxic, expensive and may need a complex recovery process aer use, which hampers their practical application in industry.
In the past, the polymerization of lignin in aqueous solutions was assessed.Ibrahim et al. 16 polymerized soda lignin with 2-acrylamido-2-methylpropane in 1 wt% NaOH solution, and the product, soda lignin polymer with a molecular weight of 2.6 Â 10 6 g mol À1 , was used as a drilling mud additive.Fang et al. 17 also produced a corn-stalk lignin-acrylamide polymer in NaOH solution and used it as an adsorbent for dye removal from wastewater.However, there is no report about the polymerization of kra lignin and acrylic acid in aqueous acidic suspensions, which is studied for the rst time in this work.
8][19] However, there are contradictory reports on the role of phenolic group in the polymerization of lignin: (1) the phenolic group acts as an inhibitor owing to the quinoid structure produced in the polymerization, which was observed in the polymerization of styrene with lignosulfonate 18 or with hydrochloric sowood lignin; 19 (2) the phenolic group acts as an active centre for the polymerization.It was observed that the conversion rates of acrylic monomers, i.e., AA, acrylonitrile and methyl methacrylate, were signicantly accelerated in the presence of lignosulfonate. 4,11Therefore, the role of phenolic group on the polymerization of KL and AA is still not clear, but is crucial for understanding the polymerization of KL and AA from academic and industrial points of views.In the polymerization reaction, the aliphatic hydroxyl groups of KL might react with carboxylate group of poly acrylic acid (PAA) formed during the lignin-AA polymerization reaction through esterication, which would also gra PAA onto lignin. 20It is not clear if the esterication reaction would happen in an acidic heterogeneous condition, which is the second objective of this study.
In the work presented herein, the polymerization of kra lignin with acrylic acid in an acidic aqueous solution was conducted using K 2 S 2 O 8 as an initiator.The main aim of this study was to generate water soluble lignin-AA polymer with a high molecular weight, which will facilitate its application as a occulant in wastewater systems.In addition, the inuence of phenolic group on the polymerization efficiency was identied.This study also intended to report how the functional groups and molecular weight of sowood kra lignin will be affected by this polymerization.The properties of the lignin-AA polymer were determined using a light laser scattering technique, and the occulation performance of the resulting lignin-AA polymer for an aluminium oxide suspension, as a model suspension system for representing wastes of the mining industry, was evaluated by a photometric dispersion analyzer.
Polymerization of KL with AA.The reactions were carried out in a nitrogen atmosphere in 250 mL three-necked asks equipped with magnetic stirrers.At rst, 2 g of lignin was suspended in 30 mL of deionized water at room temperature and 300 rpm for 20 min in three neck asks.Aer that, the required amount of AA was added to the asks and the nal pH of the suspensions was adjusted to 3.5, using 1.0 mol L À1 NaOH solution.Subsequently, the temperature of the asks was adjusted by keeping the asks in a water bath and the solutions were purged with nitrogen for 20 min.Aerwards, the predetermined amount of initiator (K 2 S 2 O 8 , wt% based on lignin's weight) was added to the asks in order to initiate the reaction at 300 rpm.A continuous supply of nitrogen was maintained throughout the reaction.The polymerization reaction was repeated under different temperatures (60 C, 70 C, 80 C, 90 C and 95 C), time intervals (0.5 h, 1 h, 2 h, 3 h, 4 h and 5 h), initiator dosages (0.5 wt%, 1.0 wt%, 1.5 wt%, 2.0 wt%, 2.5 wt% and 3.0 wt%, based on lignin's weight), AA to lignin molar ratios (1.35, 2.70, 5.4, 8.0, 10, 13.5 and 16.3) and lignin concentrations (0.07 mol L À1 , 0.1 mol L À1 , 0.15 mol L À1 , 0.22 mol L À1 and 0.38 mol L À1 ) in order to optimize the reaction conditions.
Extraction of lignin-AA polymer from reaction media.The purication of lignin-AA polymer was carried out according to the procedure developed in our earlier paper. 21The reaction solution was rstly acidied to pH of 1.5 to precipitate the lignin-AA polymer from the solution, and then centrifuged to remove the formed PAA homopolymer and unreacted AA monomer from lignin-AA polymer.The detailed procedure was stated in the ESI.
PAA preparation.PAA was prepared in the presence of KL under the conditions of pH 3.5, 0.15 mol L À1 lignin concentration, 8 AA/lignin molar ratio, 80 C, 3 h and 1.5 wt% initiator.In the absence of KL, PAA was also produced under the same conditions.Aer the reactions, the solution was treated according to the procedure used in our earlier paper 21 (available in ESI †).
Reaction of KL with PAA.In one set of experiments, the PAA, produced in AA system, was reacted with KL under the conditions of 0.15 mol L À1 of lignin, 8.0 mol mol À1 of PAA/KL, 1.5% initiator, 80 C, 3 h and pH 3.5.Aer the reaction, the solution was puried as stated above and the nal product was used for H-NMR analysis.
