TEMPO Mediated Oxidation Optimization on Thermomechanical Pulp for Paper Reinforcement and Nanomaterial Film Production
David Myja, Éric Loranger,* and Robert Lanouette
The 4-acetamido-TEMPO mediated oxidation system is well known on pure cellulose, but further investigation on more complex cellulose sources, such as thermomechanical pulp that is also composed of hemicelluloses and lignin, is needed. The reaction on wood pulp allows improvement in paper strength and nanofibril material production. However, the effects of the reaction parameters are not known enough to scale up the oxidation. With the help of two experimental designs, the chemical amounts, reaction temperature and time, sodium hypochlorite injection time, and nanofibril dispersion time were all studied. During the experiments, it was possible to observe pulp bleaching or delignification. Increasing the 4-acetamido-TEMPO amount promoted the pulp oxidation. In contrast, a large excess of sodium bromide and sodium hypochlorite was advantageous for pulp bleaching and delignification. High temperature favored the oxidation but delignification was induced by the oxidation. For the paper reinforcement, chemical amounts were optimized according to the end user needs. For nanomaterial production, both oxidation and delignification were needed. The reaction had to generate a significant delignification and increase pulp carboxyl content higher than 1600 mmol/kg to be able to produce nanofibril material. The results showed new leads on the various times required for future industrial implementation.
Keywords: 4-Acetamido-TEMPO; Conditions optimization; Paper strength; Nanofibril material
Contact information: Lignocellulosic material research center, Université du Québec à Trois-Rivières, P.O. Box 500, Trois-Rivières, Québec, Canada, G9A5H7; *Corresponding author: email@example.com
2,2,6,6-Tetramethylpyperidine-1-oxyl (TEMPO) or 4-acetamido-2,2,6,6-tetra-methylpyperidine-1-oxyl (aTEMPO) mediated oxidation has been long studied on polysaccharides (de Nooy et al. 1995; Bragd et al. 2001). The TEMPO or aTEMPO, sodium bromide (NaBr), and sodium hypochlorite (NaOCl) oxidation systems have about the same oxidative effect and are well known for oxidizing primary alcohols into aldehydes and ultimately to carboxyl groups (Okita et al. 2010; Isogai et al. 2011), thus introducing carboxylic groups on cellulose fibers. This oxidation is easily controllable to achieve low carboxyl content and best used as paper reinforcement (Ma et al. 2009, 2011), or a high carboxyl content, suitable for nanocellulose production (Saito et al. 2007; Fujisawa et al. 2011; Fukuzumi et al. 2013). Saito and Isogai (2004) started by studying the influence of reaction time and sodium hypochlorite amount on carboxyl content and cellulose polymerization degree. Saito et al. (2006) proposed a reaction scheme of the TEMPO-mediated oxidation on cellulose, and Iwamoto et al.(2010) even studied other TEMPO derivatives for the cellulose oxidation. However, most such studies have been mainly focused on native cellulose. Only a few researchers have investigated the TEMPO-mediated oxidation effect on other cellulose sources such as thermomechanical pulp (TMP).
In addition to cellulose, TMP also contains various amounts of hemicelluloses and lignin. Okita et al. (2009) analyzed the composition of the softwood TMP before and after the TEMPO-mediated oxidation. Their results showed that pulp weight decrease is directly related to the dissolution of hemicelluloses and lignin in the reaction medium (water). Sodium hypochlorite is a well-known TMP bleaching agent that removes surface lignin. However, their results have shown a higher pulp weight decrease with the TEMPO oxidation than with a simple sodium hypochlorite bleaching. This observation was confirmed by Ma et al. (2012), who focused on the aTEMPO oxidation effect on lignin. The aTEMPO oxidation on TMP was found to be effective for paper reinforcement by increasing the tensile strength (Ma et al. 2011; Myja et al. 2016). Yet, the oxidation also brings about a decrease in the fiber intrinsic strength (Ma et al.2009; Ma and Zhai 2013), which means that an optimization is needed to obtain the strongest paper possible. TEMPO-oxidized nanofibril production has also been investigated on mechanical pulp (Kuramae et al. 2014) and were a bit longer than cellulose nanofibrils.
In this paper, the authors have continued this effort to further understand the reaction on TMP with many aTEMPO oxidation conditions. First, an optimization for low oxidation conditions was made to increase conventional paper strength. A second optimization for high oxidation conditions was realized to observe nanofibril film properties. Both optimizations were undertaken as preliminary work in anticipation of a future scale-up, from laboratory to pilot plant scale, of TMP aTEMPO oxidation.
The TEMPO-mediated oxidation was made on a secondary unbleached softwood thermomechanical pulp. The pulp was from Kruger S.E.C. (Trois-Rivières, Canada) and was refined to a Canadian Standard Freeness (CSF) of approximately 150 mL in the authors’ research center pilot refining unit (Valmet CD300 refining system; Valmet, Espoo, Finland). The chemical, which were used as received, included sodium hypochlorite from Sigma-Aldrich (Oakville, Canada), aTEMPO for the reaction from Chemos (Regenstauf, Germany), sodium bromide from Thermo Fisher Scientific Chemicals Inc. (Ward Hill, USA), and hydrogen peroxide from Thermo Fisher Scientific Chemicals Inc. (Fair Lawn, USA).
