High-yield grafting of carboxylated polymers to wood pulp fibers

Poly(ethylene-alt-maleic acid), PEMAc, is a linear polymer that, along with its isomer polyacrylic acid, has the highest carboxylic acid content of any polymer. The goal of this work was to elucidate the mechanisms that control the amount of PEMAc that is permanently fixed on pulp fibers after the impregnation of dry pulp with a dilute PEMAc solution followed by drying/heating (curing). Two mechanisms by which PEMAc is fixed to cellulose fibers were discovered, chemical, and physical fixation. With room temperature drying only physical fixation is operative. Evidence supports the explanation that physical fixation is a consequence of the slow swelling and dissolution of thick dried PEMAc deposits on fiber surfaces. Chemical fixation includes grafting to cellulose plus enhanced cohesion within thick PEMAc layers, possibly due to interchain crosslinking. The pH of the PEMAc impregnation solution determines the fixation mechanism for curing temperatures above 100 °C. Physical fixation dominates when pH > 8 whereas chemical fixation dominates for impregnation pH values ≤ 7, suggesting the curing reactions require partial or complete protonation of the succinic acid moieties. The yield of impregnated polymer fixed to the fibers after washing depends upon the fixation mechanism. When chemical fixation dominates, the yields for low and high molecular weight PEMAc doses less than 0.1 meq/g (6.3 kg PEMA/tonne dry pulp) were close to or equaled 100%. By contrast, when the primary mechanism is physical fixation, yields are ~50% for high molecular weight PEMAc and 0–20% for low MW PEMAc. These results show that high PEMAc fixation yields can be achieved under curing conditions that could be implemented in pulp drying machines producing dry market pulp.


Introduction
Although softwood kraft pulps are known to be one of the strongest wood pulps for paper products, the literature contains many examples of efforts to improve the mechanical properties of pulp by increasing the density of charged groups on ber surfaces. Examples include cellulose oxidation, (Kitaoka et al. 1999) (Fras et al. 2005) carboxymethylation, (Wågberg et al. 1987) (Laine et al. 2003) (Barzyk et al. 1997) growing charged polymers from cellulose (Lepoutre et al. 1973) (Vitta et al. 1985) and attaching (grafting) charged polymers to cellulose. (Belgacem and Gandini 2005) (Laine et al. 2000) (Laine et al. 2002) Surface charge enhancement can give stronger ber/ ber joints, increased ion-exchange capacities, increased water absorbency, and increased functional groups for subsequent surface modi cation. Despite these potential advantages, kraft market pulps with enhanced surface properties are not widely marketed because bleached cellulose bers are barren, unreactive surfaces that are di cult to chemically modify under the aqueous conditions in a pulp mill.
The overall goal of our work has been to identify high charge-density water-borne polymers that can be xed to kraft pulp bers in the pulp mill without catalysts, oxidants, monomers, or other low molecular weight reagents that present environmental challenges in process e uents. Based on the work of others (Rätzsch 1988) (Johnson 2010) (Pompe et al. 2003), followed by our initial screening study, (Zhang et al.) we identi ed maleic anhydride copolymers as a promising approach. Succinic anhydride moieties on the copolymers can covalently couple to cellulose by forming ester linkages. The grafting chemistry is shown in Figure 1 for the case of poly(ethylene-alt-maleic anhydride), PEMA. Based on Charles Yang's publications on maleic anhydride copolymer grafting to cotton (Chen et al. 2005;Gu and Yang 1998;Yang 1993;Yang and Wang 1996a;Yang and Wang 1996b) and paper,   (Yang et al. 1996) (Yang and Xu 1998)  ) the rst step in grafting is likely to be the formation of succinic anhydride moieties. In a second step the anhydrides react with cellulosic alcohols to form ester linkages.
Our initial survey identi ed PEMA as the most promising of seven commercial maleic anhydride copolymers. (Zhang et al.) PEMA is not water-soluble; yet, it is easily hydrolyzed to poly(ethylene-altmaleic acid), PEMAc, which is very water-soluble owing to the high content of ionizable carboxyl groups. PEMAc does not adsorb onto bleached pulp bers in water. Therefore, we impregnated dry pulp bers, in the form of paper handsheets, with PEMAc solutions followed by drying and heating (curing). However, any movement towards the commercial application of PEMAc pulp treatments includes satisfying two technical requirements. First, the dried treated pulp must have su ciently low wet strength to facilitate redispersion in water, particularly when the pulp is sold in the form of dense bales. In other words, the dried pulp must be repulpable. Second, when repulped, most of the added polymer must remain xed to the ber surfaces and should not leach into water; the xation yield must be high. This paper reports the results of a systematic study of factors in uencing the xation yield for PEMAc on bleached northern softwood pulp bers. We demonstrate that an additional xation mechanism is operating in parallel to that shown in Figure 1. The issue of wet strength and repulpability will be addressed in our next contribution.