AA conversion analysis.The AA conversion in the reactions was determined using H-NMR.In this set of experiments, 0.2 mL of reaction solution was collected and then mixed with 0.8 mL D 2 O containing 5 mg mL À1 trimethylsilyl propanoic acid (TMSP) as an internal reference.The NMR spectra of these samples were recorded using an INOVA-500 MHz instrument (Varian, USA) with a 45 pulse and relaxation delay time of 1.0 s.The area of the peak at 5.95-6.05ppm was considered for determining the unreacted AA concentration, C 1 , in the reaction solution.The AA conversion was calculated using eqn (1) where C 0 was the initial AA concentration in reaction solution, mol L À1 and C 1 was the unreacted AA concentration in reaction solution, mol L À1 .Analysis of initiator consumption.A set of reaction was carried under the conditions of at 80 C for 3 hours in a nitrogen atmosphere in 250 mL three-necked asks, but without AA.At rst, 25 mg of lignin was suspended in 50 mL of deionized water.Aerward, the initiator was added to the asks to make an initial K 2 S 2 O 8 concentration of 0.75 g L À1 .The nal pH of the suspensions was adjusted to 3.5 or 10.5 using 1.0 mol L À1 NaOH and HCl solutions.A continuous supply of nitrogen was maintained throughout the reaction.Similar reactions were carried out only with K 2 S 2 O 8 at the same pHs (i.e., control samples).The concentrations of initiator under acidic and alkaline conditions before and aer the reactions were analysed by back titration with standardized potassium permanganate with a standardized ferrous ammonium sulfate solution. 2210 mL of persulfate containing solution was pipetted to a 100 mL Erlenmeyer ask.Then, 10 mL of 0.5 M H 2 SO 4 solution and 10 mL of 0.025 M ferrous ammonium sulfate solution were added to the ask while stirring constantly.Aer one min of stirring, the solution was titrated against 0.002 M KMnO 4 to a permanent pink endpoint.A blank titration was conducted on 10 mL of ferrous ammonium sulfate solution in 10 mL of the 0.5 M H 2 SO 4 .The concentration of persulfate in the solution was calculated based on eqn (2) where E is the volumes (L) of KMnO 4 solution used for the endpoint for blank sample.E 0 is the volumes (L) of KMnO 4 solution used for the endpoint for actual sample, M is the molarity (mol L À1 ) of KMnO 4 , V is the volume (L) of the sample and m is the molar mass of K 2 S 2 O 8 .Acetylation of lignin.In order to understand if AA can react with other reaction sites on KL (in addition to phenolic hydroxyl groups) under acidic conditions, the acetylation of lignin was carried out according to the method described by Andes et al. 22 In this set of experiments, 0.5 g of KL was dissolved in 6 mL of pyridine-acetic anhydride (1/1, v/v) by stirring for 30 min at 300 rpm, 25 C and then kept in the dark at room temperature for 72 h.The solution was added drop-wisely to 120 mL of cold water and then centrifuged and washed 3 times.The solvent was removed from the sample using a freeze dryer and the nal product was considered as acetylated lignin sample.
Methylation of KL.In order to understand the relationship between Ph-OH group content of KL and polymerization, KL was methylated according to the method reported in the literature. 23The reaction scheme is available in ESI (Fig. S1 †).In this method, only Ph-OH group of KL can be methylated, while the aliphatic hydroxyl group of KL is methylated marginally. 23A 1.0 g of KL was dissolved in 15 mL of 0.7 mol L À1 NaOH solution at room temperature by stirring at 300 rpm for 20 min.Aer that, 0.25 mmol, 0.50 mmol or 1.0 mmol of dimethyl sulphate was added per each mmol of total phenolic hydroxyl groups of KL, and the solution was stirred at room temperature for 30 min.The solution was then heated to 80 C for 2 h.During the reaction, the pH of the solution was kept at 11-11.5 by continuous addition of 0.7 mol L À1 NaOH solution.Upon the completion of reaction, the solution was acidied to pH 2.5 using 2 mol L À1 HCl solution and precipitates were washed with excess amount of deionized water until neutral pH was obtained and then the precipitates were freeze-dried.The nal product was considered as methylated KL.The methylation conditions and the content of phenolic hydroxyl group of methylated KLs were listed in Table 1.
H 2 O 2 treatment of kra lignin.In order to further testify the inuence of Ph-OH content of KL on the polymerization efficiency in the presence of carboxylate groups, hydrogen peroxide was used to treat KL to prepare the treated kra lignin samples with different Ph-OH and carboxylate group contents.In this treatment, the HOO-, which is produced by H 2 O 2 under alkaline conditions, reacts with the quinone-methide intermediates that are formed by ionization of the phenolic lignin moieties or by alkali-induced cleavage of the cyclic a-aryl ether bonds in phenolic lignin moieties. 24The quinone-methide intermediates (see Fig. S2 in ESI †) are nally oxidized into dicarboxylic acids by the cleavage of the benzene ring in KL, resulting in the reduction of Ph-OH group content in KL. 25,26 The treatment conditions and the properties of treated samples are listed in Table 2. Aer the treatment, the solution was neutralized using 1.0 mol L À1 H 2 SO 4 and dialyzed for 48 h using the aforementioned membrane dialysis.The sample collected from membrane was dried and considered as peroxide-treated KL in this study.
Polymerization of treated lignins with AA.The polymerization conditions of acetylated KL, methylated KLs and peroxidetreated KLs with AA were xed at pH 3.5, lignin concentration 0.15 mol L À1 , AA/lignin molar ratio 8.0, 80 C, 3 h and 1.5 wt% initiator.Aer polymerization, the polymer was puried according to the procedure stated earlier.The carboxylate group content was measured according to the method detailed in the following section, and the increased carboxyl group of lignin was determined (via subtracting the carboxylate group of lignin-AA polymer from the total carboxylate group of KL).
Presence of unreacted kra lignin in the polymerization.As stated previously, the unreacted KL may exist in the nal lignin-AA polymer aer the reaction.In order to evaluate this, acetone was used to extract the unreacted lignin from lignin-AA polymer samples extracted from the reaction medium in soxhlet extractor for 12 h.In this case, the unreacted lignin can be extracted from the lignin-AA polymer as kra lignin is soluble in acetone, but lignin-AA polymer is insoluble. 27,28Initially, a certain amount (about 2.0-2.5 g) of sample, M 0 , was maintained in a pre-extracted lter paper, and then the samples were extracted with acetone for 12 h.Aer the extraction, the sample was air-dried and then dried in an oven at 105 C for 12 h.The nal mass of the sample, M 1 , was weighed and the percentage of unreacted lignin was calculated using eqn (3): Functional group analysis.The carboxylate group and Ph-OH contents of lignin and lignin-AA polymers were measured using an automatic potentiometric titrator (785 DMP Titrino, Metrohm, Switzerland).About 0.06 g of dried KL or lignin-AA polymer, m, was added to 100 mL of deionized water containing 1 mL of 0.8 mol L À1 potassium hydroxide in a 250 mL beaker.Aer stirring at 200 rpm for 5 min, 4 mL of 0.5% para-hydroxybenzoic acid solution was added as an internal standard, and the solution was titrated with 0.1 mol L À1 hydrochloric acid solution.During the titration, with the decrease in the pH of the sample solutions, three endpoints appeared in sequence (V 0 1 ; V 0 2 and V 0 3 ; respectively).The corresponding three endpoints in the titration curve of blank sample were specied as V 1 , V 2 and V 3 , respectively.The carboxylate group and Ph-OH contents of samples were calculated according to eqn (4) and (5). 29The reported data in this paper is the average of three repetitions.
