Increased recyclability of wet strengthened liquid packaging board, through synergetic e ﬀ ects of combining CMC and PAE – A case study in full scale

: There is an ever-increasing demand for renewable, recyclable and biodegradable packaging solutions, from consumers, producers of goods, and producers of packaging materials. Closing the material loop and increasing recyclability of ﬁ ber-based package materials is one of the keys to move forward towards a more sustainable future. While the recyclability rates of ﬁ ber-based packaging are high, packaging boards with high wet-strength can pose problems due to problems with repulping. This manuscript investigated the possibilities to reduce the use of polyamide-epichlorohydrin resins (PAE) by supporting the process system with carboxy-methylated cellulose (CMC) through machine trials in mill scale production. The focus of the investigation was on the wet-strength of the board product and the repulpability value. The results given from the full-scale trial were positive, indicating a potential in using CMC while reducing the PAE addition. The results showed that wet tear strength and wet tensile strength of the board were maintained, while higher repulpability rates were given, encouraging better recyclability of the board material. This will be bene ﬁ cial to the environment both with lower use of non-renewable chemicals and possibilities for higher degrees of recycling of the board products.


Introduction
The need for sustainable packaging solutions that are renewable, biodegradable, and recyclable is growing rapidly among consumers, goods producers, and packaging material manufacturers, and will be necessary to fulfil the 17 goals of sustainable development (United Nations 2015).To advance towards a more sustainable future, it is essential to close the material loops and increase the recyclability of fiber-based packaging materials.Although fiber-based packaging has high recyclability rates, wet-strong packaging boards can present challenges to repulping, posing a problem in this regard.Strength performance of paper products are mainly related to inter-fiber bonded area in the fiber network and to some extent also individual fiber strength.Although there is still a debate as to the specific mechanisms of the bonds, there are no arguments of the bonds being reversible in water (Hirn and Schennach 2015;Wågberg 2022;Wohlert et al. 2022).To enable the formation of stable networks between fibers, it is sometimes necessary to modify the fiber surface through supportive refining.Regardless of the initial pulping process, further refining can be performed to improve surface accessibility and alter paper properties.Refining involves mechanically treating the fibers, causing fibrillation, increased flexibility and partial structural changes by shortening their length (Paulapuro 2000).This process disentangles the fibers through stretching, exposing the fiber surface to some extent (Eklund and Lindström 1990).Additionally, refining procedures enhance the pore structure within the fiber wall, increasing the surface area of cellulose and improving fiber flexibility through delamination of the fiber wall (Laine et al. 2002).Carboxyl groups of cellulosic fibers also plays a major role, impacting various unit operations in pulp and paper manufacturing, as well as affecting the physical properties of the final product.Carboxyl groups contributes to a large part of the surface charge of the single fiber (Lindström 1992), and also affecting the chemical interactions during the manufacturing process.The cationic demand of the fibers is directly affecting the amount of cationic additive that could be used, before charge saturation to both fiber and process system (Eklund and Lindström 1990).It is important to note that surface charges and internal charges of fibers both can be affected, depending on the size of the charge altering agents as well as the morphology of the fiber structure and porosity of the fiber wall.
Since fiber-based materials are characteristically weak in water, certain measures have to be undertaken in order to be able to use fiber-based packages in wet environments.Most commonly, synthetic and/or polymeric wet strength agents (WSA) are added in the wet-end of the papermaking process to improve the wet strength of the final products (Aarne et al. 2013;Lindström 2009).The wet strength can also be increased with surface sizing (Lindström 2009).For a paper product to be considered to have high wet-strength, the ratio between wet and dry strength must be at least 0.15 (Eklund and Lindström 1990).
In a study conducted by Au and Thorn (1995), they proposed that the origin of wet strength resistance in paper products can be attributed to two primary mechanisms.The first mechanism involves the interaction between polymer molecules, known as homo-crosslinking.This process restricts the swelling of the fiber network when exposed to water, often depicted as a protective layer resembling a balloon or a net enveloping the fibrous structure.The second mechanism, co-crosslinking, occurs when the polymers react with the cellulose interface.Co-crosslinking reinforces the fibrous network and is analogous to a "zip lock," holding the fibers together even when they come into contact with water.However, recent research challenges this earlier perspective.According to Onur et al. (2019), wet strength resistance is actually a combination of both these mechanisms.This combined effect serves to both shield and strengthen the fiber network, effectively preventing water intrusion.
The crosslinking process can extend over days or even weeks if the temperature is optimal.It has been demonstrated that the crosslinking and full encapsulation of the Polyamideamine chains around the fibers synergistically occur in the drying section of the paper machine.This implies that elevated temperatures are necessary for the complete execution of the reaction (Aarne et al. 2013).It's important to note that this crosslinking behavior is irreversible, which has implications for the recyclability of the fibrous material, as discussed by Onur et al. (2019).
Polyamide-epichlorohydrin resins (PAE) are by far the most common WSA when producing papers with increased wet strength (Aarne et al. 2013;Eklund and Lindström 1990;Onur et al. 2019).The main downside when using PAE is that the repulpability will also be reduced due to the increase in wet strength (Siqueira et al. 2015).An additional problem of PAE are possible residuals originating from side-reactions in the production of PAE, such as Adsorbable Organic Halogens (Onur et al. 2019).Aarne et al. (2013) proposed that Wet strength performance can be achieved with additions of CMC without getting the problems related to crosslinking chemistry originating from PAE causing the product to be harder to recycle.They showed that when using a combination of CMC and Polybrene, with CMC first altering the surface charge of the cellulosic fibers, and Polybrene later added and adsorbed to the fibers, wet strength was increased.Siqueira et al. (2015) examined the interactions between CMC and PAE and possible advantages of using a CMC-PAE complex in the production of wet strong paper products.When using CMC and PAE together, they construct complexes that could potentially lower the net charge of the system, compared to only adding PAE to the fiber surface (Siqueira et al. 2015).Laine et al. (2000Laine et al. ( , 2002Laine et al. ( , 2003) ) showed that CMC can modify pulp fibers in a selective way through irreversible adsorption.According to work by Laine et al. (2000Laine et al. ( , 2002Laine et al. ( , 2003)), as well as a study later made by Aarne et al. (2012), CMC was suggested to have a two-fold effect when added onto the pulp fiber; (i) a strong irreversible absorption onto the fiber surface, (ii) thus changing the charge density of the fiber.CMC align with the cellulosic surface of the fibers and through carbohydrate-carbohydrate interactions, creating an outer layer of additional anionic charges.To be able to adhere to the negatively charged fibers, the CMC polymer should have a low enough degree of polymerization, to avoid too high repulsive electrostatic forces, while still being a big enough molecule to stay on the fiber surface (Laine et al. 2000(Laine et al. , 2002(Laine et al. , 2003)).Adding CMC to fiber surfaces has also been shown to increase dry strength of the paper product (Bäckström et al. 2004).The CMC adsorption is dependent on multiple parameters and parameters such as pH, temperature, conductivity, time and CMC specifications must be favourable in order to get efficient attaching.
The purpose of this work was to examine what happens to a commercial board product with respect to wet strength performance and repulpability when reducing the amount of PAE with an addition of CMC.The investigation was conducted in full scale at Stora Enso AB, Sweden.

