Wet strength is one the most difficult paper properties to improve. Whereas polyethylene films have the same strength wet or dry, paper products undergo a catastrophic loss of strength when exposed to water. Individual wood pulp fibers are somewhat weaker wet than dry; unbleached pulp fibers have about the same strength wet or dry, whereas bleached fibers can be up to 30 % weaker (Gurnagul and Page 1989). However, fiber weakening does not explain the low strength of wet paper. Instead, water swelling of fiber-fiber joints severs hydrogen bonds and both weakens the joints and renders the fiber wall more susceptible to delamination. The current solution is the use of wet-strength resins that strengthen wet fiber-fiber joints. The chemical companies supplying the paper industry continue to evolve wet-strength resin technologies. The gold standard for evaluating new wet-strength resins is a papermachine trial. However, much of the early development work typically employs wet tensile measurements of laboratory made paper. We started working on wet strength chemistries in 2002 and have developed a new laboratory wet strength measurement that augments handsheet wet tensile measurements, giving much additional information. The goal of this paper is to describe the technique which we call “Wet-peeling”, illustrating the utility of wet-peel measurements with data collected over the last 15 years.
Wet-peeling is a variation of an approach first described by McLaren in 1948 (McLaren 1948). Two wet, regenerated cellulose membranes are stacked with a thin layer of wet-strength resin between them. The stack is pressed (laminated), dried, and then rewetted. The resulting laminate is a physical model for wet fiber-fiber joints in paper. Wet laminate strength is measured as the 90-degree peel delamination force required to separate the two membranes.
Most of the results presented below employed examples from three types of commercial wet-strength resins; polyamideamine-epichlorohydrin (PAE), polyvinylamine (PVAm), and glyoxalated cationic polyacrylamide (GCPAM). PAE, a modified condensation polymer, is arguably the most common type of wet-strength resin used today. The literature reports a very broad range of weight-average molecular weights (7 to 1,100 kDa) and very broad molecular weight distribution (Obokata et al. 2005). PAE has a moderate cationic charge; our sample gave 2 meq/g, based on polyelectrolyte titration. With drying and heating, azetidinium groups form covalent crosslinks and covalent grafts to carboxyl groups on fiber surfaces – see Figure 1. The lowest molecular weight fractions of PAE can enter fiber wall pores, strengthening the fibers (Taylor 1968) (Andreasson et al. 2005).
PVAm is a linear, highly cationic (8 meq/g, hydrochloride salt and 50 % ionization), high molecular weight water-soluble polymer that, with drying, forms imine and aminal grafts to aldehyde groups on oxidized cellulose – see Figure 1. Finally, GCPAM is a modified cationic polyacrylamide, often with high molecular weights (106–107 Da) and intermediate charge contents. With drying, the glyoxal adducts form covalent hemiacetal bonds with cellulosic alcohol groups – see Figure 1.
In addition to grafting, PAE and GCPAM form covalent crosslinks within and between the polymer chains, contributing significantly to joint strength – example crosslinking chemical structures are given in Espy’s review (Espy 1995). By contrast, PVAm molecules do not form covalent crosslinks so the cohesive strength of PVAm films depends upon physical interactions.
In summary, all three wet-strength resin types are cationic, promoting electrostatic adsorption of the resins onto anionic fiber surfaces. PAE may penetrate small pores in the fiber wall, whereas high molecular weight GCPAM and PVAm cannot. The grafting chemistries vary: PAE requires cellulosic carboxyls, drying and heating; PVAm requires cellulosic aldehydes and drying; and, GCPAM requires cellulosic hydroxyls, and drying. The specific products used in this work were chosen at random – no effort was made to identify a “best-of-breed”.
Materials and methods
Regenerated cellulose membranes (Spectra/Por2, MWCO 12–14 kDa, 76 mm diameter tubing, product number 132684) were purchased from Spectrum Laboratories, US. TAPPI Standard blotter papers were purchased from Labtech Instruments Inc., Canada. Moisture-resistant double-sided medical tape 1522 was purchased from 3M, US. Glyoxalated cationic polyacrylamide (GCPAM) was Luredur Plus 555 provided by BASF, US. Polyamide-epichlorohydrin (PAE) was Kymene 5221 provided by Solenis, US. Polyvinylamine (PVAm) samples were provided by BASF, Germany. Some PVAm samples were further hydrolyzed (DiFlavio et al. 2005) to remove residual formamide moieties and all PVAms were dialyzed for 1 week and freeze-dried before use. All other chemicals were used as received. TEMPO (2,2,6,6-tetramethylpiperidine 1-oxyl) and all other chemicals were purchased from Sigma-Aldrich, Canada. Carboxymethyl cellulose (CMC) used in this study was Mw 250 kDa with DS 0.9. Water type 1 (as per ASTM D1193-6, resistivity 18.2 MΩ/cm) was used in all experiments.
Pre-treatment of regenerated cellulose membranes
Regenerated cellulose dialysis tubes (12 cm diameter) were cut into strips, either 6 cm × 2 cm for the “top strips” or 6 cm × 3 cm for the “bottom strips”, as shown in Figure S1. The membranes were cleaned by extraction in water at 60 °C for 1 h and stored in water at 4 °C.
The regenerated cellulose membranes were oxidized by TEMPO-mediated oxidation. In our standard method for membrane oxidation, 68 mg of TEMPO was dissolved in 2 L of water along with 680 mg NaBr and followed by the addition of 300 mg NaClO. The concentration of NaClO solution was determined by available chlorine titration (TAPPI, Test Method T 611 cm-07). The pH of the solution was adjusted to 10.5, after which wet cellulose membranes (10 g dry weight) were added. The solution was stirred and the pH maintained at 10.5 for 15 minutes, after which the oxidation was stopped by the addition of 10 mL ethanol. The oxidized membranes were washed thoroughly with water and stored at 4 °C.