Phenolic hydroxyl group mmol g where C HCl is the concentration of HCl solution (0.1 mmol L À1 ) as titrant, m is the mass (g) of the sample.Molecular weight analysis.The molecular weight of samples was determined using a Gel Permeation Chromatography system, Malvern GPCmax VE2001 Module + Viscotek TDA305 with multi-detectors (UV, RI, viscometer, low angle and right-angle laser detectors).For KL measurement, the columns of PAS106M, PAS103 and PAS102.5 were used with a xed ow rate of tetrahydrofuran (THF) at 1.0 mL min À1 .For lignin-AA polymer measurement, the columns of PAA206 and PAA203 were used with a xed ow rate of 0.1 mol L À1 NaNO 3 solution at 0.70 mL min À1 .The column temperature was set at 35 C in both systems.Polystyrene polymers were used as standards for organic system and poly (ethylene oxide) for the aqueous system.Meanwhile, an attempt was made to cleave the PAA chain from lignin-AA via treating lignin-AA polymer in different acidic environments so that the molecular weight of PAA chain attached to lignin can be identied.However, this experiment was unsuccessful as the lignin part of lignin-AA polymer was also decomposed under acidic conditions prohibiting this analysis.
About 100 mg of air dried KL sample was initially suspended in 4.0 mL of acetic anhydride/pyridine 1/1 (v/v) solution by stirring for 30 min at 300 rpm and 25 C; and then the solution containing KL was kept in a dark place at 25 C for 24 h to acetylate KL.The resulting solution was then poured in an excess amount (50 mL) of ice water and centrifuged/washed 3 times.Aerwards, the solvent was removed from the samples using a freeze dryer.The acetylated KL was dissolved in 10 mL of tetrahydrofuran (THF) by stirring at 300 rpm for 30 min at room temperature, and then ltered with a PTFE lter having a diameter of 13 mm and pore size of 0.2 mm.The ltered samples were used for a molecular weight analysis.For lignin-AA polymer analysis, about 50 mg of air dried polymer samples was dissolved in 10 mL of 0.1 mol L À1 NaNO 3 solution and ltered with a 0.2 mm nylon lter (13 mm diameter).The ltered solutions were used for molecular weight analysis.
Hydrodynamic diameter of lignin-AA polymer.The hydrodynamic diameters of lignin-AA polymers were measured using a dynamic light scattering instrument (type BI-200SM Brookhaven Instruments Corp., USA).The light source was a solid state laser with a maximum power of 35 mW and a wavelength of 637 nm.The experimental procedure was adopted as described by Yan et al. 30 The lignin-AA polymer was dissolved in 1 mg mL À1 NaCl solution at pH 10.5 to make 0.2 wt% polymer concentration.The obtained solution was stirred for 30 min at 300 rpm and 25 C.Then, the solution was kept for 24 h without stirring to have a well-dissolved polymer in the solution.Aer that, the sample solution of 20 mL was ltered with a 0.45 mm disposable syringe lter prior to the size measurement.Five measurements were performed for each sample and the mean value was reported.The analysis was conducted at 25 AE 0.02 C. The scattering angle was set at 90 .Elemental analysis.The elemental analysis was performed for KL and lignin-AA polymer using Elementar Vario EL Cube Elemental Analyzer by a method described in the literature. 31he samples were rstly dried in the oven at 105 C overnight in order to remove any moisture.Approximately, 2 mg of samples were used for determining carbon, hydrogen and oxygen contents of the samples.
Fourier transform infrared (FTIR).Fourier transform infrared spectroscopy (FTIR) analysis was conducted on KL and lignin-AA polymers.The samples were rstly dried in the oven at 105 C overnight and 0.05 g of the samples was used for analysis using FTIR (Bruker Tensor 37, Germany, ATR accessory).The spectra were recorded in a transmittance mode in the range 600 cm À1 and 4000 cm À1 with 4 cm À1 resolution, and 32 scans per sample were conducted.
1 H-NMR analysis.The KL and lignin-AA polymer were analyzed with 1 H-NMR analysis.The samples of dried KL and lignin-AA polymer were dissolved into D 2 O with 10.2 pH at a 40-50 g L À1 concentration.The solution was stirred for 30 min to fully dissolve the materials.The NMR spectra of these samples were recorded using an INOVA-500 MHz instrument (Varian, USA) with a 45 pulse and relaxation delay time of 1.0 s.Solubility analysis of KL and lignin-AA polymer.The solubility of KL and lignin-AA polymer was determined based on the method described by Lappan et al. 32 About 0.5 g of KL or lignin-AA polymer was added to 50 mL of deionized water at different pHs using 1.0 mol L À1 NaOH or 1.0 mol L À1 H 2 SO 4 solution in a 125 mL Erlenmeyer ask.The suspension was immersed into a water bath shaker (Innova 3100, Brunswick Scientic, Edison, NJ, USA) and shaken (100 rpm) at 30 C for 2 h.Then, the suspension was centrifuged at 1000 rpm for 5 min.The supernatants were collected and dried at 105 C, which helped determine the solubility of KL and lignin-AA polymer in water at different pHs.To determine the solubility of KL or lignin-AA polymer, the mass of NaOH or H 2 SO 4 added for adjusting the pH was taken into account.