PEMA Hydrolysis to PEMAc. In a typical hydrolysis experiment, PEMA (1 g) powder was dispersed in 49 g 1 mM NaCl solution at neutral pH with constant stirring at room temperature. After 2 days, the PEMAc solution was clear and the solution pH dropped to 2.3.
PEMAc Quanti cation. Conductometric titration was used to measure PEMAc concentrations in solutions and on pulp bers. To a wet pulp sample (dry mass 0.2 g) was added 90 mL, 4 mM NaCl solution. The initial pH was adjusted to < 3.0 by adding 1 M HCl. 0.1 M NaOH solution was added at a rate of 0.05 mL/min up to pH 11.5 by using an auto titrator (MANTECH,Benchtop Titrator Model,. Titrations were repeated with fresh samples at least three times. The volume of base consumed by the weak carboxyl groups was determined by the points of intersection of three trendlines going through the linear sections of the titration curve. PEMAc Gel Content The pH values of PEMAc solutions (0.1 wt%) were adjusted to 2, 4, or 8 by adding NaOH or HCl, and then were ltered using a 60 mL syringe tted with a 0.2 μm Supor syringe lter. The PEMAc concentrations before and after titration were used to estimate the gel fraction.
PEMAc to PEMA by Isothermal Thermogravimetric Analysis (ITGA). The Mettler TGA/DSC 3+ thermogravimetric analyzer was used for PEMAc kinetic dehydration analysis. The PEMAc samples were prepared by freeze-drying hydrolyzed copolymer solution whose pH value was adjusted to 2, 4, or 8.
About 5 mg PEMAc powder in 70 μL aluminum crucible was loaded on the TGA instrument, argon gas at a rate of 10 mL/min was allowed to ow over the sample during the measurement. The samples were heated at 80 °C for 30 min to evaporate free water, then the temperature was raised to the nal temperature of 120 or 150 °C with a rate of 10 °C/min, holding this isotherm for 3 h. The conversion was based on the weight loss assumed to be one water molecule per anhydride group formed.
Pulp Ion Exchange. Most results were obtained with as-received never-dried pulp. A few pulp samples underwent the following treatments to change the counterions before drying and polymer impregnation. To protonate the ionizable groups, 5.4 g wet pulp (solids content 28 wt%) was diluted in 600 mL, 0.01 M HCl for 30 min with constant stirring (3 cm diameter propeller, 500 RPM). The acidi ed pulp was ltered on a 15 cm Büchner funnel tted with Whatman® 5 qualitative lter paper. Excess acid was removed by dispersion in 600 mL deionized water in a 1 L beaker. After 10 min stirring, the pulp was ltered. Identical second and third washes were performed.
The acidi ed, washed pulps were converted to the calcium form by dispersion in 0.05 M CaCl 2 at pH 8 for 30 min with constant propeller stirring. Excess calcium and chloride were removed by the same washing procedure used to clean the acidi ed pulp, yielding the Ca-pulp.
In other experiments, chelation was used to sequester calcium ions in the never-dried pulp. Never-dried pulp, 5.4 g of wet pulp (28 wt% solid content) was added to 600 mL, 0.01 M EDTA at pH 4.5 for 2 h with constant stirring (3 cm diameter propeller, 500 RPM) to remove adsorbed metal ions, including Ca 2+ , from the pulp. The pulps were then washed three times with 600 mL deionized water as described above.
PEMAc Adsorption on Pulp Fibers. To 200 mL pulp suspension (1 g, dry mass) in 1 mM NaCl, 1 mL PEMAc (20 g/L) solution was added, with the pH set to 2. PEMAc was allowed to adsorb onto the pulp for 4 h with constant stirring (3 cm diameter propeller, 500 RPM). The adsorbed pulp was ltered on a 6 cm Büchner funnel tted with Whatman® qualitative lter paper, Grade 5. The charge content of the bers was later measured by conductometric titration.
Handsheet Preparation. Pulp sheets for polymer impregnation (75 g/m 2 ) were prepared with never-dried bleached pulp (15 g, dry mass), which was diluted to 2 L with deionized water and disintegrated in a British disintegrator (Labtech Instruments Inc., model 500-1) for 15,000 revolutions. 200 mL of 0.75% pulp was added to a semiautomatic sheet former (Labtech Instruments Inc., model 300-1) where the pulp was further diluted to 0.019% with deionized water before dewatering. Wet handsheets were pressed (Standard Auto CH Benchtop Press, Carver, Inc., US) between blotter pads with a pressure of 635 kPa for 5 minutes at room temperature. The pressed sheets were then placed in drying rings to dry overnight at 50% relative humidity and 23 o C.
