View Article Online
Materials Advances
Paper
by weakening the hydrogen bonds. 34 Later, SHP may react with the intermediate of the anhydride to esterify the cellulose. 35 Guo et al. 36 investigated the effects of varying weight ratios between CNF and BTCA and revealed that a weight ratio of 10 : 1 (CNF : BTCA) was optimal for the intended application, and the amount of SHP was the same of BTCA. Similar research was conducted by Dinesh et al. 37 revealed that a 10 : 1 ratio of CNF to CA showed the highest degree of crosslinking. Therefore, in this study, a 10 : 1 weight ratio of CNF to BTCA or CA was applied, and the weight of SHP was the same as the poly- carboxylic acid in each formulation. PAE is another common crosslinker in the papermaking industry. The primary mechanism of PAE crosslinking with cellulose is attributed to the ester bond formation between carboxyl groups on cellulose and azetidinium groups of PAE, as shown in Fig. S2 (ESI † ). 38 There is also a secondary mechanism of homo-crosslinking of PAE itself, 39–41 where the carboxyl end group on one polymer chain reacts with the secondary amine group on another chain and forms crosslinking network. 42–44 It has been discovered that in the interactions between PAE and cellulose, a dual crosslinking mechanism was evident. 19 The self-crosslinking involves azetidinium groups crosslinking with the carboxyl groups and primary amino groups, as well as the secondary amine on the backbone. This self-crosslinking phenomenon results in the formation of a water-insoluble network structure and further increases mechanical performance. However, the degree of biodegradability and the repulping ability of cured cellulose would be decreased if too much PAE had been used. Therefore, the utilization of PAE is typically minimal, with dosages generally constituting approximately 1% of the cellulose dry weight. 1.5% PAE was used based on the previous literatures and reports. 15,17,19 Despite this relatively low concentration, the crosslinking was observed to be highly effective. FTIR analysis was conducted on CNF samples crosslinked with BTCA, CA, and PAE respectively to verify the occurrence of crosslinking, and pure CNF and CNF/CMC mixture were also tested as control groups. 10CNF/1CMC/1BTCA/1SHP, 10CNF/ 1CMC/1CA/1SHP, and 100CNF/10CMC/1.5PAE were tested. 10CNF/1CMC and pure CNF were also tested as control groups. Fig. 1 shows the FTIR of the formulations. Peaks at 3338 cm 1 (O–H stretching), 2902 cm 1 (C–H stretching), and 1642 cm 1 (O–H bending) were observed among all five CNF-based sam- ples, which are characteristic peaks for CNFs. 42,43 Peaks at 1594 cm 1 were due to the asymmetric stretching vibration of the carboxylate (COO–) moiety, 27,44 and this peak was observed in all samples with CMC except the CA crosslinked, which could be attributed to the masking by the adjacent peaks at 1642 cm 1 . Peaks at 1726 cm 1 for BTCA crosslinked and CA crosslinked samples and peak at 1742 cm 1 for PAE crosslinked samples were attributed to C Q O ester bond stretching vibration. 15,19,36–38,45 For BTCA crosslinked and CA crosslinked samples, the esterification reaction happened between the hydroxyl groups on cellulose and the carboxylic groups on the polycarboxylic acids (BTCA or CA). For PAE crosslinked sample, the esterification occurs between the carboxyl groups and azetidinium groups of PAE. It is normal for these ester groups
Rule Dies (Medford, NJ) following ASTM D6438 Type IV standard. Unless stated otherwise, the dimensions were measured by a digital caliper (Fisherbrand t Traceable t ) or a micrometer (Mitutoyo Corporation). The dog-bone sample has a total length of 60 mm, neck width of 4.80 mm, and gauge width of 12 mm. The thickness of the dry coating sample was around 200 m m, and thickness of the coated MP trays was around 1.7 mm. The specimens were first preconditioned at room temperature and 10% RH for a day. Sandpapers were used between the grips and specimens to avoid slippage. The crosshead speed was set at 1 mmmin 1 . The ultimate stress was taken from the highest stress before break, strain is calculated from the crosshead displacement, and the modulus is calculated using the steepest slope method. For formulated CNF-based coatings, wet strength of the coatings were also tested. The dog-bone specimens were first soaked inside the DI water for 4 hours followed by the uniaxial tensile test. The samples were lightly pressed with blotting paper to remove excess water before testing. The testing procedure is the same as the one for dry specimens.
3. Results and discussion 3.1. Turbidity and FTIR
Common crosslinking agents for cellulose nanofibrils (CNFs) include polycarbodiimide (CDI), 1,2,3,4-butanetetracarboxylic acid (BTCA), citric acid (CA), and polyamide epichlorohydrin (PAE). We first screened crosslinkers for a CNF/CMC formula- tion to determine which provided the best properties. Our previous work has shown that CMC exhibits a strong affinity for adsorption onto the surface of CNF and has a preferred weight ratio of CNF to CMC was 10 : 1. 27,33 The CMC acts to allow molding, extrusion and very high solids content by colloidally stabilizing the CNF. Our previous work on MP coatings used CNF/CMC and for all work here, this material was used. CDI requires acidic conditions to be activated for cross- linking. The crosslinking process for CNF cords can be effec- tively accomplished by immersion in an acidic solution, thereby meeting the requisite conditions. 27 However, when copious quantities of CNF paste balls are to be prepared, a prolonged immersion period in acidic solution post-preparation is necessitated. The diffusion of acid within CNF paste balls may not be homogeneous. Consequently, a concentration gradient is likely to be established between the surface and the interior of the paste balls. Alternatively, the introduction of acidic compo- nents during the mixing process may be considered. However, this approach is also potentially problematic as corrosive effects on the stainless-steel mixing vessel itself may be induced. Furthermore, in the context of CNF application as a coating material for molded pulp trays, it is deemed preferable to utilize the material under neutral conditions. BTCA and CA are polycarboxylic acids that are commonly used to crosslink cellulose via esterification reaction. Sodium hypophosphite (SHP) has proven to be an efficient catalyst for the reaction. In the presence of SHP, BTCA or CA may form cyclic anhydride at temperatures lower than the melting point
Mater. Adv. , 2025, 6 , 2833–2844 | 2837
© 2025 The Author(s). Published by the Royal Society of Chemistry
Made with FlippingBook flipbook maker