Borohydride Reduction of Oxidized Cellulose Membranes. Sodium borohydride was used to reduce C6 aldehydes on oxidized cellulose to the corresponding alcohol. Wet, oxidized cellulose membranes (1 g dry weight) were immersed in 100 mL of 10 mM phosphate buffer at pH 8.5, followed by the addition of 0.5 g sodium borohydride. The pH of the suspension was adjusted to 8.5 using 1 M acetic acid. The reduction was carried out at room temperature and pH 8.5 for 48 hours. After the reaction, the cellulose membranes were washed with water and stored at 4 °C.
CMC was irreversibly adsorbed to cellulose membranes following the method (Laine et al. 2000). CMC was dissolved in water and followed by the addition of CaCl2 giving a final treatments solution composition of 1 g/L CMC in 50 mM CaCl2 at pH 8.0. Wet, regenerated cellulose membranes (10 g dry weight) were immersed in 400 mL of the treatments solution and the CMC was allowed to deposit to cellulose under mild stirring for 3 hours at 80 °C and pH 8. The CMC-treated cellulose membranes were washed with water and stored in 1 mM NaCl solution at 4 °C.
Lamination with wet-strength resins
In this work, two methods were used to apply resin in laminate joints, “adsorption application” and “direct application”. In a typical adsorption application experiment, wet cellulose membranes (1 g dry weight) were soaked in 50 mL 0.2 g/L polymer solution with 1 mM NaCl at pH 7 in a plastic Petri dish for 30 min. Then the membranes were transferred to another Petri dish containing 50 mL rinsing solution (1 mM NaCl, pH 7) to rinse the non-adsorbed polymers from the membrane surfaces. The rinsing solutions were changed 3 times during the 10 min of rinsing. After the rinsing two wet cellulose membranes, each covered by a monolayer of polymer, were laminated. To permit easy separation at one end of the laminate, a piece of Teflon tape was applied across one end of the bottom membrane before lamination – see Figure S1. After lamination the Teflon tape was removed and the top membrane could easily be attached to the jaws of the Instron used to measure the peel force. The overall lamination procedure is illustrated in Figure S2.
In the direct application method, a top and a bottom membrane were blotted free of excess water, the water content was approximately 54 wt% after the blotting. 15 μL of wet-strength resin solution was carefully applied to the bottom membrane using a micropipette, after which a top membrane was applied, and the two membranes laminated using the Teflon spacer described above – see Figure S2. Normally, our resins are dissolved in 1 mM NaCl at pH 7. However, the pH and ion strength of the resin solution are variables which can be varied systematically to give mechanistic information. The polymer coverage in the laminate joints (mg/m2) can be calculated from the laminate surface area (typically 10 cm2) and the concentration and volume of the resin solution added.
Lamination and Wet-peeling. Wet laminates were placed between two 8 square inch TAPPI blotter papers and pressed in a Standard Auto CH Benchtop Press (Carver, Inc., US) with an applied pressure of 323 kPa for 5 min. The pressed laminates were then dried unrestrained at constant temperature and humidity (23 °C, 50 % RH) for 24 h.
For never-dried samples, lamination was performed at 23 °C. With once-dried laminates, hot pressing can be used to promote grafting and crosslinking. Most the results herein were with room temperature pressing and drying.
Before delamination, the dried laminates were soaked in rewetting solution for 30 minutes. Our normal rewetting solution is 1 mM NaCl at pH 7. However, varying the pH and ionic strength can give mechanistic information. Dilute buffer solutions can give better pH control. We do not recommend deionized water for rewetting because the swelling and electrostatic forces are exaggerated.
The wet-peel force was measured with an Instron 4411 (Instron Corp., US) universal testing system. The lower jaw was replaced with a freely rotating 14 cm diameter aluminum peeling wheel running on SKF-6,8-2RS1 radial bearings (SKF, Scarborough, ON, Canada). The peel wheel has a 40 mm wide smooth outer surface. Our peel wheel is based on a design from Paprican (now FPInnovations), Point Claire, QC, Canada (Skowronski and Bichard 1987).
In a typical test, the rewetted laminate was removed from the soaking solution and excess water was removed by placing the laminate between two blotter papers and pressing once with a 2.4 kg hand roller. The wet laminate was fixed to the peeling wheel using moisture-resistant double-sided tape and the end of the top membrane was peeled off the Teflon tape spacer and fixed to the crosshead jaw of the Instron – see Figure S3. Our standard peel rate was 20 mm/min and the resulting peel force, measured with a 50 N load cell, was recorded as a function of displacement.
The Instron software was used to determine the average peel force over a user-defined displacement, corresponding to steady-state peeling. The steady state peel force is normalized with the width of the top membrane giving reported results with the unit N/m. At least three replicates were performed for each test and the average value and standard deviation were reported. The laminates were weighed wet before mounting on the wheel and after peeling, as well as after drying. The solids content of air-dried (23 °C, 50 % RH) membranes was 92–93 %, whereas the solids contents of the wet laminates were in the range of 47–54 wt% during the delamination. The solids contents increased a couple of percent during peeling because of evaporation.
Never-dried wet adhesion. Wet laminates were pressed (323 kPa) between two standard TAPPI blotter papers for times ranging between 0.1 min – 10 min; the longer the pressing time, the lower the laminate water content. The pressed wet laminates were put in a sealed plastic bag to prevent water evaporating and transferred immediately to the Instron. The wet-peel force was measured at a peel rate of 20 mm/min. The laminate wet weight was taken as the average value before and after testing – typically the wet mass decreased 2–3 wt% during peeling. The solids content is calculated as the oven-dry mass divided by the wet mass.
Cellulose membrane characterization
Mechanical properties of cellulose membranes were measured with an Instron 4411 (Instron Corp., US) universal testing system with 500 N load cell. The tensile rate was 20 mm/min and the grip distance was 3 cm. All membranes were cut to the size of 6 cm × 2 cm. For never-dried membranes, the samples were tested immediately out of water solution. For once-dried membranes, the samples were air-dried in a constant temperature and humidity room (23 °C, 50 % relative humidity) for one day, and thereafter rewetted in 1 mM NaCl solution at pH 7 for 1 hour before the measurement.