As shown in Fig. 1, aTEMPO reactions were carried out in a 5-L glass reactor disposed in a thermostated bath. A conventional stirrer was used during the reaction at 260 rpm. A pH controlling system was also used with sodium hydroxide and hydrochloric acid buffer solutions at 0.1 M to maintain the pH at 10 at all times. Finally, a metering pump was used for sodium hypochlorite addition to the reactor.
For every oxidation, 30 g of dry pulp was used. The pulp was first disintegrated in deionized water at a consistency of 1.5%. At the same time, the needed mass of aTEMPO and sodium bromide were dissolved in a small amount of deionized water before their addition. Pulp and chemicals were then introduced into the reactor, and deionized water was added to get the consistency of 1% in a 3-L final volume. The temperature and pH of the pulp were then adjusted. As sodium hypochlorite was mandatory for the aTEMPO system, the reaction time only started when the first drop was added. To stop the reaction after the required time, 100 mL of hydrogen peroxide of 3% concentration was poured into the reactor to neutralize the sodium hypochlorite excess. The oxidized pulp was then filtered, and all the recovered pulp was weighed. A small amount of the pulp was used to determine the pulp dryness to calculate the oxidation yield. Finally, the pulp was stored at 6 °C until further use.
Fig. 1. aTEMPO mediated oxidation experimental setup
Handsheet preparation and testing
Pulp obtained after low oxidation was used to make handsheets according to the TAPPI T205 sp-02 (2006) standard. The ISO brightness was measured according to the TAPPI T571 om-03 (2006) standard, and the tensile and tear tests were made according to the TAPPI T220 sp-01 (2001) standard.
Carboxylic groups measurement
The carboxylic groups in the pulp after the oxidation were measured by conductimetric titration following Beatson’s (1992) method. The pulp was first protonated twice with hydrochloric acid at 0.1 M during 40 min. After filtration and pulp washing with 1 L of deionized water, the pulp was added to 450 mL of salted water (10-3M) prepared with sodium chloride. Then 10 mL of hydrochloric acid at 0.1 M was added before the titration with sodium hydroxide (0.1 M). The pulp carboxylic groups content could then be determined according to the titration curve.
Nanomaterial film preparation and testing
Pulp obtained after the high oxidation was used to make nanomaterial film. A mass of oxidized pulp corresponding to 1.5 g of dry pulp was dispersed in deionized water at 0.5% consistency with a household blender. Each film was made from 50 mL of the dispersion poured into a 10-cm diameter aluminum dish. The dishes were then placed in a forced air enclosure for 48 h at room temperature. After this time, films can be recovered from the dishes without any apparent defects.
Films were then cut in 1.5-cm-wide strips for tensile testing with an Instron 4201 (Instron, Norwood, MA, USA) device at 10 mm/min. Those tests allowed the authors to measure the films’ Young’s modulus, tensile stress, and tensile strain at the breaking point. However, only the Young’s modulus and the tensile strain results will be analyzed because of the strong correlation between Young’s modulus and tensile stress. The light transmission through each film was measured with a Tint Meter Inspector (TM200; Laser Labs Inc., Scituate, MA, USA) that measured the light transmitted from a source to a receiver. This apparatus is normally used by police to evaluate tinted windows on cars.
Experimental design software and conditions
Both experimental designs were made and analyzed with JMP Pro 13 software (SAS Institute Inc., Cary, NC, USA). For each optimization, five factors were studied to observe potential interactions with a manageable number of tests. Both optimizations were made with a response surface design and, more precisely, a central composite design with uniform precision.
For the first optimization, chemical amounts (per g of dry pulp), reaction temperature, and reaction time were studied. For the second optimization, the chemical amounts were still studied for a high amount of sodium hypochlorite and low amount of aTEMPO and sodium bromide. The sodium bromide and sodium hypochlorite amounts were calculated with a ratio of aTEMPO. The second design allowed those ratios to be optimized. In addition, the sodium hypochlorite injection time and the dispersion time for nanomaterial film production were both studied. Such experimental testing will enable the authors to reduce secondary phenomenon at high sodium hypochlorite content, namely pulp bleaching, on TMP. The conditions and factor levels for both designs are given in Table 1.
Table 1. Oxidation Optimization Conditions and Factor Levels
RESULTS AND DISCUSSION
Low Oxidation: Paper Reinforcement
Table 2 shows all of the conditions tested for the low oxidation optimization with the yield, handsheet tensile index, tear index, and ISO brightness obtained. The minimum and maximum of each response were also reported at the bottom of the table. The difference between each minimum and maximum was wide enough to consider that the modifications of aTEMPO oxidation conditions had a significant effect on final results.