Zeta potential and occulation of aluminium oxide suspension.The zeta potential of aluminium oxide suspensions was analyzed (2.5 wt% at pH 6 and 8) via a zeta potential analyzer, ZetaPALS (Brookhaven Instruments Co., USA).In this study, 1 mL of the suspension was mixed with 20 mL of a 1 mM KCl solution.The zeta potential of the mixture was analyzed aer 30 seconds of mixing.All the measurements were carried out three times at room temperature.The occulation performance of lignin-AA polymer in an aluminium oxide suspension was evaluated by a photometric dispersion analyzer (PDA, PDA 3000, Rank Brothers, UK), which was connected to a dynamic drainage jar (DDJ).In this set of experiment, 450 mL deionized water with different pHs was rstly poured into the DDJ without any mesh.The system circulated water through PDA and DDJ for 10 min to reach a steady ow rate of 50 mL min À1 .Then, 50 mL of a 2.5 wt% aluminium oxide suspension at different pHs was added to the DDJ while stirring at 100 rpm.The suspension was circulated in the system continuously at a ow rate of 50 mL min À1 .Aer reaching a steady state condition, the lignin-AA polymer solution with 0.1 g L À1 concentration was added into DDJ to induce the occulation process.The degree of occulation was presented as a relative turbidity, which was calculated from the variations in the DC voltages of the PDA analyzer before and aer adding lignin-AA polymer according to the eqn (6). 33,34The reported data in this paper is the average of three repetitions.
where s i is the initial turbidity of the aluminium oxide suspension (before adding lignin-AA polymer); s f is the nal turbidity of aluminium oxide suspension (aer adding lignin-AA polymer); V 0 is initial base DC voltage (water solution); V i stands for the DC voltage of aluminium oxide suspension (without lignin-AA polymer); and V f is the DC voltage of the aluminium oxide suspension aer adding lignin-AA polymer.

Results and discussions
Reaction mechanism of polymerization of KL and AA Proposed reaction scheme.The free radical polymerization at low pH is usually suppressed, because the efficiency of homolytic scission of K 2 S 2 O 8 in acidic conditions is considered lower than that in alkaline and neutral conditions. 35To investigate the inuence of lignin on initiator decomposition, the nal concentrations of the initiator under acidic and alkaline conditions was investigated when treated at 80 C aer 3 hours with an initial K 2 S 2 O 8 concentration of 0.75 g L À1 .It was discovered that the consumption of K 2 S 2 O 8 under acidic conditions and alkaline conditions was similar in absence of lignin (i.e., concentration of unreacted K 2 S 2 O 8 was 0.40 g L À1 in acidic vs. 0.37 g L À1 in alkaline solutions).However, the unreacted K 2 S 2 O 8 was 0.22 g L À1 in acidic and 0.72 g L À1 in alkaline solution in the presence of lignin, indicating that lignin probably inhibited the consumption of K 2 S 2 O 8 under alkaline conditions.Thus, more radicals were produced at low pH by the redox systems in the presence of lignin, promoting polymerization under acidic conditions.The results in this study indicated that the mechanism of generation of free radicals was not solely attributed to the production of two sulfate radicals from the initiator.The redox system of persulfate-phenolic moieties of lignin could be also involved and affected by pH. 36The selfdecomposition of potassium persulfate in the initiator system generate sulfate radicals. 37Sulfate radicals can remove hydrogen from the hydroxyl groups on lignin and generate lignin macro radicals.Once the macromolecular radicals are generated, they react with the monomers and initiate the polymerization. 38,39he nal KL-AA polymer from acidic and alkaline systems has very different charge density (1.86 meq g À1 under alkaline vs. 7.22 meq g À1 under acidic condition) and molecular weight (0.46 Â 10 5 g mol À1 under alkaline vs. 7.4 Â 10 5 g mol À1 under acidic condition).The AA conversion of AA homopolymerization system at pH 3.5 and 10.5 was measured and presented in Fig. 1.The AA conversion of reaction at pH 3.5 reached 90.1% in 2 h, however, it was only 38.8% at pH 10.5.Also, the molecular weights of PAAs from the reaction conducted at pH 3.5 and pH 10.5 were determined to be 4.26 Â 10 5 g mol À1 and 0.83 Â 10 5 g mol À1 , respectively.This phenomenon was also observed in another work. 35One can conclude that the higher charge density and molecular weight of KL-AA polymer, prepared under acidic condition (pH 3.5), could be due to the fact that lignin accelerated the radical formation and thus more AA conversion.
In this study, the polymerization of KL and AA was carried out in an acidic aqueous suspension through heterogeneous reaction.To clarify if KL and PAA react via esterication, the PAA prepared in this study was used for investigating the reaction between KL and PAA with and without an initiator.The 1 H-NMR analysis of the products of KL and PAA reaction, which is available in ESI (Fig. S3 †), showed that no PAA was graed onto KL; illustrating that (i) the esterication reaction between carboxylate groups of PAA and aliphatic hydroxyl groups of KL did not occur in the acidic system, and that (ii) the terminated PAA formed during the polymerization reaction of KL and AA cannot be reinitiated to form KL-AA polymers.