Full-scale trial
The full-scale trial was conducted at a full-scale board machine at Stora Enso AB, Sweden, running one board grade with high wet strength.The board machine used consist of five layers.Print ply runs with bleached softwood sulphate pulp, the three middle layers with CTMP-pulp and the bottom ply with unbleached sulphate pulp (softwood).No recycled paper is used.The machine speed interval is between 540 and 690 m/min and the grammage span is 290-410 g/m 2 .The idea of the trial was to stepwise lower the dosage of the wet strength agent PAE in the board furnish and then introduce CMC, finally the PAE was again increased with the highest CMC level.The absolute amounts of PAE and CMC, corresponding to 100 % in Table 1, are within the dosage range of 2-4 kg/tonne PAE and 1-3 kg/tonne CMC.The relative amounts related to these are used to evaluate the effects of the additives.Table 1 presents the percentage dosage levels of PAE and CMC used for each sample of the trial.
PAE was added close to the head box in the specific board machine used for the trial and CMC was dosed as early as possible in the process for the print ply.The pulp consistency was 3.8-4 % in this position, and varies little throughout the print ply, until dilution to head box concentration shortly after P4 in Figure 1.Head box concentration varies between 0.2 and 1 %.Wet pulp samples were collected directly after PAE and CMC additions, before the head box.Produced dry paper samples were collected after the machine, at the reel section.The CMC used in the experiments was restricted to one product, a purified technical grade with a Degree of Substitution (DS) of 0.7, medium molecular weight and a viscosity of 200-500 mPa × s for 4 % dryness and 25 °C.The PAE that was used had a molecular weight of 700,000-1000,000 Da.The molecular weight distribution was quite large due to the difficulty to properly measure a crosslinked polymer complex.The time between dosing of CMC and PAE was approximately 40 min to 1 h, as CMC was dosed at the beginning of the system and PAE just before the head box.The volumes in the system are relatively small, so this was assumed as an approximate residence time for the system.Cationic starch originating from potatoes was also used during trials, but it was not assumed to compete with the PAE-CMC, rather, CMC-Starch is a well-known combination in general, but in that case, starch is applied before CMC, and then we assume that the mechanism of building charge layers on the fiber does not apply in the same way, since the starch consumes the anionic charges on the fiber surface before the CMC can interact with the fibers.The starch had a constant dosage of 3-6 kg/tonne in the print ply and 12-18 kg/tonne in the center ply.