Pulp Impregnation. Using the base case as an example, 3 mL PEMAc solution (2 wt%, the mass fraction of the parent PEMA solution) at pH 8 was added dropwise across the surface of a dry pulp sheet (~ 1.5 g). The wet pulp sheet was then placed between two blotter papers and rolled with two passes using a TAPPI-standard brass couch roller (102 mm diameter and 13 kg mass) to remove excess polymer solution. The pressed pulp sheet was weighed before impregnation and after pressing to facilitate calculating the mass of the applied polymer. Typically, the applied polymer corresponded to 2.3 mL with the remaining 0.7 mL transferring to the blotters during pressing.
The impregnated sheet was cured between two new blotting papers on a speed dryer (Labtech Instruments Inc.) at 120 °C for 10 min. In some experiments, the impregnated sheets were dried at 50 % relative humidity and 23 °C overnight.
Washing for Fixation Yield. The amount of polymer that could be washed off the pulps was measured to estimate the quantity of polymer remaining xed to the bers. Speci cally, a pulp sheet was torn into small pieces that were added to 200 mL of 1 mM NaCl at neutral pH, in a 250 mL beaker. After stirring for 30 min with a magnetic stirring bar, the pulp was ltered to separate the bers. The washing procedure was repeated. The polymer contents of the washing solutions were measured to calculate the PEMAc xation yield based on the wash solution. The PEMAc content of the bers was also directly measured by conductometric titration. In cases where the wet strength was high, the pulp sheets were repulped using a NutriBullet Baby Bullet blender. For the second wash, the damp pulp was dispersed in 600 mL of 1 mM NaCl in a 1 L beaker. The pulp suspension was stirred with a propeller stirrer for 10 min at 500 RPM and ltered again. The third wash was a repeat of the second.
Repulping and Washing for Papermaking. Treated pulps for papermaking were prepared from the impregnated and cured pulp sheets. First, a 1.5 g treated pulp sheet was torn into small pieces and added to 2 L of 1 mM NaCl, and dispersed with 15,000 revolutions. Next, the dispersed pulps were ltered on a 15 cm Büchner funnel tted with Whatman® qualitative lter paper, Grade 5. For the second wash, the damp pulp was dispersed in 600 mL of 1 mM NaCl in a 1 L beaker. The pulp suspension was stirred with a propeller stirrer for 10 min at 500 RPM and ltered again. The third wash was a repeat of the second.
Polyelectrolyte Titration The quantity of xed PEMAc on exterior ber surfaces was measured by polyelectrolyte titration. Approximately 0.1 g dry mass of wet, washed PEMAc grafted pulp were added to 40 mL PDADMAC (1.18 meq/L) in 1mM NaCl. The suspension was mixed with a magnetic stirring bar for 30 min at pH 10 to facilitate PDADMAC adsorption. The suspensions were then ltered on a 4.7 cm Büchner funnel tted with Whatman® 5 qualitative lter paper. The unadsorbed PDADMAC concentration in the ltrate was determined by titration with PVSK (1 meq/L). The endpoint was determined with a Mütek PCD-03.
Fourier-Transform Infrared spectroscopy (FTIR). FTIR spectra of uncured and cured impregnated pulp sheets were recorded on a Thermo Nicolet 6700 FTIR spectrometer in ATR mode.
Grafted Pulp Hydrolytic Stability. The grafted pulp (1 g dry mass) after washing was suspended in 400 mL 0.1 M pH 7 PBS buffer with constant stirring using a stir bar at room temperature. Pulp samples (90 mL) were removed at various hydrolysis times. The pulp samples were ltered on a 6 cm Büchner funnel tted with Whatman® qualitative lter paper, Grade 5. The damp pulp was titrated using the protocol described above. This procedure was repeated at 0.1 M pH 4 acetate buffer and 0.1 M pH 10 carbonatebicarbonate buffer. The hydrolysis experiment of grafted pulp was also carried out in 0.1 M NaOH at 70°C using the same method described above.