During tensile testing, the solids contents of the test specimens were typically in the range of 40–45 wt%. The reported results are the average of at least 3 replicates. The error bars depict the standard deviation.
The aldehyde content of oxidized membranes was measured from the fluorescence emission of dansylhydrazine labeled membranes, following Dementev’s procedure (Dementev et al. 2009). Cellulose membranes were cut into 10 mm × 20 mm strips. In a typical reaction, wet cellulose (45 mg dry weight) was immersed in 30 mL of 0.14 mM dansyl hydrazide in methanol, followed by the addition of 1 mL of 1 M HCl. After 24 hours at room temperature and in the dark, labelled cellulose was rinsed with methanol and stored in the dark. The fluorescence intensity was determined using a ChemiDoc MP Imaging System (Bio-Rad Laboratories, Inc. US). The labelled cellulose strips were fixed to glass microscope slides (Pre-cleaned, Corning) with 3M 1522 double sided medical tape. UV Trans illumination (302 nm) was used as the excitation source and Standard Filter (580 ± 120 nm) was used as the emission filter. Images were analyzed using Image Lab version 4.1 (Bio-Rad Laboratories, Inc. US).
Calibration solutions of dansylhydrazine/acetone (1:2 molar ratio) in methanol were prepared with dansylhydrazine concentrations in a range of 0.2–4 mmol/L. 50 µL solution was applied to oxidized cellulose strips (10 mm × 20 mm) fixed to glass slides and air-dried in the dark. The resulting fluorescence intensities were a linear function of the dansylhydrazine contents.
Pulp treatment and handsheet testing
Pulp treatments and handsheet preparation: Unbeaten northern bleached softwood kraft pulp (SBK) suspensions were prepared from dried market pulp. CMC-treated pulps were prepared by treating a 2.5 % consistency pulp in 50 mM CaCl2 with CMC (50 mg per o.d. gm of pulp) and the pH was adjusted to 8. The suspension was heated to 95 °C and mixed for two hours.
TEMPO oxidized pulps were prepared as follows: Disintegrated pulp (25 g) was diluted to 4 L with DI water and TEMPO (60 mg) and NaBr (600 mg) were added. The mixture was stirred for about 30 min until the TEMPO dissolved. The reaction was initiated by addition of 12.5 mL NaClO (1.4 M) and the reaction pH was maintained at 10.5 by addition of NaOH (1 M). After 15 min, the reaction was quenched by addition of excess ethanol. The oxidized pulp was filtered and carefully washed multiple times and then stored at 10 wt% in refrigerator until further use.
Handsheets with a target basis weight of 60 g/m2 were prepared using a semi-automatic sheet maker (Labtech Instruments Inc., Model 300-1, Canada) following TAPPI method T205 sp-95. Pulp with a consistency of 0.25 % were mixed with 2 % PAE, PVAm or GCPAM (based on dried-pulp weight) for 15 min at pH 7 prior to sheet-making. All of the handsheets were dried on a speed drier (Labtech Instruments Inc., Canada) at 120 °C for 10 min and stored at 23 °C and 50 % RH.
Paper wet tensile index: Wet tensile indices of handsheets were measured following TAPPI method T494 om-96. Paper specimens (150 mm × 15 mm) were rewetted in 1 mM NaCl at pH 7 for 5 min. Excess water was removed by blotter papers before the tensile test. Tensile strength was measured with the Instron fitted with a 50 N load cell at a stretching rate of 20 mm/min. Each experiment was repeated at least 4 times.
The results are presented in two parts: The wet-peel methods and Demonstrating utility. Whereas the experimental section describes the details, the following section describes the results of each step of the procedure, highlighting what we believe to be the most important issues. The second section, Demonstrating Utility, shows examples of experiments that can reveal features not accessible using traditional wet-tensile measurements on handsheets.
The wet-peel methods
The steps for wet-peel methodology are summarized in Figure 2. This workflow measures either the once-dried wet adhesion or the never-dried wet adhesion. We propose that once-dried wet adhesion measurements are related to wet paper strength. In contrast, we propose that the never-dried wet adhesion measurements reflect the chemical contribution of polymers to wet-web strength. The detailed procedures corresponding to each step in Figure 2 are given in the experimental section, whereas the following paragraphs describe the key features of each step.
Cellulose membrane pretreatment
We used dialysis tubing as the source of regenerated cellulose membranes. Some membrane properties are summarized in Table 1. In all cases the cellulose membranes had to be pretreated. In the simplest case, pretreatment was water washing to remove glycerol and other water-soluble impurities. However, the regenerated cellulose membranes have insufficient non-hydroxyl surface groups to give wet strength with most wet-strength enhancing polymers. Herein we present results for membranes with a surface layer of CMC and for TEMPO oxidized surfaces. Our CMC-treatment protocols are directly based on the work of Laine and Lindstrom who demonstrated that CMC could be deposited, virtually irreversibly onto cellulose fibers (Laine et al. 2000). By contrast, low temperature CMC solutions in low ionic strength solutions show little tendency to bind to clean wood pulp fibers.
Membrane treatments by TEMPO mediated oxidation follow Saito’s early work (Saito and Isogai 2006). Both CMC treatment and TEMPO oxidation generate surface carboxyl groups that are grafting sites for PAE (see Figure 1) and that promote adsorption of cationic polymers. TEMPO mediated oxidation also generates aldehyde groups (Saito and Isogai 2006) that are grafting sites for PVAm (DiFlavio et al. 2005).