Laboratory study
The full-scale study was combined with a laboratory study on the same pulp.Wet bleached pulp samples were extracted from the paper machine, with and without additions of PAE and CMC, and Z-potential was measured on the samples with a Mütek SZP-10 (BTG Instruments AB, Säffle, Sweden).The pulp was taken from the board machine once, at a fixed point, and 10 repetitions of measuring Z-potential were made determining the statistical mean value of the machine, the standard deviation, confidence interval (95 %) and the variance (%) were calculated.This variance was later assumed to be the same for all Z-potential samples, and is displayed throughout the manuscript.
A cross strip at the top of each tambour was collected.This cross strip was divided into 16 sheets.The ten samples carried out at each test point are from the 10 central samples.Three sheets on each edge were removed to avoid edge effects.When it comes to chemical changes in the pulp, the machine's position should not affect it, as the orientation of fibers is more driven by machine speed, inlet velocity, pressure, and concentration.Variations are usually observed at the edges when it comes to strength properties.However, we still chose to centralize the samples.Basic physical board properties were measured on the paper samples, according to the following list; Grammage according to ISO 536:2019, Density according to SCAN P:88-01, and Thickness according to SCAN P88-01.Important mechanical board properties were also measured; Tensile Strength for MD and CD according to ISO 1924-3:2005, Wet Tensile Strength for MD and CD according to ISO 3781:2011, Tear Strength for MD and CD according to TAPPI 496 cm-85, Wet Tear Strength for MD and CD according to TAPPI 496 cm-85.The relative wet tear strength for MD and CD according to ISO 3781:2011 was calculated as the ratio between wet and dry tear strength.
To evaluate the recyclability of fiber-based materials, the method PTS-RH 021/97 was used.The method evaluates recyclability of fiberbased materials according to a set of tests, similar to those used on an industrial scale: disintegration, screening and sheet forming operations.The criteria for recyclability given for the one of the methods categories (II), which is including package materials, stating the recyclability of a specific packaging material is given by the reject rate.<20 % reject means recyclable, 20-50 % means recyclable with development potential and >50 % means not recyclable with conventional methods.Important to note when calculating the reject rate, is the actual fiber content of each sample.The method does not take into account whether the reject consists of fiber, barrier, laminates or other residualsit only calculates the total reject.Samples for the recyclability test were prepared as 2 × 2 cm test pieces which were post-cured in a drying cabinet at 60 °C for 6 days.The post-curing for 6 days in the method is a way of simulating the materials' natural aging and is supposed to reflect the real aging of one year for the product.The post-curing also had the purpose of making sure that full curing of all wet strength resins has taken place.After the post-curing, the samples were conditioned in a standard climate (ISO 187:1990) for 2 days.The solids content was measured with a halogen dryer (MyCal AB, EM-120HR, Sweden).The disintegration (repulping) of the samples was performed for 6 and 20 min (18 000 and 60 000 revolutions) in a standard disintegrator (Zellcheming, ZM V/6/61).The samples were then diluted in tap water (40 °C) to a concentration of 2.5 %.After the disintegration, the samples were diluted to a concentration of 0.5 % and homogenized during 5 min in a standard distributor.Thereafter the samples were screened in a Brecht-Holl fractionator during 5 min with plate holes of a diameter of 0.7 mm.Rejects from the screening were washed with water on a filter paper and dewatered using a Büchner funnel.The dried weight of the reject was determined and the reject rate is calculated according to Equation (1).

Reject rate =
Weight of reject Total sample weight (1) The weights of the reject and the total sample weights in Equation ( 1) correspond to the dried weight of the reject and the dry weight of the initial sample.