Water Retention Value (WRV). The grafted pulp after washing was suspended in 1 mM NaCl solution at pH 7 and a testing pad consisting of grafted pulp (~ 0.8 g, dry mass) was formed by dewatering on a 15 cm Whatman® qualitative lter paper, Grade 5, which was then placed in the holding unit. Centrifugation (Allegra 64R Series Refrigerated Microcentrifuges, Beckman Coulter) was performed at 4095 RPM (3000 g) for 15 min at a constant temperature of 23 °C. Immediately after centrifugation was stopped, the test pad was weighed and dried in an oven at a temperature of 105 ± 2 °C overnight. The WRV was calculated from the wet mass of the pulp sample after centrifugation and the corresponding dry mass of the sample.
Polymer Distribution on Fibers. 1 mL 6-amino uorescein solution (2 mg/mL in acetone) was added to a solution of PEMA (2 g) in 20 mL acetone. The reaction mixture was stirred overnight at room temperature. This was followed by dialysis and hydrolysis in 1 mM NaHCO 3 for two weeks. The nal product (6-Amino uorescein grafted PEMAc, A-PEMAc) was obtained by freeze-drying. A-PEMAc solution (2 wt% in 1 mM NaCl, pH 8) was used to impregnate pulp sheets, grafting 10 min at 120 °C. The A-PEMAc grafted pulp sheet was conditioned overnight at 50 % relative humidity and 23 °C before characterizing using a confocal laser scanning microscope (Nikon A1).

Results And Discussion
PEMAc Solutions. PEMAc was obtained by hydrolyzing PEMA (see structures in Figure 1), a waterinsoluble, alternating copolymer of ethylene and maleic anhydride. Two molecular weight PEMA/PEMAc samples were employed and their properties are summarized in Table 1. H-PEMA was supplied by Sigma and L-PEMA was Vertellus E60. The molecular weight distributions for PEMAc are broad, a consequence of the alternating polymerization mechanism. (Rätzsch 1988) Although PEMAc appears to be water-soluble over the 2-8 pH range, H-PEMAc showed a signi cant gel fraction at pH 2 and 4 -see Table 1. L-PEMAc had a much lower gel fraction. Johnson's thesis reports Mw 71 kDa and Mw/Mn of 1.8 for ZeMac™ E60 (L-PEMA); Mw 610 kDa, Mw/Mn of 2.8 for E400 both manufactured by Vertellus. PEMAc, along with its isomer of polyacrylic acid, has the highest carboxyl contents of any polymer (i.e., 72 meq/g for the protonated polymers). According to the titration results in the literature, (Bianchi et al. 1970) the degree of PEMAc ionization is 0.37 at pH 4 and 0.9 at pH 8.2. Figure S 1 in the supporting information shows the ionization behavior over the pH range and a t using a model based on 2 pKa values. PEMAc's high carboxyl contents enabled accurate measurements of PEMAc concentrations in solution and the contents of PEMAc grafted to pulp bers, using conductometric titration. Table S 2, in the supporting information, shows results from three replicate titration results. The standard deviation was 4% of the mean for grafted pulp whereas the standard deviation rose to 19% when titrating untreated bleached pulp with very low charge contents.
PEMAc Adsorption onto Bleached Kraft Pulp. A suspension of pulp bers (0.5 wt%) and PEMAc (2 wt%) at pH 2 or 3 were mixed at room temperature for 4h. After which the pulps were washed three times with 1 mM NaCl at pH 2 or 3. In all cases, titration of the bers after washing did not indicate adsorbed PEMAc. These observations were expected; however, they serve to emphasize that polymer adsorption from dilute polymer solutions is not a viable approach to xing large amounts of PEMAc onto wood pulp bers. Therefore, the impregnation of dry pulp sheets was used to prepare the grafted pulps.

Impregnation Results
Impregnation of Dried Pulp with PEMAc Solutions. Sheets of dried pulp with a target dry mass of 1.5 g and a basis weight (mass/projected area) of 75 g/m 2 , were treated with 3 to 3.5 mL PEMAc solution. The pulp sheets were dried and heated to promote polymer grafting to the ber surfaces. After heating, the pulp sheets were repulped and extensively washed. Our initial hypothesis was that the only mechanism for the xation of PEMAc to pulp bers was covalent grafting. However, the following results will show that at least two xation mechanisms are operative.
Table S 1 in the supporting information summarizes many of the impregnation experimental conditions and pulp properties. The mechanical strength data will be addressed in a subsequent manuscript. The quantity of PEMAc applied to the pulp sheet (i.e., the Dose in Table S 1), was calculated from the mass of polymer solution in the sheet before drying. Conductometric titrations were used to measure PEMAc concentrations in solutions and xed to bers. The quantities of PEMA/PEMAc applied to the pulp and the quantities of PEMAc xed to the bers is expressed as equivalents of carboxyl groups per gram of dry ber (COOH, meq/g). Note, to obtain the mass of applied PEMA per tonne of dried ber, we multiply meq/g by the carbonyl equivalent weight of PEMA which is 63.05 Da.
