Cationized fibers from pine kraft pulp: advantages of refining before functionalization

Refining and fiber characterization:

Pine kraft pulp was diluted to 1.5% consistency and disintegrated at 3000 rpm for 10 min, with a Lorentzen and Wettre (Stockholm, Sweden) device that conforms to ISO 5263 (ISO TC/6 2011). Refining was performed by means of a Maskin’s Mark VI PFI mill (Hamar, Norway) (0.37 kW) following ISO 5264/2. Seven refining intensities were chosen at 3500, 3750, 4000, 4250, 4500, 4750 and 5000 PFI revolutions. The drainage capabilities of the unbeaten and beaten pulps were measured with a Canadian Standard Freeness (CSF) tester, in accordance with the Tappi method T 227 (1999). Enough replicates were made so that the relative standard deviation (StD) remained below 2%. The dimensions and populations of fibers were measured by a MorFi fiber analyzer from Techpap (France). Each suspension was prepared by diluting 1 g of pulp in 600 ml of water. The software (V. 7.9.13E) was adjusted to stop imaging after counting 5000 fibers and then the average values were calculated. Each measurement was repeated three times (see Moral et al. 2010).

Alkalization:

All reactions were carried out in a 2 l three-neck spherical glass reactor equipped with a refluxing condenser and magnetic stirring. Among the eight pulps (the original one as control pulp and the seven refined ones), the refined pulps at 0, 3500, 4000, 4500 and 5000 PFI revolutions were selected for alkalization and cationization. In each experiment, 20 g of pulp (b.o. dry pulp) was soaked in a 10% NaOH aq. solution. After 1 h of vigorous stirring at room temperature, the fibers were separated by filtration over a Whatman Glass Microfiber GF/D filter (2.7 μm), and washed with demineralized water.

Cationization:

Alkalized cellulose was mixed with the reagent in a CHPTAC/AGU mole ratio of 4, adding a 5% NaOH solution dropwise so that the NaOH/CHPTAC mole ratio became 1, and enough isopropyl alcohol was added to have a 1 l suspension. Heating to 120°C for 2 h (heating mantle) was controlled via a Pt-100 probe and an electronic PID device. The heating time began from T_max. Afterwards, the cationized pulp was diluted to lower the pH, separated from the liquid with the aforementioned filter, carefully washed, and dried in a vacuum furnace at 45°C. Washing ensured that the small portion of cellulose that became water soluble after cationization was removed.

Determination of cationicity:

The N-content of the cationized samples was measured by an LECO CNS-2000I elemental analyzer (LECO Instrumentos, Madrid, Spain). The charge density was determined via automatic potentiometric titration [Charge Analysis System (CAS) device from AFG]. The equipment is designed for liquids, and thus a back titration mode was applied instead of the direct approach. A small amount of fibers (less than 0.3 g) were soaked in excess sodium polyvinylsulfate (3 ml of PVSNa, 1.8 meq l⁻¹) as an anionic polyelectrolyte. No stirring was applied, because the surface charge and not the total charge should be neutralized. Water was immediately added to make up to 10 ml and the resulting liquid was titrated with polydiallyldimethyl-ammonium chloride (PDADMAC, 2.2 meq l⁻¹) as a cationic polyelectrolyte. The endpoint of the titration was the isoelectric point (0 mV). Reversely, to measure the charge density of the control pulp (pine kraft pulp before treatments), a sample was soaked in PDADMAC and the titration was carried out against the anionic electrolyte.

Characterization of pulps:

The original pulp and the alkalized samples were grinded and submitted to a PANalytical’s powder XRD equipped with X’Pert software. The 2θ angle ranged from 10° to 45°. The limiting viscosity number of all cationic samples and of the original pulp was determined according to the ISO 5351-1 standard, in a capillary viscometer and an aqueous copper (II) ethylenediamine (Cuen) solution as solvent.

The bulk density was measured in a novel way. In each case, a suspension of disintegrated pulp was dewatered on a wire screen to obtain a test pad of fibers. After drying at room temperature at a relative humidity (RH) of 50% for 48 h, the density of the test pad was determined with a pycnometer for solids and liquids, as described in T258 om-02 for the basic density determination of pulpwood chips, but gallium at 30–50°C was the agent instead of water. In this temperature range, the surface tension of gallium is even higher than that of mercury at 25°C, and thus this liquid metal does not penetrate into the pores (Poole 2004). The raw density value was multiplied by 6.095 g cm⁻³, the density of gallium at 30°C (Hardy 1985).

Selected pulps were observed by scanning electron microscopy (SEM) with magnifications of 100× and 1000× (JEOL device, model JM-6400). Fibers were put on a cylindrical slide, which was dried at 45°C and 200 mbar and coated with gold. Another device from JEOL, a model JSM-6335F, was used to study the distribution of PCC on the surface of handsheets, with and without cationic fibers.