It was reported that the polymerization of styrene and hydrochloric acid lignin did not occur on aliphatic group of lignin molecules. 19,40In our previous work, we described that the polymerization did not occur on those groups in alkaline homogenous reaction of KL and AA. 21Due to the lowest bond dissociation energies of C 6 H 5 O-H (89.8 kcal mol À1 ), and C 6 H 5 CH 2 -H (90 kcal mol À1 ) among chemical groups in KL, 41,42 the predominant lignin radicals, which are induced by the initiator radicals SO 4 2À c through the homolytic rupture of a bond in KL during acidic polymerization reaction, are phenoxyl radicals from phenolic lignin units, and with smaller probability, benzylic radicals from non-phenolic lignin units.It was also reported in the literature 42 that acetylated lignin models had signicantly lower free radicals (benzylic radicals) than the untreated lignin samples.To clarify whether the polymerization of KL and AA occurred through the benzylic radicals in the acidic conditions, the acetylated KL was used to polymerize with AA.The nal product was analyzed using 1 H-NMR (Fig. S4 in ESI †) and showed that the characteristic peaks of PAA did not exist on the product of acetylated KL and AA; demonstrating that (i) the polymerization of AA onto KL through benzylic radicals was not detectable in acidic system, (ii) in the absence of phenolic hydroxyl group, the aromatic ring, methoxyl group and aliphatic portion of lignin molecules did not participate in the reaction, and (iii) in the absence of phenolic hydroxyl group in KL, the polymerization of AA and KL was not noticed in acidic conditions.In addition to the lignin radicals formed by the initiator, the chain transfer reactions of KL and the growing PAA radicals may form lignin radicals.To testify if KL functions as a radical transfer in this system, the molecular weight of PAA formed in the absence and presence of KL was measured and shown in Fig. 2.
As presented in Fig. 2, the molecular weight of PAA in the presence of KL (111 700 g mol À1 ) was much lower than that in the absence of KL (426 300 g mol À1 ), illustrating that (i) KL functioned as a chain transfer agent in this polymerization system, and (ii) the chain transfer reaction between KL and PAA chain radicals formed some lignin radicals for this polymerization.As it is well known, the chain transfer in polymerization system not only results in a reduced molecular weight of the polymer, but may also affect the polymerization rate, which depends on the reinitiation reaction rate between chain transfer radicals and monomers.In order to understand the effect of KL on the polymerization rate, the AA conversion in KL-AA polymerization system and AA homopolymerization system was determined and shown in Fig. 3.It is apparent that KL slightly increased the AA conversion, illustrating that the reinitiation reaction rate between lignin radicals and monomers is similar or slightly higher than that of propagation reaction of AA chain radicals, which is also consistent with the ndings in the polymerization of lignosulfonate with AA in the literature. 43The reason for this behaviour can be ascribed to a greater decomposition of the initiator in the presence of KL (as discussed earlier).
Based on the analysis above, the proposed reaction scheme of KL and AA polymerization is shown in Scheme 1.As sowood  KL is known to be composed principally of coniferyl alcohol units, 44 it was chosen to present KL in this scheme.In this polymerization reaction, the sulfate radicals can initially be formed by heat decomposition (reaction (1)), which then react with phenolic hydroxyl group of KL to generate phenoxy radicals and its resonance radicals (reaction (2)).The formation of the KL radicals will lead to more decomposition in reaction (1) and more sulfate radicals.Also, the sulfate radicals can initiate AA to form AA radicals (AAc).The AA radicals then react with other AA monomers to form PAA chain radicals (reaction (3)). 9,11The KL radicals (ligninc) react with monomer (AA) to form propagated lignin-AA chain radicals (reaction (4)).In addition, the PAA chain radicals and lignin-AA radicals in the system can react with KL to form KL radicals through a chain radical transfer reaction (reaction (5)), and the KL radicals then reinitiate AA to form lignin-AA chain radicals (reaction (6)).Finally, the propagated AA chain radicals and propagated lignin-AA chain radicals react with each other to produce PAA homopolymer and lignin-AA polymer through termination reaction as shown in reactions (7).As a result, the carboxylate groups are introduced onto KL and the molecular weight of KL is increased.
Impact of Ph-OH group on the KL-AA polymerization.As shown in Scheme 1, the OH group on the phenolic structure of KL was converted to ether or carbonyl structures in the nal polymer product, which would reduce the Ph-OH content of the polymer.To clarify this, the Ph-OH contents of KL before and aer treating with initiator or polymerization reaction were measured.The Ph-OH content of KL was originally 1.73 mmol g À1 , but it decreased to 1.59 mmol g À1 aer treating with K 2 S 2 O 8 .Aer polymerization with AA, the Ph-OH content of KL was further decreased to 0.55 mmol g À1 .One can conclude that the Ph-OH content in KL declined during the polymerization, which was mainly attributed to the participation of Ph-OH in the polymerization.In another set of experiments, the impact of Ph-OH content of KL on the polymerization efficiency was determined via treating KL with hydrogen peroxide, which can reduce the Ph-OH content of lignin. 45,46The H 2 O 2 treatment conditions and the Ph-OH content of the treated lignin are listed in Table 2.As seen in Table 2, by increasing the dosage of H 2 O 2 , the temperature and time of the treatment, the Ph-OH content of the resulting KLs decreased; whereas, the carboxylate content of the treated KLs increased, which was due to the oxidation of KL by H 2 O 2 . 46,47The treated and untreated KLs were polymerized with AA, and the carboxylate group of the resulting lignin-AA polymer was measured.Fig. 4 presents the impact of Ph-OH of KL on carboxylate content of KL-AA polymer.It can be seen that the Ph-OH content of lignin has a linear relationship with the increased carboxylate group of the nal polymer, indicating that the OH group attached to the phenolic structure of KL is the reaction site for the polymerization.
To further assess the relationship between phenolic hydroxyl group of KL and increased carboxyl group of KL-AA polymer, the methylated lignins with different amounts of phenolic group were polymerized with AA.The increased carboxylate group had a linear relationship with the phenolic hydroxyl group of KL (Fig. 4), demonstrating the importance of phenolic hydroxyl group of KL on polymerization.It is also seen in this gure that, at a high amount of phenolic hydroxyl group on either peroxide-treated lignin or methylated lignin, the increased carboxylate groups for lignin-AA polymers are similar.However, with decreasing the amount of the phenolic hydroxyl group, the increased carboxylate group of the resulting polymer was more pronounced for peroxide-treated KL.