The maximum dose of applied PEMA was 2.6 meq/g (Row 39 in Table S 1) which corresponds to 164 kg of PEMA per tonne of dry pulp, a very high loading. Most of the high dose conditions are 0.45 meq/g (28 kg/tonne). However, for papermaking applications where to goal is a cost-effective surface coverage of carboxyl groups, the applied dosages are more likely to be < 0.05 meq/g (< 3.15 kg/tonne). Finally, if we assume the speci c surface area of the pulp, accessible to high molecular weight polymers, is 1 m 2 /g and a monolayer coverage on the ber surfaces is 1 mg/m 2 , the required polymer dose is 1 kg/tonne or 0.016 meq/g. In summary, 0.45 meq/g is a very high load and roughly corresponds to 28 monolayers of PEMAc uniformly distributed on exterior ber surface. Applying the same assumptions, 0.05 meq/g, our proposed upper limit for practical dosage, corresponds to 5 monolayers, and 1 monolayer corresponds to 1 kg/tonne or 0.016 meq/g.
The PEMAc contents xed to bers were measured by two methods: 1) conductometric titration of the pulp ber suspension after washing ("Fiber Titration"); and/or, 2) conductometric titration of the wash water solution ("Wash Titration"). Direct ber titration is commonly used (Katz and Scallan 1984;Wågberg and Annergren 1997) and should give the most accurate results. However, pulp sheets with very high wet strength were di cult to repulp and titrate; in these cases, we used the wash titration.
The PEMAc xation yield is de ned as the amount of polymer remaining on the pulp after impregnation, curing, and washing, divided by the dose of the applied polymer. Yield is important because a low yield means un xed PEMAc is released to the aqueous phase during repulping. Yield calculations are based on the xed PEMAc contents that were measured either by ber titration or wash titration. Figure S 2 (in supporting information) shows the yields from wash water titration plotted against the corresponding direct ber titration for the cases in Table S 1 where both yields were measured. Yields based on wash water estimates were up to 20 % higher than those from the direct ber titrations. We believe the ber titrations to be more accurate because the mass balance used in the wash water analysis does not account for losses from the wash solution due to the migration of polymers to the blotters during drying and curing, material that would be considered in this analysis as bound to the bers. The following sections illustrate the roles of the most important experimental parameters in the PEMAc ber treatment process: the pH of the PEMAc impregnation solution; the curing time and temperature; the amount of PEMAc applied to the pulp; and the PEMAc molecular weight.
Impact of PEMAc Solution pH on Fixation. PEMAc is a polyelectrolyte whose properties are a function of pH. The relative concentrations of each of the three dissociation states of PEMAc are shown as three curves in Figure 2. Also shown in Figure 2 are the carboxyl contents of the washed, grafted pulps as a function of the pH of the PEMAc impregnation solution. The highest PEMAc contents were impregnated at pH 2-5, however, even at pH 11, the charge content was 0.17 meq/g, three times greater than the untreated pulp. We see that the highest contents of grafted PEMAc correspond to the low pH values where PEMAc is partially or completely protonated. At pH 11, PEMAc was entirely ionized in the impregnation solution and the PEMAc content is the lowest but was still substantial. We anticipated no grafted PEMAc when the pH of the impregnation solution was 11 because anhydride formation during curing seems unlikely for the fully ionized form of PEMAc. (Higuchi et al. 1963) Contrary to our expectations, there was signi cant xation with PEMAc solutions having pH values in the range 8-11. Therefore, we propose that two PEMAc xation mechanisms are operative depending on the pH of the impregnation solution. Chemical xation is dominating at pH < 7 and physical xation dominating for pH > 8. We believe that curing reactions, shown in Figure 1, occurs at elevated temperatures after most of the water has evaporated. It is remarkable that the pH of the solution, before drying and heating has such a large in uence. A more detailed description of these mechanisms is presented after summarizing the results. Figure 3 shows the xation yield dependencies on the pH of the impregnation solution, the drying/curing temperature, and PEMAc molecular weight. Figure 3A shows the impregnation yield as a function of impregnation solution pH and the PEMAc molecular weight for pulp sheets cured at 23 o C. At this temperature, no chemical conversion of succinic acid moieties to succinic anhydrides is expected. Physical xation is the only operative mechanism. The yield for H-PEMAc (100-500 kDa) was about 50% from pH 2 -11 with a peak of about 70 at pH 4 where PEMAc had a gel content of 16% (Table 1). The yields for L-PEMAc were much less; the role of molecular weight is discussed below in the section titled Explaining Physical Fixation.