The surface treatments have little impact on the mechanical properties of the wet membranes under our standard conditions. We also measured membrane mechanical properties after exposure to PVAm and PAE. The results, summarized in Table S1, show that the polymers did not have a significant influence on the membrane mechanical properties. Therefore, we propose that wet-peel values from these treated membranes can be directly compared.
Measuring the density of surface functional groups is a challenge because of the very low total surface area in the membranes. We used dansyl hydrazine to label the aldehyde groups, and fluorescence to estimate the total aldehyde content of the labeled membranes. Figure 3 shows the labeling chemistry and superimposes on the calibration curve the properties of the unmodified membrane, TEMPO oxidized membrane, and, TEMPO oxidized membrane that was treated with sodium borohydride to reduce the aldehydes. Only the TEMPO mediated oxidation gave a significant aldehyde content of 0.71 μmol per square meter of membrane top surface (10 μmol per dry gram). However, we have shown previously that confocal images of membrane cross sections revealed a uniform distribution of fluorescent groups in the z-direction (Liu et al. 2013). Therefore, to estimate the surface density of aldehydes we assumed that the “surface aldehydes” corresponded to the top 5 nm layer of the membrane. We have expressed our estimate of the aldehyde density in Table 1 as the area per aldehyde group of approximately 62 nm2. TEMPO mediated oxidation targets the C6 hydroxyl and the estimated area per exposed C6 hydroxyl on crystalline cellulose is 0.6 nm2. (Fleming et al. 2001) Therefore, based on our estimate, only a small fraction of the accessible C6 hydroxyls are present as aldehydes in our experiments.
Wet membranes were 130 ± 10 µm thick. Once-dried membranes were dried at 23 °C. The surface aldehyde density is based on the assumption of a 5 nm surface layer. RMS roughness values measured by AFM (Yang 2018). Error limits depict the standard deviation based on at least 3 replicates.
Finally, Liu et al. reported that the CMC (250 kDa, DS 0.9, 30 mM CaCl2) adsorption density on regenerated cellulose was 0.34 mg/m2 (Liu et al. 2011). This corresponds to an area per carboxyl group of 1.2 nm2. Note that an adsorbed layer of CMC will have loops and tails extending into solutions. Therefore, the carboxyls are dispersed through the adsorbed layer thickness.
Wet-strength resin application
Two methods have been developed for applying wet-strength resins to the membranes – direct application and adsorption, as show in Figure S2. Direct application involves spreading 15 μL of polymer solution over a joint area of 10 cm2. In early work we were concerned that pressing would squeeze out some of the wet-strength resin solutions. Figure S4 shows photographs of a laminate and the blotters after pressing. The laminate was formed with a Congo Red solution to facilitate the visualization of lost solution. There was no obvious leakage with pressing. The hand blotting of membranes just before adding the 15 μL of polymer solution is a critical step. The blotted membrane has the capacity to hold the directly added polymer solution.
We show below that direct application is useful for comparing a range of polymer types, while keeping constant the polymer content in the laminate joint. For much of our work the wet-strength resin coverage (dry polymer mass/joint area) was 15 mg/m2 corresponding to a dry thickness of 15 nm, corresponding to approximately 15 layers of polymer molecules between the cellulose membranes. However, compared to conventional adhesive technologies, these are very thin adhesive layers.
In adsorption application one or both membrane surfaces are exposed to a polymer solution to give an adsorbed monolayer on each surface treated. Adsorption application is the closest to the situation in papermaking. However, adsorption has two serious drawbacks. First, the coverage (dry polymer mass/joint area) is difficult to measure. Second, it is impossible to control coverage which, in turn, means a series of polymers cannot be compared at the same coverage. Typically, adsorbed water-soluble polymers give coverages in the range 0.1–1 mg/m2 on an individual surface. Forming a joint between two such surfaces doubles the polymer coverage in the joint.
For both direct and adsorption application, it is important to control the pH and ionic strength of the polymer solution. The degree of ionization and the polymer configuration are sensitive to aqueous solution conditions.
We believe the membranes are manufactured in an extrusion process that generated striations in the peeling direction. Atomic force microscopic reveals a wet RMS roughness of only 10–15 nm. However, the images suggest parallel striations – see Figure S5. We prepared a series of laminates with fluorescently labeled PVAm and used confocal microscopy to image the plane of maximum fluorescence between the cellulose membranes. The top image in Figure 4 shows a laminate made by direct application. The directly applied polymer is present as bands reflecting the striations in the membranes. By contrast, adsorption from solution gave a much more uniform distribution of polymer. Finally, the striations were parallel to the long axis of the parent dialysis tubing which was also our peeling direction.
Lamination. Pairs of polymer treated, wet membranes are pressed together between standard handsheet blotters to ensure good contact. Originally our pressing pressures were high (2150 kPa), for no particular reason. Our current standard pressing pressure is 323 kPa. Figure 5 shows that wet-peel force is not very sensitive to lamination pressure over a great range. The highest wet-peel forces corresponded to the highest lamination pressures.
For never-dried wet-peeling, the laminate water content is an important variable. Because pressing between dried blotters lowers the water content, pressing time is a convenient way to control never-dried water content.
GCPAM, PVAm, and PAE resins all require water removal for covalent bonds to form within in the resin (crosslinking) and to the cellulose surfaces (grafting). For GCPAM and PVAm the once-dried laminates were normally removed from the lamination press and dried at 23 °C, 50 % RH overnight. Under these conditions, the equilibrium laminate water content was about 7.5 %. PAE crosslinking reactions are facilitated by drying at elevated temperatures (Obokata and Isogai 2007). In these cases, we employed a heated lamination press, combining the heating and drying step. We have found that the upper temperature limit for the laminates is about 70 °C for PAE treated membranes. Above this, we observe membrane failure during the wet-peeling possibly indicating the joints were stronger than the membranes.
Once-dried laminates must be rewetted-before wet-peeling. The laminates are hydrogels and the degree of swelling will depend upon the pH and ionic strength of the rewetting solution – these properties should be controlled. Typically, we use 1 mM NaCl at neutral pH as the rewetting solution. Deionized water should not be used as it exaggerates joint swelling. A dilute buffer is recommended for pH sensitive adhesive chemistries.