Results and discussion
The full-scale trial was performed during September 2022.The purpose of the trial was to investigate the effect given by a step-wise lowering of the wet strength additive (PAE), while simultaneously introducing CMC with increasing dosages.The hypothesis was that the wet physical properties would be maintained within the board specification, which means that the wet/dry tear and tensile strength ratio should be at least 0.6.Wet and dry tear and tensile strength were evaluated in both the machine direction (MD) and the cross direction (CD).The basic board properties that were measured for all trial points are presented in Table 2. Figure 2 shows how CMC addition affect the measured Z-potential at the addition point of CMC and PAE.Laine et al. (2000Laine et al. ( , 2002Laine et al. ( , 2003) ) and Aarne et al. (2013) showed that CMC could be used to modify fiber surfaces under specific process conditions.These conditions are affected by temperature, residence time, pH and conductivity and are important for modification with CMC to be efficient.As indicated by Laine et al. (2000), the bonding of CMC to cellulosic fibers hinges on the cellulose content of the employed pulp.Hence, pulps rich in cellulose content facilitate the adherence of CMC more readily than those derived from CTMP and other wood-based sources.The influence of temperature on the interaction between CMC and fibers indicate that higher temperatures give stronger bonding, with the effect plateauing at temperatures surpassing 120°.The temperature measured post CMC addition and homogenization was lower than the temperature range deemed optimal in Laine et al. 's investigation (2000).Regarding variations in electrolyte concentration, no attachment occurs in an electrolyte-free environment.Throughout the trial, both conductivity and pH were maintained constant, aligning with the recommendations of Laine et al. (2000), which emphasize that pH levels could impact CMC bonding if conductivity is low.An additional pivotal parameter to consider is the dwell-time, as underscored by Laine et al. (2000), to enable effective adsorption of CMC onto fiber surfaces.The research advises a minimum of 60 min for achieving a satisfactory degree of attachment, with extended times potentially offering further benefits.There is, based on this, a chance that CMC did not adhere as effectively as it might have been, and this can certainly be developed further.If the conditions are enough, CMC can attach irreversibly to the surfaces of cellulose, where the binding mechanism is co-crystallization (Laine et al. 2000(Laine et al. , 2002(Laine et al. , 2003)); Aarne et al. (2013).As shown in a pilot scale study by Ankerfors et al. (2013), CMC can attach well to fiber surfaces using a temperature of 95 °C and reaction time of 90 min.These conditions were sufficient for CMC to attach