The corresponding yields for pulps cured at 120 o C are shown in Figure 3B. Note that the high yield samples could not be repulped for titration, so the yields were based on wash water measurements. When the impregnation solution is acidic, the yields are high and independent of PEMAc molecular weight suggesting chemical curing. On the other hand, with basic solutions the H-PEMAc yield levels at ~ 0.4 due to physical xation, whereas no L-PEMAc remained on the washed pulp.
The In uence of Curing Time and Temperature. Figure 4 shows the in uence of curing time, curing temperature, and PEMAc molecular weight, on the PEMAc content of washed bers. With pH 8 impregnation, extended curing times or temperatures slowly increased the xed H-PEMAc content from the physical xation limit determined by the room temperature curing. In contrast, with pH 4 impregnation, most of the added polymer was xed after 10 min curing at 120 o C; therefore, increasing the curing time or temperature had little impact. L-PEMAc gave much lower polymer contents than did H-PEMAc. Physical xation was far less effective with L-PEMAc.
Varying the Quantity of Applied PEMAc. The concentration of PEMAc in the impregnation solution was varied to yield a series of pulps spanning a range of PEMAc contents. Figure 5A shows the in uence of applied H-PEMAc content on the charge contents of the bers after curing and washing to remove un xed polymer. For pH 8 impregnation, which is dominated by physical xation, the immobilized PEMAc content approaches a constant value of 0.35 meq/g when the added PEMAc impregnation concentration exceeds 0.5 meq/g. The numbers beside the data points give the corresponding yields (i.e. polymer content on washed bers/polymer content added). The pH 8 yields are low with 0.55 being the highest. In contrast, impregnation with pH 4 PEMAc solution gave high yields up to the highest applied PEMAc (R46 in Table  S 1, 2.12 meq/g, yield 0.76, ber titration). Figure 5B shows the xation yields as a function of the dosage of applied PEMAc for pH 4 treatments. The vertical columns of points correspond to series of experiments where the applied dosage was constant whereas the curing temperatures, T, and/or curing times, t, were varied. Yields at or near 1 (i.e., 100%) were achieved for low and high molecular weight PEMAc doses less than 0.1 meq/g (6.3 kg PEMA/tonne dry pulp). For very high dosages of 0.3 meq/g and higher, the maximum achievable xation yields decreased.
Impregnation at pH 8 resulted in much lower yields. All 23 H-PEMAc treatment conditions at pH 8 in Table   S 1 gave xation yields between 0.11 and 0.55, values too low for commercial application. Therefore impregnation at pH 4 looks to be a promising approach for ber treatment technology whereas the yields at pH 8 were too low. The following sections explore more deeply the nature of physical and chemical xation.
Polyacrylic Acid (PAA) vs PEMAc. The comparison of PAA with PEMAc gives further evidence for physical xation. Whereas upon heating, dry PEMAc readily forms 5-member anhydride groups (see Figure 1), the neighboring carboxyl groups on PAA are much less likely to form anhydride rings or to graft to cellulose under the catalyst-free, mild curing conditions. (Yang and Wang 1996b) (Greenberg and Kamel 1977) In other words PAA is unlikely to form covalent ester linkages to cellulose or form covalent crosslinks at the curing temperatures employed in this study. On the other hand, PAA like PEMAc does exhibit physical xation. Pulp was impregnated with PAA (450 kDa) solution and pH 8 followed by curing at 120 o C for 10 min. The yield, based on ber titration, was 0.28, about half the yield for the PEMAc under similar conditions. The PAA yields were not very sensitive to drying/curing temperatures between 70 and 120 o C -see Table S 3. The similar behaviors of PEMAc and PAA is further evidence that polymer xation to bers at pH 8 was due to physical interactions and not chemical grafting.

Fixation Mechanisms
Explaining Physical Fixation. In the 70's, Allan showed that when polyethyleneimine or proteins were impregnated into wood pulp bers under pH conditions where the chains were collapsed, the polymers remained xed when bers were eluted (washed) under pH conditions where polymer chains would expand and become physically entrapped inside ber pores. (Allan and Reif 1971;Allan et al. 1970) In our experiments, PEMAc chains are collapsed at low pH and washing was performed at neutral pH. Perhaps this mechanism contributes to the peak xation yield at pH 4 for H-PEMAc cured at room temperature in Figure 4A. However, this mechanism cannot account for physical xation of H-PEMAc at high pH.