Our standard rewetting time of 30 minutes is usually sufficient to saturate the laminate. In early work we were concerned that water soluble PVAm could diffuse out of laminates. We showed that laminates bonded with 7.5 mg/m2 could be soaked for 2000 hours with no loss in wet-peel force (DiFlavio et al. 2005).
The laminates are removed from the soaking solution, hand blotted, and weighed just before and after wet-peeling. Typically, each laminate takes approximately 3 minutes to weigh and delaminate. The solids content increases about 1 % per minute at 23 °C, 50 % RH, giving an average of about 50 % during the delamination.
Laminate adhesion was measured as the ninety-degree peel delamination force values – see Figure S3 for photographs of a wet-peel experiment. The reported wet-peel results are expressed in units of N/m obtained by dividing the measured forces by the width of the top membrane. Note that wet-peel forces are dimensionally equivalent to J/m2, corresponding to the total delamination work per area of laminate joint.
The mechanics of peeling has been studied and modeled in the pressure sensitive adhesives literature (Kendall 1975) (Gent and Hamed 1977) (Kinloch 1982) (Zhang Newby and Chaudhury 1998) (Pesika et al. 2007). Peel work is much greater than the thermodynamic work of adhesion (Zhang Newby and Chaudhury 1998) (Li et al. 2001) reflecting energy consumption by stretching the tape backing and viscoelastic deformation within thick adhesive layers. In the case of our experiments, non-adhesive contributions to the peel work could include stretching of the top membrane after separation, and dissipation within the two-sided medical tape fixing the bottom membrane to the peel wheel. Example stress/strain curves for wet cellulose membranes are shown in Figure S6. The strain corresponding to a wet-peel force of 40 N/m was less than 1 %, and the corresponding work of stretching was only 0.13 N/m when the overall wet-peel force was 40 N/m. Table 1 compares the mechanical properties of treated and untreated membranes. For the < 1 % membrane strains in the wet-peel experiments, changes in membrane properties due to surface treatments should have little impact on the membrane stretching contributions to peel work.
The peel forces for pressure sensitive adhesive tapes are a strong function of peeling rate because much energy is consumed deforming the thick (typically 100 μm) viscoelastic adhesive layer (Satas 1989). By contrast, the adhesive layers in our cellulose laminates are very thin (3–30 nm), mimicking fiber-fiber joints in paper. Figure 6 shows two examples of peel force versus peel rate curves. The thicker adhesive layer showed some rate effects, whereas the very thin layer did not. If we assume that the water content of the adhesive layer was 50 %, the corresponding thickness of the wet adhesive layers was about twice the dry PVAm coverage, 15 nm for thicker layer and 3 nm for the thin layer. Virtually all of our work has employed a fixed peeling rate of 20 min/min.
Figure 7 shows four replicates of raw peel-force versus displacement curves, as well as the mean and standard deviation plots. The mean peel force is calculated from the user-defined horizontal section of the curve; the values for the curves in Figure 7 were 32, 27, 28, and 29 N/m. The results in Figure 7 reveal two types of variation. Focusing on the steady-state portions beyond a displacement of 5 mm, each curve shows significant noise around a mean value. In addition, there is laminate-to-laminate variation. Normally we perform triplicated measurements and the steady-state portions of the curves are averaged and the curve-to-curve variation is a used as a measure of experimental error. Flawed experiments are usually obvious from the absence of a horizontal section in the peel curve. Such results are rejected and additional samples are measured.
One of the most useful applications of wet-peeling is comparing the efficacy of wet-strength polymers at the same coverage. Table 2 summarizes unpublished early work (2003) where we compared some common polymers used in papermaking. None of the polymers gave significant wet adhesion with untreated regenerated cellulose membranes. On the other hand, with oxidized cellulose, PVAm gave adhesion values of 42 N/m. PAE laminates were weaker, however, PAE gives stronger laminates than PVAm if the laminates are heated, promoting crosslinking and grafting. It is not surprising that CMC (applied without calcium ions and elevated temperatures) and PolyDADMAC do not impart wet strength. PEI (polyethyleneimine) is known to give some wet strength (Espy 1995). The CPAM had a slight strengthening effect on TEMPO oxidized cellulose – similar results have been reported for handsheets (Saito and Isogai 2007).
Comparing cellulose surface treatments
Wet-peeling is also adept at showing how wet-strength resins respond to cellulose surface pretreatments. Figure 8 shows the influence of TEMPO oxidation time on the wet-peel force with PVAm. Adhesion increases with oxidation time, reflecting an increase of the concentration of carboxyl and aldehyde groups. The aldehyde groups are grafting sites for PVAm (see chemistry in Figure 1) whereas the carboxyl groups promote PVAm adsorption at the cellulose/water interface. However, these experiments employed the direct application procedure, giving a constant, high PVAm content in the laminates, removing the extent of PVAm adsorption as a factor. With conventional handsheet studies, the coverage of adsorbed PVAm would increase with oxidation time because of the increased carboxyl content. Finally, at very long oxidation times, cellulose degradation leads to membrane failure during the wet-peeling.
The polymer solutions contained 1 mM NaCl adjusted to pH 7 and direct application method gave a coverage of 11 mg/m2. Laminates were dried at 23 °C in 50 % RH. The membranes were modified by TEMPO mediated oxidation.
Table 3 compares adhesion results from experiments comparing three wet-strength resin types (PAE, GCPAM, and PVAm) and two types of cellulose treatments, TEMPO oxidation and CMC treatment. In these experiments the resins were applied by adsorption. Therefore the surface treatment can impact both the coverage of resin in the laminate and the extent and type of bonding between the resin and the membrane. The results in the first three rows show that without a wet-strength resin, TEMPO oxidation gives some increase in wet strength, presumably due to the formation of hemiacetal bonds between the membranes. Similar results from wet handsheet testing have been reported (Jaschinski et al. 2003, Saito and Isogai 2005).