The reasons for this phenomenon might be (i) the more open structure of lignin aer peroxide treatment compared with that aer methylation, and (ii) the lower molecular weight and higher solubility of peroxide-treated KL compared with methylated lignin.In other words, the more open structure provides higher accessibility of AA to the reaction sites on KL.To clarify this, the acetylated peroxide-treated lignin was reacted with AA and the nal product was analyzed using 1 H-NMR (Fig. S5 † in ESI †).The results showed that AA was not polymerized with the acetylated peroxide-treated lignin, illustrating the fact that there were no other reaction sites on the peroxide-treated lignin.Therefore, it can be concluded that the more signicant increase in the carboxylate group of peroxide-treated lignin-AA polymer was attributed to the easier accessibility of AA to the phenolic hydroxyl group of KL, as the reaction site for the polymerization was still phenolic group of peroxide treated KL.This phenomenon was also observed by Phillips et al. 19 in the polymerization of styrene and calcium lignosulfonate.
Participation of KL in polymerization.To investigate the participation of lignin in this heterogeneous polymerization with AA, acetone was used as a solvent to extract unreacted KL (if present) from the lignin-AA polymer aer the polymerization, since KL has a high solubility in acetone but lignin-AA polymer was insoluble in acetone solution.The results of this analysis are listed in Table 3.In the absence of AA in the reaction, the percentage of unreacted lignin reached 98.8%, demonstrating that the lignin's properties were not affected either by the initiator treatment or the acidic treatment.In the absence of the initiator, the percentage of the unreacted lignin was also higher than 99%, which could be regarded as evidence that basically no reaction occurred.However, in the presence of AA, initiator and KL in the polymerization, the amount of unreacted lignin was minimal (only 0.61-1.13wt%).Furthermore, the reaction was very fast as the amount of unreacted lignin was marginal even aer 0.5 h of the reaction.One can conclude from the analysis that almost all of the KL participated in the polymerization with AA in acidic conditions.

Effects of reaction conditions
Initiator.The effect of initiator dosage on the carboxylate group content and molecular weight of lignin-AA polymer was investigated and the results are shown in Fig. 5.With the increase in the initiator dosage from 0.5 wt% to 1.5 wt%, the carboxylate group content increased from 4.0 mmol g À1 to 5.38 mmol g À1 .Further increase in the dosage marginally increased the carboxylate group content.However, the molecular weight of lignin-AA polymer decreased from 7.9 Â 10 5 g mol À1 to 4.8 Â 10 5 g mol À1 when the dosage of the initiator   increased from 0.5 to 3.0 wt%.It was well known that the lower the dosage of the initiator, the fewer the radicals (graing cites) would form on lignin, which would result in a polymer with a longer chain. 48,49The polymerization of lignin and AA can occur under two possibilities (i) a small number of long PAA segment attached to lignin in nal KL-AA polymer or (ii) a large number of short PAA segment attached to lignin in nal KL-AA polymer.In the former case, the hydrodynamic diameter of the polymer would increase due to the presence of long PAA segment on lignin.In the latter case, as a large number of short PAA segment would be graed on lignin, it may not change the hydrodynamic diameter of lignin signicantly.To clarify this, the hydrodynamic diameters of lignin-AA polymers produced were measured via using different initiator dosages and presented in Fig. 5.With the increase in the initiator dosage, the hydrodynamic diameter of the polymer decreased, illustrating the polymer produced at high initiator dosage probably had more of shorter PAA segment.It should be noted that, the hydrodynamic diameters of KL were 6.1 nm.The distributions of hydrodynamic diameter of KL and KL-AA polymer are provided in Fig. S6 in ESI † and the results depicted better uniformity for KL-AA polymer than for KL in the solution.
Reaction time.The effects of reaction time on the carboxylate group and molecular weight of lignin-AA polymer are shown in Fig. 6.It can be seen that both carboxylate group and molecular weight increased with extending the reaction time.The carboxylate group and molecular weight of lignin-AA polymer signicantly increased from 2 meq g À1 to 5.32 meq g À1 and from 1.0 Â 10 5 to 5.0 Â 10 5 g mol À1 , respectively.The increases in the carboxylate group and molecular weight could be attributed to the addition of monomers to the growing of graed chains at an extended reaction time. 50,51A/lignin molar ratio.The inuence of AA/lignin molar ratio on the carboxylate group and molecular weight of the polymer is shown in Fig. 7.It can be observed that by changing AA/lignin molar ratio from 1.4 to 10.8, the carboxylate group and molecular weight increased rapidly to 7 meq g À1 and 7.1 Â 10 5 g mol À1 , respectively.Further increase in the ratio marginally affected the charge density and molecular weight of the polymer.The increases in the carboxylate group and molecular weight showed an acceleration in the polymerization rate of lignin and AA due to an increased AA content in the reaction medium. 52However, when the ratio of AA/KL was higher than 13.5, the homopolymerization of AA to generate PAA was probably dominated that decelerated the polymerization of KL and AA (carboxylate group and M w were constant). 50,53emperature.The effects of reaction temperature on the carboxylate group and molecular weight of lignin-AA polymer are presented in Fig. 8.It can be seen that with rising the reaction temperature from 60 C to 80 C, the carboxylate group and molecular weight of lignin-AA polymer increased dramatically from 1.12 mmol g À1 and 0.4 Â 10 5 g mol À1 to 6.52 mmol g À1 and 5.4 Â 10 5 g mol À1 , respectively.The increases in the carboxylate group and molecular weight are attributed to the more effective access of AA monomer to the reaction sites on the lignin at a higher temperature, which could be due to the extended conformation of lignin molecules and the dissociation of lignin from its self-assembly of aggregates at a high temperature. 54,55In addition, the results in Fig. 8 suggests that the polymerization of KL with AA is endothermic reaction as it was promoted at a higher temperature.When the temperature was higher than 90 C, both carboxylate group and molecular   weight decreased, which was due to the fact that the higher temperature made the initiator less effective. 4,9Also, a high temperature favored the chain termination and chain transfer reactions as well as the competing homopolymerization (PAA) reaction. 10,28,43,52ignin concentration.The effects of lignin concentration on carboxylate group and molecular weight of lignin-AA polymer are shown in Fig. 9.7][58] When lignin concentration was higher than 0.15 mol L À1 , the phenolic hydroxyl radicals had more chances to interact with other lignin radicals, such as benzyl or phenolic hydroxyl radicals, by disproportionation or radical coupling reactions. 4,50Thus, the active radicals, capable of initiating the polymerization of monomers, were ineffectively consumed, leading to the declined molecular weight and carboxylate group.
