Introduction

Nanocelluloses are promising bio-based nanomaterials. They can be used as both nanofillers to reinforce polymer matrices and versatile bulk materials such as transparent films, hydrogels, and aerogels (Habibi et al. 2010; Klemm et al. 2011; Moon et al. 2011; Isogai et al. 2011; Isogai 2013). Natural and synthetic rubbers have unique properties and are used as components of a wide range of products, including tires, belts, hose pipes, packings, sealants, O-rings, rollers, and cushions. The mechanical properties of rubbers are improved by vulcanization and addition of fillers such as carbon black and silica (Karasek and Sumita 1996; Hshim et al. 1998; Wang 1998). The reinforcement of rubbers with various nanofillers such as nanoclays (Varghese and Karger-Kocsis 2003; Gatos et al. 2005; Kader et al. 2006) and carbon nanotubes (Endo et al. 2008; Deng et al. 2011) has been investigated in recent years. Nanocelluloses have also been used as fillers for rubber matrix reinforcement (Bendahou et al. 2010; Bras et al. 2010; Abraham et al. 2013; Annamalai et al. 2014; Parambath Kanoth et al. 2015), and nanocellulose/rubber composite materials with good mechanical and thermal properties have been reported.

When plant cellulose/water suspensions at pH 10 are subjected to catalytic oxidation using 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) and NaBr with NaClO as the primary oxidant, significant amounts of sodium carboxylate groups are formed densely, regularly, and position-selectively on the crystalline cellulose microfibrils in the plant cellulose fibers. The mechanical disintegration of fibrous TEMPO-oxidized plant celluloses with carboxylate contents > 1 mmol/g in water under moderate conditions gives highly viscous and transparent gels. The gels consist of completely individualized TEMPO-oxidized cellulose nanofibrils (TOCNs) with homogeneous widths of ~ 3 nm and high aspect ratios, i.e., > 150, the morphologies of which are different from those of other nanocelluloses (Saito et al. 2006; 2007; Isogai et al. 2011; Shinoda et al. 2012). Moreover, because there are significant amounts of sodium carboxylate groups on the surfaces of the crystalline TOCN elements, other metal ions and hydrophobic alkylammonium ions can be introduced onto the TOCN surfaces by simple ion exchange in water (Fujisawa et al. 2012a, b, 2014; Shimizu et al. 2014a, b, 2016).

TOCN films prepared from TOCNs with sodium carboxylate groups (TOCN-Na) by casting and drying have high Young’s moduli of 6–7 GPa, high tensile strengths of 200–300 MPa, low coefficients of thermal expansion (CTEs) of ~ 3 ppm/K, and low oxygen transmission rates of ~ 0.001 mL µm/(m2 day kPa) under dry conditions (Fukuzumi et al. 2009; Isogai et al. 2011). The thermal and mechanical properties of TOCN-reinforced rubber composite films are therefore expected to be better than those of neat rubber films or other nanocellulose-reinforced composite rubber films because TOCNs have higher aspect ratios and more homogeneous and smaller widths than other nanocelluloses. When counterion-exchanged TOCNs (TOCN-X) and surface-hydrophobized TOCNs were used as fillers to reinforce polystyrene, poly(lactic acid), and cellulose triacetate matrices, the mechanical and thermal properties of the TOCN/polymer composite films were significantly better than those of the neat polymer films (Fujisawa et al. 2012a, b, 2014; Soeta et al. 2015).

In a previous study, TOCN-X was mixed with hydrogenated acrylonitrile–butadiene rubber (H-NBR) latex under aqueous conditions, followed by casting and drying to prepare TOCN-X/H-NBR (TOCN/H-NBR = 0–5/100, w/w) composite films. H-NBR is a high-performance polymer, with better oil resistance, thermal resistance, and weather resistance than other natural and synthetic rubbers (Aimura 1997). The TOCN-X/H-NBR (TOCN/H-NBR = 5/100, w/w) composite films had high Young’s moduli, high storage elastic moduli at temperatures above the glass-transition point, and low CTEs under dry nitrogen conditions. However, microscopy images show that the hydrophilic TOCN elements form heterogeneous grid-like networks around hydrophobic H-NBR latex particles in the composite films. If TOCN/H-NBR composite films are prepared using surface-hydrophobized TOCNs in a homogeneous system using an H-NBR/organic solvent solution (rather than the aqueous H-NBR latex described above), TOCN/H-NBR composite films with different mechanical or thermal properties should be obtained.