Table 2. Tests and Results for the Low Oxidation Optimization
To analyze all of those results, models were calculated by the step-by-step method with a p-value of 0.05. The estimated coefficient and a summary of some statistics are reported in Table 3 for each response model.
Table 3. Estimated Coefficient and Adjustment Information for a Model with a Significant Probability at 95% for the Low Oxidation Optimization
An initial model was calculated for the tensile index with all the experiments, but an important lack of fit was observed for this model. The authors’ investigation has identified the tensile index value obtained from trial number 20 as responsible for this lack of fit. That test condition had an extreme level of temperature that was not repeated elsewhere in the experimental design. To achieve a valid model, other calculations were made only for the tensile index by excluding trial 20. The estimated coefficients of both tensile index models were nearly the same, but the statistics of the second model were far better. The corrected tensile index model was used for the results analysis reported here.
According to the calculated models, temperature was an important condition in aTEMPO oxidation for paper reinforcement. It had a significant effect on each studied response, a significant interaction with all other conditions for at least one response, and a quadratic effect on all paper properties. As explained previously, the aTEMPO oxidation on TMP is in competition with pulp bleaching, and both phenomena are influenced by the temperature (Sain et al. 1997; Sun et al. 2005). For the same reason, the sodium hypochlorite amount had significant effects and interactions on all responses, but more so on paper properties. The aTEMPO and sodium bromide amounts have naturally significant effects on the responses because they are directly linked to the aTEMPO oxidation kinetic (Sain et al. 1997). Finally, the reaction time did not have an important impact on responses in comparison to the other conditions. Ma et al. (2012) showed that increasing the reaction time should increase the fiber delignification. However, the time variation in the experimental design did not allow the authors to observe such a delignification effect. This was mainly related to the long time period used.
To optimize the TEMPO oxidation conditions, the condition effects on each response are represented in Fig. 2. The reaction time had a rather small impact on paper properties, so it will be kept to the lower value studied. This decision was made with industrial implementation in mind, where a low reaction time can be more beneficial. At this point, other conditions cannot be fixed because of their important impact and interactions between them. However, a few observations can be made with respect to Fig. 2. Increasing the aTEMPO amount at fixed sodium bromide and sodium hypochlorite increased the oxidation yield, but it also reduced the tensile index and ISO brightness, which was expected because a high aTEMPO content helped the sodium hypochlorite be used more for the aTEMPO oxidation rather than for fiber delignification. The same observation was made for sodium bromide, but a large excess led to an excess of sodium hypobromite.
Fig. 2. Representation of condition effects on each response for the low oxidation optimization
Usually, the sodium hypobromite activates the aTEMPO to oxidize the fibers selectively (Saito et al. 2006), but sodium hypobromite is an oxidant and could oxidize fibers directly, as observed from the decreasing yield at high sodium bromide amounts. As mentioned previously, an excess of sodium hypochlorite induced a higher bleaching effect, which promoted pulp delignification. Delignification has consequences, such as a drop in reaction yield, an increase in tensile strength from better fiber-to-fiber bonds, and a decrease in the fibers’ intrinsic strength, which induce overall tear strength decrease.
Finally, the temperature presents an optimum value at approximately 47 °C for ISO brightness. This temperature is in the lower range of known values for pulp bleaching (Sain et al. 1997). However, the temperature had many interactions with other conditions, so it was the last condition optimized.
An important observation for the optimization was the interaction between sodium bromide and sodium hypochlorite. However, any optimum value could be fixed for the aTEMPO amount, which is why the interaction was studied for different aTEMPO amounts.
Figures 3, 4, and 5 present isoresponse profiles of tensile and tear index, yield, and ISO brightness according to sodium bromide and sodium hypochlorite at various aTEMPO quantities. As expected, the tensile strength and ISO brightness increased with a high sodium hypochlorite and a low sodium bromide content, because the sodium hypochlorite bleaching was responsible for this effect. In contrast, increasing the sodium bromide and decreasing sodium hypochlorite favored the oxidation over bleaching. This condition induced a yield increase, but it induced a drop in tensile and brightness as well. The tear index was negatively affected by both bleaching and oxidation because of the decrease in the fibers’ intrinsic strength.
The highest tear strength was obtained at low chemical amounts. Moreover, the tear strength decrease was smaller due to increased sodium bromide rather than sodium hypochlorite. This meant that bleaching was more harmful to fiber intrinsic strength than aTEMPO oxidation. In the same way, increasing the aTEMPO amount resulted in a decrease of tensile index and brightness in favor of a yield with no impact on tear strength. However, for all aTEMPO levels, the same point could be chosen as an optimum of tensile and tear index with a good brightness value and without a huge decrease in yield. The chosen point was at 0.24 mmol/g of sodium bromide and 2.95 mmol/g for sodium hypochlorite. Depending on what is preferable for the end user, it was possible to adapt the aTEMPO amount to increase either the tensile strength and brightness or treatment yield.
Fig. 3. Isoresponse profiles of (1) tensile and tear index and (2) yield and ISO brightness, with respect to sodium bromide and sodium hypochlorite at 0.038 mmol/g of aTEMPO