To investigate the relationship between carboxylate group and molecular weight of lignin-AA polymer, the data presented in previous gures is plotted in Fig. 10.As seen, the molecular weight of the polymer increased linearly with the increase in the carboxylate group content, indicating that the increase in the molecular weight of the polymer was mainly attributed to the PAA segment in the KL-AA polymer.The linear relationship between carboxylate group content and molecular weight is Y ¼ 1.0618X À 0.9738, R 2 ¼ 0.9462, where Y is the molecular weight (Â10 5 g mol À1 ) and X is the carboxylate group of KL-AA polymer (mmol g À1 ).This formula can be used to correlate the molecular weight of lignin-AA polymer with its carboxylate group content.In addition to the molecular weight, the H y of lignin-AA polymers with different carboxylate group contents for the same samples were measured and presented in Fig. 10.The H y of lignin-AA polymers did not show a linear relationship with the carboxylate group, illustrating that even though lignin-AA with a determined molecular weight can be generated, the formed lignin-AA polymer may have different molecular conformations (coiled/linear) in solutions if they were produced under different conditions.The higher H y of samples with lower carboxylate group may imply that, when the amount of carboxylate group (and molecular weight) of polymer was low, the polymer was linear or more lignin was involved, but with high carboxylate group (and high MW) the polymer generated a coil shape conformation.
Based on the results shown in previous gures, the optimal conditions for producing lignin-AA polymer were 0.15 mol L À1 KL, AA/KL molar ratio 10.0, initiator 1.5 wt% (based on lignin mass), 80 C and 3 h.Under these conditions, the carboxylate group and molecular weight of KL-AA polymer reached 7.37 mmol g À1 and 7.4 Â 10 5 g mol À1 , respectively.This lignin-AA polymer was selected for further analysis.
Characterization of lignin-AA polymer FTIR.The FTIR spectra of KL-AA polymer prepared under the optimal conditions and KLs are presented in Fig. 11.0][61] The absorption band at 1700 cm À1 in the spectra of KL and KL-AA polymer is assigned to carbonyl groups conjugated with an aromatic ring. 62In the spectrum of KL, two absorption bands were observed at around 1266 cm À1 and 1140 cm À1 , which are assigned to C-O stretch of guaiacyl unit and C-H stretch of guaiacyl unit, respectively, illustrating that KL was a sowood lignin. 63The characteristic bands for the aromatic skeletal vibration of KL were located at around 1591, 1510 and 1425 cm À1 , respectively. 63In the spectrum of KL-AA polymer, two new strong absorption peaks appeared at 1558 cm À1 and 1406 cm À1 , which were absent in the spectrum of KL.These two peaks belong to carboxylic acid and symmetrical stretching vibrations of carboxyl anions -COO À , which illustrates the existence of PAA chain segment in the KL-AA polymer. 64,65The absorption peaks at 1510 cm À1 and 1425 cm À1 in the spectrum of KL-AA polymer, which were assigned to aromatic skeletal vibration of KL and demonstrated the existence of aromatic ring of KL, 43 was evident for successful polymerization of KL and AA.More interestingly, the results in Fig. 11 show that the relative intensity of the band at 1028 cm À1 , which belongs to non-etheried Ph-OH groups in KL-AA polymer, was weaker than that in KL, suggesting that lignin participated in the polymerization reaction through its active phenolic hydroxyl groups. 64emental analysis.The elemental analysis of KL and KL-AA polymer was carried out and the results are tabulated in Table 4.It is clear that the oxygen content of KL-AA polymer was increased from 27.04% in KL to 31.69 wt% in the polymer.Moreover, with the increase in the carboxylate group and the molecular weight, the oxygen content increased, while the carbon and hydrogen contents decreased.These changes on the carbon, hydrogen and oxygen contents of lignin in the polymer (compared to KL) were mainly because of the presence of PAA chains on lignin-AA polymer.In another report on the polymerization of amylopectin and AA, the hydrogen and oxygen contents of amylopectin were increased from 6.15% and 46.86% to 7.25% and 54.27%, respectively, but its carbon content decreased from 46.99% to 38.48%. 48 1 -NMR analysis.The 1 H-NMR spectra of KL and KL-AA polymer are shown in Fig. 12, respectively.In Fig. 12, the peak at 9.97 ppm is attributed to the carboxylate group in KL; at 9.20 ppm is attributed to the hydrogen linked to the aldehyde group; at 8.30 ppm is associated with unsubstituted phenolic protons; at 7.42-5.99ppm is attributed to aromatic protons (f on the gure).Moreover, the peak at 5.75-5.15ppm is attributed to aliphatic protons including H a and H b ; at 3.9-2.557][68] Peaks appeared at 4.5-4.9ppm are assigned to the solvent of D 2 O.In Fig. 12, it can be observed that the peaks for PAA chain segment appeared at 1.4 ppm, 2.0 ppm and Fig. 11 The FTIR spectra of KL-AA polymers and KL.2.4 ppm, respectively.Peaks appeared at 1.4 ppm is attributed to C-1 (a), at 2.0 ppm is attributed to C-2 (b), and 2.4 ppm is assigned to carboxylic acid end hydrogen of PAA (c). 28The peaks at 3.20 ppm, 2.55-3.0ppm, 5.15-5.75ppm, 5.99-7.42ppm, 8.30 ppm and 9.2 ppm belonging to KL illustrate the successful polymerization of KL and AA.In addition, the peak at 4.10 ppm is observed in the spectrum of KL-AA polymer, which is absent in that of KL and assigned to the protons of -CH 2 -(d in the gure) connected with aromatic structure through ester bond (-CH 2 -O-C 6 H 5 ). 69,70This also conrms that the Ph-OH groups are the active sites participating in the polymerization reaction, which is consistent with the FTIR results.The decrease in the peak intensity at 8.30 ppm, assigned to unsubstituted phenolic group of KL in the KL-AA polymer, demonstrated a decline in the residual Ph-OH group content of the polymer, which is consistent with the results discussed in the previous section.