In this study, surface-hydrophobized TOCN/H-NBR composite films were prepared in heterogeneous system with water and a homogeneous system with N,N-dimethylformamide (DMF). The mechanical and thermal properties of TOCN/rubber composite films prepared using the two systems were compared to clarify the effects of different preparation systems on the TOCN/H-NBR composite film properties.

Experimental

Materials

TEMPO-oxidized cellulose with a sodium carboxylate content of 1.1 mmol/g (TOC-Na) was prepared from wood cellulose according to a previously reported procedure (Fukui et al. 2018). A commercial H-NBR latex (solid content 40%, w/v, Zetpol 2230LX, Nippon Zeon Co., Ltd., Tokyo, Japan) was kindly provided by Nippon Zeon. The chemical structure of H-NBR is shown in Fig. 1. The aqueous H-NBR latex (2 mL) was freeze-dried, and the freeze-dried H-NBR was stirred in DMF (20 mL) to prepare a 4% (w/v) H-NBR/DMF solution. Aqueous N(n-Bu)4OH solution was purchased from Sigma-Aldrich (USA). Other chemicals and solvents (laboratory grade, Wako Pure Chemical Ind. Ltd., Osaka, Japan) were used as received.

Fig. 1
figure 1

Chemical structure of hydrogenated acrylonitrile–butadiene rubber (H-NBR)

Full size image

Preparation of TEMPO-oxidized cellulose nanofibrils

TOCNs with tetra-n-butylammonium counterions [TOCN-N(n-Bu)4] were prepared from fibrous TOC-Na. The TOC-Na was counterion-exchanged to TOC-H with 1 M HCl, and the TOC-H was counterion-exchanged to TOC-N(n-Bu)4 by neutralization with N(n-Bu)4OH (Shimizu et al. 2014a; Fukui et al. 2018). The TOC-N(n-Bu)4 was mechanically disintegrated in water using double-cylinder-type and ultrasonic homogenizers under the conditions reported in a previous paper (Fukui et al. 2018). The fibrous TOC-N(n-Bu)4/water suspension was solvent exchanged to give a TOC-N(n-Bu)4/DMF suspension by centrifugation several times with fresh DMF. The TOC-N(n-Bu)4/DMF suspension was mechanically disintegrated to give a TOCN-N(n-Bu)4/DMF dispersion. The TOCN-N(n-Bu)4/water and TOCN-N(n-Bu)4/DMF dispersions were both centrifuged at 12,000×g for 10 min to remove the small amounts of unfibrillated fractions. The supernatants were used as ~ 0.1% (w/v) TOCN-N(n-Bu)4/water and TOCN-N(n-Bu)4/DMF dispersions.

Preparation of TOCN-N(n-Bu)4/H-NMR composite films

Certain amounts of TOCN-N(n-Bu)4/water dispersion were added to aqueous H-NBR latex (2 mL). The mixtures were stirred for 30 min, cast in glass Petri dishes, and dried at 40 °C for 3 days to prepare TOCN-N(n-Bu)4/H-NBR composite films with TOCN/H-NBR weight ratios of 0–5/100 (without including the weights of the TOCN counterions) in the heterogeneous TOCN/H-NBR latex/water system (Fig. 2a). Certain amounts of TOCN-N(n-Bu)4/DMF dispersion were added to the H-NBR/DMF solution, and the mixtures were stirred for 30 min. The mixtures were then cast in glass Petri dishes, and dried in vacuo at 60 °C for 3 days to prepare TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 0–5/100, w/w) composite films using a homogeneous TOCN/H-NBR/DMF system (Fig. 2b). The neat and composite film thicknesses and densities after conditioning at 23 °C and 50% relative humidity were ~ 200 μm and 0.8–0.9 g/cm3, respectively.

Fig. 2
figure 2

Schematic diagrams of preparation of TOCN-N(n-Bu)4/H-NBR composite films in heterogeneous system with water (a) (Fukui et al. 2018) and homogeneous system with DMF (b)

Full size image

Analyses

Light-transmittance spectra of the composite films were recorded from 400 to 800 nm using a spectrophotometer (V-670, JASCO Corp., Tokyo, Japan) (Fukui et al. 2018). Tensile tests and thermomechanical analysis of the composite films were performed according to previously reported methods (Fukui et al. 2018).