Property
Mean value well to softwood fibers, although the study also used much higher CMC dosages of 10 kg/tonne.Assuming that CMC could attach to the cellulose surfaces in the present full-scale trials, the level of Z-potential should be constant until a new additive is added to the system, which affects the charge of the fiber surface and thus affects the Z-potential.According to Eklund and Lindström (1990), beating the cellulose fiber is a way to modify the fiber surface, which theoretically would increase the availability for the carboxyl groups to attach through an increased surface area.Overall, the Z-potential is less than the value at the reference point, and these lower values are maintained throughout the process without saturating the fiber surface with charge.As additions of CMC are incrementally increased at trial points 4-6 (Figure 2), the magnitude of the Z-potential also increases, indicating that more CMC is available and also that it stays on the fiber surface.
Figures 3-4 present the result of both dry and wet tear strength, in MD and CD, in relation to the PAE-dosage.The trial points were evaluated with 10 measurements in both directions, the 95 % confidence interval was calculated, showing minimal statistical significance for the tear strength for different additions.Relative values are shown in the figures, where the value for REF CD is set to 1.The dry tear strength is higher than the wet tear strength, as expected.This is true regardless of PAE-dosage in the different trial points, and both dry and wet tear strength in MD and CD were unchanged for different addition strategies, with no significant differences.Although the error-bars for wet tear strength are larger compared to the error-bars for dry strength, showing larger variances in the results for wet samples.
Figure 5 presents the relative wet tear strength in MD and CD for the board, defined as the wet/dry ratio.The error bars were calculated with error propagation from the values of the error bars of wet and dry tear strength.According to the specific board specification, 60 % of dry tear strength should be wet tear strength, in both MD and CD.Some of the bars for MD in Figure 5 are lower than the specification, although this is also true for the reference sample.This means that the actual study is not causing the drop below 60 %, rather the reference in this trial was low.
The results for tear strength with wet papers show that there is an opportunity to lower the PAE dose quite drastically (according to test point 3-6) without having to compromise too much regarding product performance.Considering both the economic aspects and the sustainability aspects that follow with smaller amounts of PAE in   the cardboard production, increased use of CMC will be able to contribute to less emissions of adsorbable organic halogens and dichloropropanols to the waste water, which is very positive.
Figures 6 and 7 show the tensile strength (MD and CD) of the samples from the trials.Relative values are shown in the figures, where the value for REF MD is set to 1. Relative to the wet/dry tear strength ratio, the wet/dry tensile strength is much lower.The reduction of PAE and introduction of CMC does not significantly affect the dry tensile strengths, much like the tear strengths, which shows potential for the possibilities of successfully introducing the method related to strength properties.
However, the results of wet tensile strength in MD and CD are presented in Figure 7, here the effects of the change in wet end chemistry can be observed.Around the positions PP4-5, the wet tensile strength is observed to be lowered.Here CMC was introduced which is assumed to be the reason for this, the CMC does not seem to affect the wet tensile strength in MD for PP4-5.Between PP5 and PP7, there is a clear increase in wet tensile strength in MD, where the strength of the reference was recovered.In CD, wet tensile strength is lower in PP3 where the PAE-addition is at its lowest dosage rate and does not recover as much as in MD, rather stabilizes at a lower level.
The increase in wet sheet tensile strength between PP4-5 and PP6-7 could indicate that CMC contributes to a higher proportion of PAE staying in the sheet, i.e. higher PAE retention, as the PAE addition in PP4-6 is equal.Although the strength increase might also come from the CMC itself, acting by increasing the bonded area between fibers (Ankerfors et al. 2013).These results indicate that CMC to some extent attaches to, and thus modifies the fiber surface and aids in the retention of PAE, this is a suggestion of the mechanism based on previous literature and present experimental results, although more work needs to be conducted to prove this hypothesis.According to Laine et al. (2000) there is a limit to the specific amount of CMC needed for this modification of the fiber surface to occur.A disadvantage of this "threshold" dosage of CMC required for surface modification is the potential risk of higher proportion of CMC in the short circulation on the board machine, which increases the energy consumption for both dewatering and drying.
Figure 8 shows the results of the recyclability measurements together with the addition of PAE.It is clear that the reject ratio from PTS-measurements is related to the additions of PAE.This means that the CMC additions are not that critical for recyclability but still enhances the product performance as shown previously.This is true for both the 6 and the 20 min versions of the recyclability test.Given the   PTS result for both 20 min and 6 min, the recyclability is strongly related to the amount of added PAE.Even with small changes in PAE addition there is a response in reject ratio.According to the results for both 20 min and 6 min, it is also quite clear that CMC addition does not affect the reject rate.Based on the results, there is an opportunity to lower PAE additives using CMC, and still keep the products within the given wet strength specification while increasing recyclability.

Conclusions
The results indicate that additions of PAE can be lowered by additions of CMC, as wet physical paperboard properties were maintained where CMC was added in combination with lower amount of PAE.According to the PTS results, the rejection rate corresponded well to the reduced PAE doses.The possible increase in PAE retention did not adversely affect the reject rate of the PTS measurements.This means that it is possible to build on the concept of CMC and PAE, by using CMC as an aid to maintain the wet physical properties of the board while making it somewhat more recyclable.The results have the potential to contribute positively to the environment in two ways: first, by reducing the consumption of non-renewable chemicals; and second, by increasing the opportunities for enhanced recycling of board products.

Figure 1 :
Figure 1: Flow chart of the print ply of the board machine.

Figure 2 :
Figure 2: Z-potential plotted against CMC and PAE addition in the board machine.Error bars correspond to a 95 % confidence interval.

Figure 3 :
Figure 3: Relative dry tear strength in MD and CD for sheets with CMC and PAE dosages.Error bars correspond to a 95 % confidence interval.

Figure 4 :
Figure 4: Relative wet tear strength in MD and CD for sheets with CMC and PAE dosages.Error bars correspond to a 95 % confidence interval.

Figure 5 :
Figure 5: Relative wet tear strength in both MD and CD, specific product requirement indicated at 60 %.Error bars were calculated with error propagation from the corresponding error bars of wet and dry tear strengths and represent a propagated 95 % confidence interval.

Figure 6 :
Figure 6: Relative dry tensile strength for MD and CD with CMC and PAE dosages.Error bars correspond to a 95 % confidence interval.

Figure 7 :
Figure 7: Relative wet tensile strength for MD and CD with CMC and PAE dosages.Error bars correspond to a 95 % confidence interval.

Figure 8 :
Figure 8: : Reject rate (%) for both 6 and 20 min together with CMC and PAE dosages.Error bars correspond to a 95 % confidence interval.

Table  :
Dosages of PAE and CMC, the absolute amounts are machine specific.

Table  :
Results from basic testing of the investigated board grade, the mean values and  % confidence intervals are displayed.