Water solubility of KL-AA polymer.The solubility of KL-AA polymer and KL was presented as a function of pH in Fig. 13.As can be seen, at pH 10, the solubility of KL dropped dramatically to less than 2 g L À1 .Interestingly, KL-AA polymer was soluble under acidic conditions to pH 4, below which KL-AA polymer became insoluble.As is well known, the pK a of carboxylic acid is around 4.75, implying that the solubility of KL-AA polymer is due to the presence of carboxylate groups at a pH higher than 4.0. 71Also, it was observed that at pH 7, KL had a very low solubility (only 0.2 g L À1 ); however, KL-AA polymer had 10 g L À1 solubility, which illustrated the dramatic increase in the solubility of KL via this polymerization.Here, it should be noted that the highest concentration of KL-AA polymer water (in pH 7) can be as high as 100 g L À1 .
Flocculation performance of KL-AA polymer for aluminium oxide suspension Aluminum oxide is an industrially important oxide mineral.The occulation of aluminum oxide particles is a key step for treatment of wastewater from mining industry. 72The occulation characteristics of KL-AA polymer at different pHs were assessed in a 2.5 wt% aluminum oxide suspension, and the results are presented in Fig. 14.With an increase in the concentration, the occulation efficiency of KL-AA was enhanced, but better results were obtained for KL-AA polymer at pH 6.As reported in the literature, occulation can be promoted via charge neutralization, bridging, and hydrophobic/hydrophobic interaction. 73,74The reasons for better occulation efficiency of KL-AA polymer at pH 6 are due to the fact that the (1) surface charge of aluminum oxide particles is positive at pH 6 (zeta potential, 12.4 mV), which can be neutralized by the negative charge of KL-AA polymer and (2) the KL segment of the KL-AA polymer offers the hydrophobic/ hydrophobic interaction with aluminum oxide. 73urthermore, at a low concentration (e.g., 1 mg L À1 concentration), the occulation efficiency of PAA is higher than that of KL-AA polymer.This is attributed to the amount of charges introduced to the particles.The carboxyl group content of PAA (i.e., 10.2 mmol g À1 ) is much higher than that of KL-AA polymer (i.e., 7.4 mmol g À1 ).At the same concentration, PAA offers more negative charges than KL-AA polymer to the surface of particles, providing a more neutralization degree to aluminum oxide particles and thus a higher occulation efficiency.At 3 mg L À1 concentration for PAA and 4 mg L À1 concentration for KL-AA polymer, the amount of charges introduced to particles are similar, i.e., 29.5 mmol L À1 with KL-AA polymer vs. 30.6mmol L À1 with PAA, inducing a similar turbidity for aluminum oxide suspensions.At higher dosages, a higher occulation efficiency was achieved for KL-AA polymer than for PAA indicating that, in addition to the PAA segment in KL-AA polymer, the lignin segment in KL-AA polymer contributed to the occulation (i.e., a more important role of KL component than PAA charge).Furthermore, the higher molecular weight (7.4 Â 10 5 g mol À1 ) and higher H y (25.2 nm) of KL-AA than those of PAA (4.26 Â 10 5 g mol À1 and 7.2 nm) might have played roles in occulation at the optimum dosage.

Conclusions
The polymerization mechanism of KL and AA under acidic conditions was comprehensively studied in this work.It was discovered that KL promoted the decomposition of the initiator under acidic conditions.Also, more PAA chain would be graed on KL under acidic conditions than alkaline.The results suggested that the phenolic hydroxyl group content of KL had a signicant inuence on the AA polymerization.The optimal conditions for the polymerization were 0.15 mol L À1 KL, AA/KL ratio of 10.0 mol mol À1 , 1.5 wt% initiator, 80 C and 3 h.Under the optimized conditions, the carboxylate group content and the molecular weight of the KL-AA polymer were 7.22 meq g À1 and 7.4 Â 10 5 g mol À1 , respectively.The FTIR, H-NMR and elemental analyses conrmed the successful polymerization of KL and AA.Additionally, the resulting KL-AA polymer was water soluble at a pH higher than 4.5 and its maximum solubility was 100 g L À1 .Compared with PAA, the KL-AA polymer was a more efficient occulant for aluminum oxide suspensions at pH 6.

Fig. 1
Fig. 1 AA conversion in to produce PAA at pH 3.5 and 10.5 as function of reaction time.

Fig. 2
Fig.2The distribution of molecular weight of PAA from KL-AA in the presence and absence of KL.

Fig. 3
Fig.3AA conversion in KL-AA system and AA system as function of reaction time.

Fig. 4
Fig. 4 Relationship between Ph-OH group of KL and increased carboxylate group of KL-AA polymer.

Fig. 10
Fig. 10 Relationship between carboxylate group content of lignin-AA polymer with its molecular weight (based on 1.5 wt% initiator).

Fig. 12 H
Fig. 12 H-NMR spectra of KL and KL-AA polymer.

Fig. 13
Fig.13The solubility of KL-AA polymer and KL as a function of pH at 10 g L À1 concentration.

Fig. 14
Fig. 14 Flocculation performance of KL-AA polymer at different concentrations and pHs for aluminium oxide suspension.

Table 1
Methylation conditions and Ph-OH content of methylated lignin 12324 | RSC Adv., 2018, 8, 12322-12336 This journal is © The Royal Society of Chemistry 2018

Table 2
Hydrogen peroxide treatment conditions and Ph-OH content of peroxide-treated lignin

Table 3
Reaction conditions and percentage of unreacted KL a