We postulated that calcium ions present in the wet pulp could form ionic crosslinks with the impregnated PEMAc, contributing to physical xation. Our kraft pulp was supplied as a wet slurry that had never been dried. Normally pulp samples were diluted and made into handsheets for impregnation. To evaluate the potential role of calcium in physical xation, calcium-free pulp, and calcium-saturated pulp sheets were prepared and impregnated with H-PEMAc at pH 8. The yields based on titration of the washed bers were 0.44 for the calcium-free pulp sheet, 0.50 for the calcium-saturated sheet, values that were close to 0.52 for untreated pulp. Therefore, we concluded that calcium played no role in physical xation.
We propose that physical xation is a result of dried deposits of H-PEMAc that are very slow to dissolve when repulped. The physics of polymer dissolution is summarized in some good reviews. (Lee and Peppas 1987;Miller-Chou and Koenig 2003) There seems to be general agreement on the dissolution phenomena which, in the case of PEMAc, would be as follows. Consider a thick layer of dried PEMAc pH 8 solution on a at surface. Based on the ionization curves in Figure 2, most of the PEMAc is present as a sodium salt dried to form a glass. Upon immersion, water will diffuse into the dry polymer, slowly converting the glassy polymer at the polymer/water interface into a swollen hydrogel. In the case of PEMAc gel, swelling is promoted by both the hydrophilicity of the polymer and Donnan swelling pressure due to the carboxylic salts. With time, the hydrogel layer thickness grows and the glassy polymer layer decreases. For chains to be released into the solution from the hydrogel, the gel must swell enough to give the polymer chains su cient mobility to untangle. The untangling process and thus the dissolution rate decreases with increasing polymer molecular weight. (Ueberreiter and Asmussen 1957) (Cooper et al. 1985) Obviously the thicker the initial glassy polymer lm, the longer the time for total dissolution. An implication of the "slow dissolution" explanation for physical xation involves the xation yield. If all the impregnated PEMAc were present as large, slowly dissolving deposits, the xation yield would be very high if only a thin layer would dissolve during pulp washing. Similarly, most of a uniformly deposited PEMAc thin lm could be removed by washing. In our experiments, the yields of H-PEMAc xation were rarely above 50% for high pH impregnation and the yields were very low for the fast-dissolving L-PEMAc.
In summary, the evidence for physical xation being due to slow dissolution is: The impregnated pulp sheet must be dried to at least 8% water for physical xation. During drying, capillary forces can drive the non-uniform accumulation of polymer in the ber mat. With complete drying, the polymer hydrogel is converted to a glassy polymer that is slower to dissolve. Without drying the xation yield was only 0.09, Row 72 in Table S 1 Low molecular weight L-PEMAc displayed very little physical xation, re ecting a reduced contribution of PEMAc chain entanglement.
Impregnation with low H-PEMAc concentrations gave low yields (yield 0.19 in Figure 5) suggesting a larger fraction of thinner deposits dissolve in a xed washing time.
Explaining Chemical Fixation. We propose that chemical xation involves heat activated changes in the composition of PEMAc under acidic conditions. Figure 1 shows the mechanism for the grafting of PEMAc to cellulose. (Yang and Wang 1996a)  The grafting reaction scheme in Figure 1 cannot explain entirely xation by chemical curing. As mentioned above, most of the PEMAc in our experiments was not in physical contact with cellulose. Instead, most of the PEMAc was present in multilayer deposits and chemical xation involves some form of heat-activated attraction between contacting PEMA/PEMAc chains. What are these cohesive interactions? The pH 4 results in Figure 3B do not show PEMAc molecular weight sensitivity suggesting PEMAc chain entanglement is not a major factor at pH 4. Two possible explanations for intermolecular PEMAc cohesion are: 1) The formation of covalent anhydride intermolecular crosslinks between PEMA chains; and, 2) The conversion of succinic acid moieties back to water-insoluble succinic anhydride groups, which could be slow to hydrolyze when con ned to deposits on ber surfaces.

Some Grafted Fiber Properties
Water Retention Values (WRV) of Treated Pulps. Figure 6 shows the in uence of grafted PEMAc content on WRV for pulps tested at neutral pH. Note that the bers were impregnated at pH 4 or 8, cured, and extensively washed before WRV measurements. The increase in swelling with the contents of xed PEMAc illustrates the contribution of polymer hydrogel to the overall water contents. For a given PEMAc content, most of the pulps impregnated at pH 8 were more swollen than those impregnated at pH 4.