Unbeaten, bleached softwood pulp was treated with 2 wt% polymer based on dry fiber. The membranes were saturated with adsorbed polymer. The laminates were dried at room temperature, whereas the handsheets were dried at 120 °C. The handsheet results were published recently (Gustafsson et al. 2017).
With PAE, both TEMPO oxidized and CMC-treated membranes gave high strengths, whereas laminates from untreated membranes were weak, reflecting the lack of carboxyl groups on the regenerated cellulose membrane surfaces.
PVAm was most effective with TEMPO oxidized membranes, whereas the CMC-treated membrane laminates were only half as strong as the TEMPO oxidized membrane laminates – see Table 3. The amine groups on PVAm interact with CMC by forming polyelectrolyte complexes (Feng et al. 2007a, Gustafsson et al. 2016) whereas with drying, amines form covalent imine and aminal linkages with aldehydes.
The glyoxalated cationic polyacrylamide, GCPAM, was the only polymer to improve the wet-strength of laminates made with untreated regenerated cellulose membranes. TEMPO oxidation gave a small improvement whereas the CMC-treated membranes, treated with GCPAM, gave the highest wet-peel values in Table 3.
The contribution of aldehyde groups to wet-adhesion was illustrated by reducing the aldehydes on the TEMPO oxidized membrane before laminate formation. With PAE, removal of aldehydes had little impact because azetidinium groups on PAE resins do not react with aldehydes. On the other hand, the results in Table 3 show that TEMPO oxidized membranes, laminated with adsorbed PVAm lose about 2/3 of their strength when the aldehydes are reduced. This example shows how the wet-peel experiments can be used to test adhesion mechanistic hypotheses.
For a given fiber type and sheet structure, there is a strong correlation between once-dried wet-peel and wet tensile index. Figure 9 shows the once-dried wet peel force as a function of handsheet wet tensile index for the results in Table 3. In spite of the wide range of fiber/membrane treatments and wet-strength resin chemistries, wet-peel force measured with laminates increases with wet tensile index measured on handsheets. Given the speed, ease, and reproducibility of wet-peel measurements, it might be tempting to replace handsheet testing completely. We do not recommend this because fiber treatments and polymers that give exceptional wet strength also can give extensive formation problems due to fiber flocculation. The roles of formation, retention, and fibrillation (beating) require paper testing.
Wet-strength resin coverage (thickness)
The influence of wet-strength resin coverage (resin mass/joint area) on wet cellulose adhesion is easily probed by wet-peeling laminates prepared using direct application of resin solution. Figure 10 compares three data sets showing once-dried wet-peel force as a function of coverage for PVAm laminated oxidized cellulose membranes. The PVAm coverage is expressed as mg of dry polymer per square meter of laminate. If the density of the dry polymer is 1 g/mL, the dry polymer thickness in nm is equal to the coverage. For example, a 10 mg/m2 dried uniform film, coating a smooth surface has a thickness of about 10 nm. Note that the very highest coverage in Figure 10 correponds to a dry adhesive thickness of 100 nm, whereas a typical pressure sensitive adhesive layer thickness is 100 μm (Satas 1989). Therefore in the context of the broader adhesive technologies, all the results in Figure 10 correspond to thin adhesive layers.
For our standard direct application procedure, the volume of polymer solution (15 μL) and the laminate bonded area (10 cm2) are constant. Therefore in going from 0.01 mg/m2 to 1000 mg/m2, the polymer concentration ranged from 0.7 mg/L to 7 g/L. Thus the solutions of the applied polymer ranged from the dilute, low viscosity regime to the concentrated, high viscosity regime.
The major trend in Figure 10 is that wet-peel force increases with adhesive coverage. The two data sets and the individual experiments were obtained by three different researchers over a twelve year period. Although the PVAm molecular weights and the laminate pressing pressures varied, the results show reasonable agreement. There is no papermaking process that can apply wet-strength resin by direct application to the fiber-fiber joint. Therefore the utility of results such as those in Figure 10 are two-fold. First, one can compare different resins at exactly the same resin content and second, results from thicker layers (>10 nm) show the contributions of polymer-polymer cohesive interactions to the overall adhesion.
Never-dried adhesion: the chemical contribution to wet web strength
On some papermachines wet-web strength is an important requirement for efficient operation. Wet-web strength depends upon water content, the sheet structure and the adhesion/friction properties of fiber-fiber contacts (Page 1993). Polymeric additives have the potential to influence water content (drainage), sheet structure (formation) and possibly fiber-fiber adhesion. Never-dried wet-peel gives insight into the contribution of polymers to adhesion without the complications of varying paper sheet formation and sheet structure. Figure 11 shows the influence of cellulose pretreatment on the never-dried wet adhesion of cellulose membranes laminated with gyloxlated cationic polyacrylamide, GCPAM. In all cases wet-peel strength increases with the solids contents. The CMC-treated membranes gave spectacular never-dried wet adhesion, whereas TEMPO oxidized membranes showed little improvement compared to untreated membranes. We propose that adhesion is due to polyelectrolyte complexation between GCPAM and the surface CMC layer. We have shown that complexes formed between PVAm and CMC in solution give good once dried wet adhesion (Feng et al. 2007b) and have also shown that high never dried adhesion is achieved using CMC-treatment and PVAm (Gustafsson et al. 2016).
With respect to the never-dried adhesion measurements, we have not settled on a standard procedure for sample preparation. The key property is the laminate water content. We recently reported never-dried wet-peel results where the lamination step was performed with a 2.4 kg hand roller pressing the laminates, sandwiched between blotters – water content was varied by blotter exposure time (Gustafsson et al. 2016). By contrast, the laminates for the results in Figure 11 were dewatered in a Carver press at 323 kPa. The water content was varied by adjusting the pressing times between 0.1 and 10 minutes.