Performance in papermaking:

The retention of fillers and fines was studied by a laboratory device DFR-05 from Mütek. Four percent cationized pulp was combined with the original pulp refined to 3500 PFI revolutions (96%) and a suspension with a 0.5% consistency was prepared with tap water of conductivity 500 μS cm⁻¹. This suspension was mixed with PCC (0.2 g PCC as filler per gram dry pulp) to obtain a total mass of 800 g. The pH of the suspension was adjusted to 7.5 by adding HCl or NaOH dropwise. The retention program of Mütek’s software was applied to the fiber mixture under stirring (200 rpm for 60 s, and then at 300 rpm for 30 s). The suspension went through a 100-mesh screen, corresponding to a wire size of 0.11 mm.

As the DFR device does not distinguish between fillers and fines, the filtrate was submitted to a complexometric titration with EDTA to determine the amount of PCC with eriochrome black T as the indicator. The pH was kept around 10 by adding a buffer solution of ammonia and ammonium chloride. The amount of PCC retained is the difference between the PCC mass furnished initially and that in the filtrate. Five isotropic sheets were made from each of the following materials: kraft pulp, kraft pulp/PCC (8/2), and kraft pulp/PCC/cationic fibers (77/20/3), while the cationic fibers were beaten to various refining degrees. To this purpose, a laboratory sheet former was applied according to the ISO standard 5269/1 (ISO TC/6 2011). Agitation was carried out by hand, with a standard stirrer. The sheets were left to dry between rings to keep them pressed, at 23°C and 50% RH. The basis weight was 60 g m⁻². The opacity was determined by means of an Elrepho spectrophotometer from Lorentzen and Wettre with a C/2° light source, following ISO 2471 (ISO TC/6 2011).

Influence of refining and chemical treatments on morphology

The dimensions of the native fibers from pine are reported in the literature (Sable et al. 2012). Figure 2 presents the results of average fiber length (weighted in length), number of fibers (length higher than 100 μm) and percentage of microfibrils over the fiber surface (area-based calculations), obtained by means of the morphological analysis. Freeness decrement is visible as a function of refining (Figure 2d). The fiber shortening (Figure 2a) and the increasing protrusion of the microfibrils (Figure 2c) are also presented. Kraft pulps from hardwoods, in contrast, may need only 3000 PFI revolutions to achieve the values obtained for the pine kraft pulp after 5000 PFI revolutions (Bhardwaj et al. 2007). Compared with other fiber sources, softwood pulp needs the most energy for noticeable freeness differences. During refining, fibers undergo shear and compression. The generation of fines (length <100 μm) is evidenced by the cutting of fibers (Figure 2a). These secondary fines increase the water retention of the web, filling the gaps between fibers and lead to lowering CSF data. Refining to 3750 PFI revolutions slightly increased the fiber population because some fibers were cut. After a long beating time, the mechanical damage resulted in a notorious fiber loss from 10.8 M g⁻¹ to 9.6 M g⁻¹.

Figure 2:

It is shown how fibers are slightly shortened by refining and how freeness is clearly reduced.

Effect of refining on (a) fiber length, (b) fiber population, (c) proportion of microfibrils on the fiber surface, and (d) Canadian Standard Freeness.

The SEM images in Figure 3 reveal that the control pulp is nearly not curled (Figure 3a) and that their surfaces contain only a few microfibrils (Figure 3b). Alkalization and cationization of this unrefined pulp increased the number of curls, but the fibers were swollen to a lesser extent than expected (Figure 3c). As can be seen from Figure 3d, the fiber surface became much rougher and the fiber wall was clearly peeled due to the chemical damage. When the pulp was refined to 5000 PFI revolutions and then pretreated and cationized, extreme internal and external fibrillation was found (Figure 3e), and at the end, the fiber walls were nearly destroyed (Figure 3f). It is obvious that refining to a certain extent before cationizing increases the specific surface area.

Figure 3:

Chemical treatments damaged the surface, especially if the pulp had been refined.

Micrographs for (a) original kraft pulp (100×), (b) original kraft pulp (1000×), (c) pulp after alkalization and cationization (100×), (d) pulp after alkalization and cationization (1000×), (e) pulp after refining to 4500 PFI revolutions, alkalization and cationization (100×), (f) pulp after refining to 4500 PFI revolutions alkalization and cationization (1000×).

Influence of refining on decrystallization and charge

The XRD patterns displayed in Figure 4 resemble the typical shape of cellulose Iβ, which is the main cellulose polymorph in wood. Miller indices are assigned to the most prominent peaks, based on the suggestions of French (2014). Despite the treatment with NaOH, there was little conversion to cellulose II. The most prevalent peaks of this polymorph, which correspond to the planes (110) and (020), should appear at 20.1° and 21.5°, respectively. At most, the (200) peak of cellulose Iβ became slightly lopsided. Nonetheless, partial amorphization of fibers is visible after pretreatments. The intensity of the crystalline peaks at 15° and 16.5° decreased after alkalization, and this decrease was found to be more abrupt in the case of a preceding beating at 4000 PFI revolutions or more. There is an evident diminishment in the area under the peaks for the plane (200) of cellulose Iβ (22.5°) and (004) of both cellulose Iβ and cellulose II (34.6°).