Presumably, the grafting to cellulose and possibly crosslinking within the PEMAc layers restricted the swelling of the chemically xed pH 4 impregnated pulps. The slopes of the lines indicate that at pH 4, 23.7 g of water were present for every g of PEMA whereas, for pH 8 treatment, the ratio was 57.1 g water/g PEMA. Superabsorbent polyacrylates can bind an order of magnitude more water, particularly when the ionic strength is very low. (Lin et al. 2001) Therefore, the xation process reduces the ability of the polymers, and by extension the pulps, to hold water.
PEMAc Distribution in Pulp Fibers. In many of the experiments described herein, the impregnated pulp sheets were loaded with approximately 30 mg of PEMA per g of dry ber. The speci c surface area of an unbeaten pulp for a high molecular weight probe is ~ 1 m 2 /g. Therefore, a uniformly impregnated pulp sheet is coated with a polymer lm with a dry thickness of approximately 30 nm and is equivalent to about 30 layers of dry polymer. These order-of-magnitude estimates emphasize that most of the added PEMAc is not in direct contact with cellulose. In earlier work, we have shown that impregnation of lter paper with non-adsorbing, water-soluble polymer (dextran) results in an uneven polymer distribution on the dried lter paper. (Pelton et al. 2003) Capillary forces during drying result in thick polymer deposits at ber-ber junctions and thinner coatings on exposed ber surfaces. Figure 7 shows uorescent micrographs of pulp bers impregnated with uorescently labeled H-PEMAc. The labeled PEMAc is not uniformly distributed on the bers. The dark regions on the ber surfaces may indicate domains where ber/ ber contacts in the dense pulp sheets, prevented polymer access during impregnation.
Hydrolytic Stability of PEMAc Treated Pulps. Commercial applications of PEMAc grafted pulps are likely to require stable cellulose-PEMAc linkages. Figure 8A summarizes results from soaking at room temperature for up to 25 h. There was no change in the titratable polymer contents at pH 7 whereas the polymer contents after 24 h at pH 4 and 10 decreased by about 10%. Results at more extreme conditions of 0.1M NaOH at 70 o C are shown in Figure 8B which shows the total ber charge versus aging time. For pH 8 impregnation where physical xation dominates, it took three days for the charge content to revert to the carboxyl contents of the untreated pulp. Since PEMAc is chemically stable under these conditions, the lowering of ber charge with time re ects the detachment of PEMAc chains from the ber surfaces. The high initial ber charge for pH 4 impregnation re ects the higher yield of the pH 4 chemical xation. Furthermore, the rate of charge loss was much slower for the pH 4 compared to pH 8 impregnated bers.
We assume that with time the ester linkages to cellulose hydrolyze under hot alkaline conditions. Based on the results in Figure 8, a dry PEMAc grafted pulp that is repulped and fed to papermachines will not lose xed polymer due to hydrolysis or other mechanisms.

Conclusions
The objective of this work was to determine curing conditions whereby poly (ethylene-alt-maleic acid), PEMAc, could be xed to bleached kraft softwood pulp bers in high yields using conditions suitable for implementation in a pulp mill. The main conclusions from this work are: 1. High yields of PEMAc xation onto bleach kraft softwood pulp can be achieved with, catalyst-free, mild curing conditions (T < 150 o C).
2. With impregnation followed by curing (T > 100 o C) there are two mechanisms by which PEMAc is xed to cellulose -chemical and physical xation. With room temperature curing only physical xation is operative.
3. Evidence supports the explanation that physical xation is a consequence of the slow swelling and dissolution of thick dried PEMAc deposits on ber surfaces.
4. Chemical xation includes grafting to cellulose plus enhanced cohesion within thick PEMAc layers, possibly due to interchain crosslinking.
5. The pH of the PEMAc impregnation solution determines the xation mechanism for curing temperatures above 100 o Physical xation dominates when pH > 8 whereas chemical xation dominates for impregnation pH values < 7, suggesting the curing reactions require partial or complete protonation of the succinic acid moieties in PEMAc.
6. The yield of impregnated polymer xed to the bers after washing depends upon the xation mechanism. When chemical xation dominates, the yields approach or are equal to 100%, whereas physical xation yields are ~50% for H-PEMAc and 0-20% for L-PEMAc.
The technological signi cance of this work is that PEMAc can be grafted to pulp bers in high yields under mildly acidic conditions, expanding the properties space of wood pulp bers. Furthermore, high PEMAc xation yields can be achieved under curing conditions that could be implemented in pulp drying machines producing dry market pulp. However, our preliminary work showed that pulp impregnated with pH 4 PEMAc gave very high wet strength, complicating pH 4 PEMAc treatment for dried market pulp due to poor repulpability. (Zhang et al.) The next paper in this series describes approaches to obtaining low wet strength for PEMAc treated pulp while maintaining high xation yields.

Declarations
Funding