Our motivation herein is to promote the wet-peeling method for evaluating and comparing paper wet-strength resins. The most compelling application of wet-peeling is the “head-to-head” comparison of a series of wet-strength resin candidates using the once-dried method. With direct application, the comparison is made with exactly the same coverage (i. e. the mass of polymer/joint area). With adsorption application, the peel-force reflects both the ability of the polymer to adsorb and then to contribute to adhesion. Although adsorption application more closely mimics the situation in papermaking, it is difficult to determine the amount of wet-strength resin adsorbed on the membrane surfaces.
Our wet, regenerated cellulose membrane laminates are physical models for the wet fiber-fiber joints in papermaking. The utility of any model is limited by the extent to which the model mimics its target. From the perspective of chemical composition, bleached kraft pulp fibers and the membranes are both mainly cellulose. The pulp will include some hemicellulose, whereas the membranes should not. Of course, fiber morphology is much more complex and diverse than is the case for the dialysis membranes. The pore volume fraction of the membranes, as judged by the water content, is about 50 %, which is less than most fiber pore volumes based on water retention values. Pulp fiber wall pore size estimation is complex issues and results depend upon the method. The introduction to Andreasson’s paper gives a good summary of the issues (Andreasson et al. 2003). Typical radii estimates for bleached kraft pulp fibers are 10 nm as estimated by NMR (Maloney et al. 1999). On the other hand, the accessibility of fiber pores to soluble polymer suggest pore throat radii of 50 nm (Andreasson et al. 2003). By comparison, the effective pore size of dialysis membranes are smaller than those in pulp fibers; a 12 kDa pullulan molecular weight standard has a radius of about 3 nm (Dubin and Principi 1989). This analysis suggests that intermediate molecular weight wet-strength resins that can diffuse into fiber walls are unlikely to enter the dialysis membranes. The mechanical properties of individual membranes were little changed by polymer treatment (see Table S1), further supporting the claim that our PAE, PVAm and GCPAM polymers did not penetrate the wet cellulose membrane pores.
Our adhesion results suggest that untreated regenerated cellulose membranes have too few surface functional groups to promote grafting and adhesion. Regenerated cellulose fibers display the same behaviors. We recently have shown that handsheets prepared with regenerated cellulose (Lyocell) fibers treated with 2 % PAE resins had a low wet tensile index of 1.6 Nm/g, whereas when the fibers were pretreated by TEMPO oxidation, the wet tensile increased to 4.1 Nm/g (Gustafsson et al. 2017).
We propose that never-dried wet-peeling measurements give insight into the contributions of polymers to wet-web strength during papermaking. Historically, web-web strength on papermachines was explained by a fiber-fiber friction induced by capillary forces (Campbell 1933, Lyne and Gallay 1954, Page 1993). However, this view has been recently challenged (van de Ven 2008). The influence of polymer additives on wet-web strength is complicated. The most likely positive contributions are increased fiber-fiber wet adhesion/friction and increased sheet solids content (i. e. better drainage). Wet-web strength measurements are tedious, and meaningful comparisons require a series of measurements as functions of solids contents. Never-dried wet-peeling gives a direct measure of the adhesion contributions of polymer additives and is easily performed as a function of water content (see Figure 11 as an example). The application of never-dried wet-peeling to elucidate wet-web strengthening mechanisms is illustrated by the following example. PDADMAC is known to increase drainage rates in some types of pulp suspensions (Hubbe 2000). On the other hand, PDADMAC gave no never-dried wet adhesion on TEMPO oxidized cellulose membranes. This means that improved drainage and press dewatering are the only mechanisms by which PDADMAC is likely to enhance wet-web strength.
Finally, for more academic and mechanistic studies, wet-peeling has proved to be useful in evaluating more exotic joint structures such as layer-by-layer (Feng et al. 2009), microgels (Wen and Pelton 2012), and colloidal polyelectrolyte complexes (Feng et al. 2007b).
Laminated wet, regenerated cellulose membranes serve as a physical model for fiber-fiber joints in wet paper. Measurements of the wet delamination force (i. e. wet-peel measurements) augment conventional laboratory wet tensile strength measurements by:
Giving comparisons of wet-strength polymers at the same content of polymer in the laminate joint without the influences of varying fines contents, formation or paper density.
Providing measurement of both the wet-strength of cured, dried joints, and the strength of never-dried joints (i. e. analogous to wet-web strength).
Showing the influence of fiber surface chemistry, including oxidation and the presence of firmly bound polymers.
Permitting the evaluation of more exotic joint structures, including layer-by-layer assemblies, microgels and colloidal polyelectrolyte complexes.
Dr. Anton Esser, BASF Ludwigshafen and Joel Soucy, BASF Canada, are acknowledged for useful discussions. Heera Marway is thanked for help with AFM experiments.