Results and discussion

Preparation of TOCN-N(n-Bu)4/H-NBR composite films

All the TOCN-N(n-Bu)4/H-NBR composite films prepared in the heterogeneous system with water had high light transparencies, i.e., ~ 80% at 600 nm wavelength, similar to those reported for TOCN-Na/H-NBR composite films prepared in the same heterogeneous system with water (Fukui et al. 2018). In contrast, the composite films prepared in the homogeneous system with DMF had low transparencies (Fig. 3). This may have been the result of light scattering from the relatively rough surfaces of the films prepared using the homogeneous system; composite films prepared in the heterogeneous system had smooth surfaces (see photographs in Fig. 2). Another possibility is that the small amount of surfactant present in the original H-NBR latex to improve the stability of the latex particles in water, which is present in the freeze-dried H-NBR, is insoluble in DMF or immiscible with the TOCN-N(n-Bu)4 and H-NBR components of the films.

Fig. 3
figure 3

Light-transmittance spectra of TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 0–5/100, w/w) composite films prepared in homogeneous system with DMF (Fig. 2b). Inset shows photographs of corresponding composite films

Full size image

Mechanical properties of TOCN-N(n-Bu)4/H-NBR composite films

Typical stress–strain curves for the composite films prepared in the heterogeneous and homogeneous systems are shown in Fig. 4. The relationships between the TOCN content and the Young’s modulus, tensile strength, strain-to-failure, and work-of-fracture were calculated from the stress–strain curves in Fig. 4 and plotted in Fig. 5. The TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite film showed a clear yield phenomenon in the stress–strain curve.

Fig. 4
figure 4

Stress–strain curves for TOCN-N(n-Bu)4/H-NBR composite films (with TOCN/H-NBR = 0–5/100, w/w) prepared in heterogeneous system with water (a) and homogeneous system with DMF (b)

Full size image
Fig. 5
figure 5

Relationships between TOCN content and Young’s modulus, tensile strength, strain-to-failure, and work-of-fracture of composite films prepared in heterogeneous system with water and homogeneous system with DMF

Full size image

As shown in Fig. 5, the Young’s moduli of the composite films prepared in the two systems increased similarly with increasing TOCN content. There was almost no difference between the Young’s moduli of the composite films prepared using the two systems at the same TOCN content. The tensile strengths of the composite films prepared in the heterogeneous system gradually decreased with increasing TOCN content, whereas those of the composite films prepared in the homogeneous system increased. Conversely, the strain-to-failure values of the composite films prepared in the heterogeneous system increased with increasing TOCN content, whereas those of the composite films prepared in the homogeneous system decreased. The work-of-fracture values increased with increasing TOCN content for the composite films prepared in the two systems, although the values were different for films prepared in the two systems at the same TOCN content. The trends in the tensile strengths and strain-to-failure values of the composite films prepared in the two systems therefore clearly differed.

Thermal properties of TOCN-N(n-Bu)4/H-NBR composite films

The CTE values in temperature regions A and B in Fig. 6 were calculated by plotting the expansion ratios of the composite films against temperature under dry nitrogen conditions. The CTEs of the composite films prepared in the two systems decreased significantly with increasing TOCN content in the temperature regions A and B in Fig. 6. No significant differences were observed between the CTE values for the composite films prepared in the two systems in the lower temperature range A.

Fig. 6
figure 6

Typical thermal expansion curve for composite film (left), and relationships between TOCN contents and coefficient of thermal expansion of TOCN-N(n-Bu)4/H-NBR composite film in temperature regions A and B

Full size image

Difference between structures of composite films prepared in heterogeneous and homogeneous systems

As described in the previous section, there were clear differences between the tensile strengths and strain-to-failure values of the TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite films prepared in the heterogeneous and homogeneous systems. The Young’s moduli and CTE values of the composite films were almost the same at the same TOCN content. Figures 7 and 8 show typical stress–strain curves for the TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite films prepared in the heterogeneous system with water and homogeneous system with DMF, respectively, and the corresponding photographs of the films during tensile tests. The composite film prepared in the heterogeneous system showed a typical yield phenomenon and necking behavior. In contrast, the composite film prepared in the homogeneous system showed neither necking behavior nor yield during tensile tests. These results show that the primary differences between the two systems did not appear in the elastic region, unlike the Young’s modulus or CTE, but appeared in the plastic region after the elastic region during tensile tests.