Figure 4:

X-ray diffraction patterns for the original pulp and for the alkalized samples after refining.

The crystallinity index (CI) listed in Table 1 was estimated from the diffraction patterns by performing a Gaussian deconvolution with Systat’s Peakfit software, identifying four crystalline peaks, and dividing their area by the total area according to Eq. (1) (Park et al. 2010). Also, as a way to estimate the conversion of cellulose I to cellulose II, the area of the distinguishable crystalline peaks associated with cellulose I after deconvolution was divided by the area of the distinguishable crystalline peaks associated with cellulose II (Eq. 2). Overlapping peaks with those of other polymorphs were omitted. This is only a comparative estimation, and it is not representative of the quantitative conversion of cellulose I to cellulose II.

Refining enhanced the interactions between the hydroxide and the fibers during alkalization and led to a crystallinity decrement. Nonetheless, its influence on the conversion of cellulose I into cellulose II is not clear (Table 1). Refining increases the negative surface charge of fibers (Banavath et al. 2011) because of the enlarged area and the higher amount of hemicelluloses in the fiber wall. Alkalization alone produces a higher value, even without refining, because the hydroxyl groups in cellulose are ionized.

Characterization and testing of cationized pulps

Elemental analysis data of CF are presented in Table 2. To calculate the degree of nitrogen substitution (DNS) of the CF, it was assumed that they consist entirely of anhydrohexose units. The amount of anhydropentose units is low because of the alkaline hydrolysis of hemicelluloses during kraft pulping (Deutschle et al. 2014; Postma et al. 2014) and their molecular mass is around 30 Da or lower. Hence, Eq. (3) contains the molecular mass of an anhydrohexose unit (162 Da), together with that of the quaternary ammonium group (152 Da) and 100 times the atomic mass of nitrogen (Moral et al. 2015).

The degree of substitution (DS) was in the high range for insoluble cationic cellulose. According to Besemer et al. (2005), the functionalization of fibers should not reach more than 30 cationic groups per 100 monomers; otherwise a lot of water-soluble polyelectrolytes may arise. Regardless of the pretreatment, the yield was approximately 95%, as the high DP of cellulose from pine wood and its high crystallinity prevented the polymer from being dissolved. Overall, the DS increased with decreasing crystallinity, because amorphous and paracrystalline cellulose is less stable thermodynamically and more reactive than cellulose I and cellulose II (Poletto et al. 2014). The exception was the pulp with the lowest crystallinity fraction, likely because the part with the highest DS was solubilized. The charge density followed the DS trend. The 4% CF would be sufficient to neutralize the global surface charge when the cationized pulp is refined.

Cationization has a positive effect on the pulp viscosity and a negative effect on the apparent density of the pulp pads. The former is due to the changes in the spatial distribution of the cellulose chains in fibers, which become stiffer (Moral et al. 2016b). The latter is explained by the strong positive charge of the cationic fibers, which are repelled from each other, and by the breakage of intermolecular hydrogen bonds.

Enhancement of retention

The performance of the CF refined to different levels is illustrated in Figure 5. Mechanical retention of PCC, without aids, did not reach 55% (Figure 5a), as the aperture size of the wire was much larger than the particle size of PCC. There is no appreciable difference in filler retention (Figure 5a). To the best of our knowledge, so far, CF have only been proved to improve kaolin retention (Sang and Xiao 2009). The filler has a strongly negative zeta potential and its retention is favored by charge neutralization. PCC may require a flocculation agent (a soluble polyelectrolyte), regardless of the surface charge of the fibers. Nonetheless, the total retention of solids was clearly improved by the presence of CF. Therefore, the insertion of cationic functional groups in some fibers increased the retention of pulp fines. The results were particularly good for the CF refined to 4000 and 4500 PFI revolutions, matching their high surface charge. In these cases, the pulp loss was less than 1%.

Figure 5:

Results for the pine kraft pulp (77%) and PCC (20%), when mixed or not with cationic fibers (3%).

(a) Total retention and retention of PCC; (b) C/2° opacity.

Because of the reduced loss of fines, the opacity was increased by the addition of CF, particularly when its DS was high (Figure 5b). Fines fill the gaps between fibers in the paper web, blocking the light in a higher degree for a given basis weight.

As the SEM images demonstrate in Figure 6 (surface of isotropic sheets with PCC without flocculants), cationic fibers modified the filler distribution. Figure 6a shows a very heterogeneous distribution of the calcium carbonate particles, which became aggregated in the area circled. These particles are more evenly distributed in Figure 6b, due to the addition of CF with the highest surface charge, enhancing fiber-filler bonding. A small amount of CF can both increase the bulk and the apparent density and improve the optical properties.

Figure 6:

Addition of cationic fibers seemed to improve filler distribution.

Micrographs of the paper surface for (a) pine kraft pulp with PCC, (b) pine kraft pulp with PCC and cationic fibers that had been refined to 4500 PFI revolutions.