Campbell, W. B. The Cellulose-Water Relation in Papermaking. Forest Service Bulletin, Dept. of Interior, Canada, (84), 1933. Google Scholar
Dementev, N., Feng, X., Borguet, E. (2009) Fluorescence Labeling and Quantification of Oxygen-Containing Functionalities on the Surface of Single-Walled Carbon Nanotubes. Langmuir 25(13):7573–7577. CrossrefPubMedWeb of ScienceGoogle Scholar
DiFlavio, J.-L. The Wet Adhesion of Polyvinylamine to Cellulose. Chemical Engineering. McMaster University, Hamilton, Canada, 2013. Google Scholar
DiFlavio, J. L., Bertoia, R., Pelton, R., Leduc, M. (2005) The Mechanism of Polyvinylamine Wet-Strengthening. In: Advances in Paper Science and Technology: Transactions of the 13th Fundamental Research Symposium. Ed. SJ, I. A. Cambridge, UK. pp. 1293–1316. Google Scholar
Dubin, P. L., Principi, J. M. (1989) Failure of Universal Calibration for Size-Exclusion Chromatography of Rodlike Macromolecules Vs. Random Coils and Globular Proteins. Macromolecules 22(4):1891–1896. CrossrefGoogle Scholar
Espy, H. H. (1995) The Mechanism of Wet-Strength Development in Paper: A Review. Tappi J. 78(4):90–99. Google Scholar
Feng, X., Zhang, D., Pelton, R. (2009) Adhesion to Wet Cellulose – Comparing Adhesive Layer-by-Layer Assembly to Film Casting. Holzforschung 63:28–32. Google Scholar
Gurnagul, N., Page, D. H. (1989) The Difference between Dry and Rewetted Zero-Span Tensile-Strength of Paper. Tappi J. 72(12):164–167. Google Scholar
Gustafsson, E., Pelton, R., Wågberg, L. (2016) Rapid Development of Wet Adhesion between Carboxymethylcellulose Modified Cellulose Surfaces Laminated with Polyvinylamine Adhesive. ACS Appl. Mater. Interfaces 8(36):24161–24167. PubMedWeb of ScienceCrossrefGoogle Scholar
Gustafsson, E., Wang, Z., Polat, O., Ostendorf, W. W., Sheehan, J. G., Pelton, R. (2017) Towards Wet Resilient Paper – Fiber Modifications and Test Method Development. In: 16th Fundamental Research Symposium. Oxford. pp. 865–894. Google Scholar
Kinloch, A. J. (1982) The Science of Adhesion.2. Mechanics and Mechanisms of Failure. J. Mater. Sci. 17(3):617–651. Google Scholar
Laine, J., Lindstrom, T., Nordmark, G. G., Risinger, G. (2000) Studies on Topochemical Modification of Cellulosic Fibres Part 1. Chemical Conditions for the Attachment of Carboxymethyl Cellulose onto Fibres. Nord. Pulp Pap. Res. J. 15(5):520–526. CrossrefGoogle Scholar
Liu, Z., Choi, H., Gatenholm, P., Esker, A. R. (2011) Quartz Crystal Microbalance with Dissipation Monitoring and Surface Plasmon Resonance Studies of Carboxymethyl Cellulose Adsorption onto Regenerated Cellulose Surfaces. Langmuir 27(14):8718–8728. Web of SciencePubMedCrossrefGoogle Scholar
Liu, J., Pelton, R., Obermeyer, J. M., Esser, A. (2013) Laccase Complex with Polyvinylamine Bearing Grafted TEMPO Is a Cellulose Adhesion Primer. Biomacromolecules 14:2953–2960. PubMedWeb of ScienceCrossrefGoogle Scholar
Lyne, L. M., Gallay, W. (1954) Measurement of Wet Web Strength. Tappi J. 37(12):694–698. Google Scholar
Maloney, T. C., Laine, J. E., Paulapuro, H. (1999) Comments on the Measurement of Cell Wall Water. Tappi J. 82(9):125–127. Google Scholar
Obokata, T., Isogai, A. (2007) The Mechanism of Wet-Strength Development of Cellulose Sheets Prepared with Polyamideamine-Epichlorohydrin (PAE) Resin. Colloids Surf. A 302(1–3):525–531. CrossrefWeb of ScienceGoogle Scholar
Obokata, T., Yanagisawa, M., Isogai, A. (2005) Characterization of Polyamideamine-Epichlorohydrin (PAE) Resin: Roles of Azetidinium Groups and Molecular Mass of PAE in Wet Strength Development of Paper Prepared with PAE. J. Appl. Polym. Sci. 97(6):2249–2255. CrossrefGoogle Scholar
Page, D. H. (1993) A Quantitative Theory of the Strength of Wet Webs. J. Pulp Pap. Sci. 19(4):175–176. Google Scholar
Pelton, R., Ren, P. R., Liu, J., Mijolovic, D. (2011) Polyvinylamine-Graft-TEMPO Adsorbs onto, Oxidizes and Covalently Bonds to Wet Cellulose. Biomacromolecules 12:942–948. PubMedWeb of ScienceCrossrefGoogle Scholar
Pesika, N. S., Tian, Y., Zhao, B., Rosenberg, K., Zeng, H., Mcguiggan, P., Autumn, K., Israelachvili, J. N. (2007) Peel-Zone Model of Tape Peeling Based on the Gecko Adhesive System. J. Adhes. 83(4):383–401. CrossrefWeb of ScienceGoogle Scholar
Saito, T., Isogai, A. (2005) A Novel Method to Improve Wet Strength of Paper. Tappi J. 4(3):3–8. Google Scholar
Satas, D. (1989) Peel. In: Handbook of Pressure Sensitive Adhesive Technology, Ed. Satas, D., Van Nostrand Reinhold, New York. Google Scholar
Skowronski, J., Bichard, W. (1987) Fibre-to-Fibre Bonds in Paper. Part 1. Measurement of Bond Strength and Specific Bond Strength. J. Pulp Pap. Sci. 13(5):165–169. Google Scholar
Taylor, D. L. (1968) Mechanism of Wet Tensile Failure. Tappi J. 51(9):410–413. Google Scholar
Yang, D. Degradable Wet Cellulose Adhesives. Ph.D. Thesis, Chemical Engineering McMaster. Hamilton, Canada, 2018. Google Scholar
The online version of this article offers supplementary material (https://doi.org/10.1515/npprj-2018-0013).
About the article
Published Online: 2018-08-18
Published in Print: 2018-12-19
Some measurements were performed in the McMaster Biointerfaces Institute funded by the Canadian Foundation for Innovation. R.H.P. holds the Canada Research Chair in Interfacial Technologies. BASF Canada is acknowledged for funding this project through a grant to R.P. entitled “Understanding Cellulose Interactions with Reactive Polyvinylamines”.
Conflict of interest: The authors declare no conflicts of interest.