Fig. 7
figure 7

Typical stress–strain curve for TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite film prepared in heterogeneous system with water, and photographs of film during tensile test, showing necking behavior at C and D

Full size image
Fig. 8
figure 8

Typical stress–strains curve for TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite film prepared in homogeneous system with DMF, and photographs of film during tensile test, showing no necking behavior

Full size image

Necking behavior of polymer and polymer composite films in the plastic region during tensile tests has been studied experimentally and theoretically for a long time (Bucknall and Smith 1965; Wu and van der Giessen 1995; Brünig 1998; Tervoort and Govaert 2000). Polymer/clay composite films, which consist of spherical polymer particles surrounded with clay nanoparticles, show characteristic necking behavior during tensile tests (Haraguchi et al. 2006). A transmission electron microscopy image of the TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite film showed that similar heterogeneous and two-phase structures may have been partly retained in the film (Fukui et al. 2018). When cellulose nanocrystals are homogeneously distributed in a rubber matrix, no necking behavior is observed during tensile tests (Nair and Dufresne 2003; De and White 1996). The results shown in Figs. 7 and 8 therefore suggest that TOCN-N(n-Bu)4 elements are homogeneously distributed in a homogeneous H-NBR polymer matrix in the composite film prepared in the homogeneous system with DMF.

Figure 9 shows typical stress–strain curves in first and second tensile tests; the second tensile test was performed before the point at which fracture occurred in the first test for the same sample specimen. Neither a yield phenomenon nor necking behavior was observed in the second tensile test for the composite film prepared in the heterogeneous system. The originally heterogeneous and two-phase structures of the TOCN-N(n-Bu)4/H-NBR composite film prepared in the heterogeneous system therefore became more homogeneous, similar to those of the composite film prepared in the homogeneous system. However, the tensile strength and strain-to-failure value of the composite film prepared in the homogeneous system, determined in the second tensile test (Fig. 9b), were still much higher and lower, respectively, than those for the composite film shown in Fig. 9a.

Fig. 9
figure 9

Stress–strain curves for TOCN-N(n-Bu)4/H-NBR (TOCN/H-NBR = 5/100, w/w) composite films prepared in heterogeneous system with water (a) and homogeneous system with DMF (b), obtained from first and second tensile tests

Full size image

Schematic structural models of the TOCN-N(n-Bu)4/H-NBR composite films prepared in the two systems are shown in Fig. 10. The procedure for preparation of the composite films in the heterogeneous system with water, without any organic solvent, is much simpler and preferable to that for preparation in the homogeneous system with DMF, both environmentally and industrially. However, the homogeneous system is better for the preparation of TOCN/H-NBR composite films with high tensile strengths and more homogeneous structures. Thermal molding or kneading of aqueous dispersions of TOCNs and aqueous H-NBR latex is a simpler and more practical method for removing water from mixtures and preparing dry TOCN/H-NBR composite materials. However, in this case, hydrophilic TOCN elements aggregate preferentially in the hydrophobic H-NBR matrix during water removal, and the TOCNs have almost no nanofiller effect on the composite materials. It is therefore necessary to establish suitable procedures for preparing TOCN/H-NBR composite films without using any organic solvent, to enable homogeneous distribution of TOCN elements in the homogeneous H-NBR polymer matrix, resulting in the TOCNs acting as efficient nanofillers in the composite films.

Fig. 10
figure 10

Schematic structural models of TOCN-N(n-Bu)4/H-NBR composite films prepared in heterogeneous system with water (a) (Fukui et al. 2018) and homogeneous system with DMF (b)

Full size image

Conclusions

Composite films consisting of surface-hydrophobized TOCN-N(n-Bu)4 and H-NBR (TOCN/H-NBR = 0–5, w/w) were prepared in a heterogeneous system with water and a homogeneous system with DMF, by mixing, casting, and drying. The two composite films both had high Young’s moduli of ~ 45 MPa and low CTEs of ~ 20 ppm/K at a weight ratio of TOCN/H-NBR = 5/100, irrespective of the system used to prepare the composite films. In contrast, the tensile strengths and strain-to-failure values of composite films prepared using the two systems clearly differed. These common and different properties of the composite films prepared in the two systems are probably the results of different distributions/structures of the two components. The composite films prepared in the heterogeneous system with water partly retain the original latex particle structure and the TOCN elements are present in the films, surrounding the latex particles. In contrast, the composite films prepared in the homogeneous system with DMF have more homogeneous distributions of TOCN elements in the homogeneous H-NBR polymer matrix. Although the TOCN surfaces were hydrophobized with tetra-n-butylammonium groups, heterogeneous distribution of TOCN elements without penetration into the hydrophobic H-NBR molecules may be partly retained in the composite films.