Morphological properties of cellulosic fibers
The morphological properties of cellulosic fibers obtained after grinding with different times were characterized using SEM and AFM as shown in Fig. 1 and Fig. 2. After ten times of grinding treatment (sample A10), the cylindrical and bunchy original cellulose fibrils were squashed under the shear force when the pulp slurry passed the gap between the static and rotating grind stone. A few individual nanofibrillated cellulose were found at the ends and edges of the squashed cellulose. But the vast majority of fibrils were still tightly aligned along the cellulose and closely connected to each other with the ‘band’ feature. It suggested that ten times of grinding treatment can’t completely break down the hydrogen bonds and cell wall structures. The size of sample A10 was basically kept at the micrometer in diameter, as shown in Table 1. After 20 times of grinding treatments, the size of cellulosic fiber was obviously decreased due to the partial breakage of hydrogen bonds and cell wall structures. As shown in Fig. 1d-f, the internal fibrillation of cellulose increased with increasing the times of grinding treatment, thus presenting a ‘tree’ feature. A mass of nanofibrillated individual cellulose appeared at the ends of squashed cellulose like the branches of trees (Fig. 2f), ranging from the large nano-size of several hundred nanometers to the small nano-size of about a few nanometers. It suggested that the amount of fibrillation was obviously improved with increasing mechanical fibrillation. But the center section of squashed cellulose with approximately several micrometers was a non-individual body like the ‘tree trunk’. Similar morphological properties have been reported in other studies (Phanthong et al. 2018; Wang et al. 2016). After 40 times of grinding treatments, most part of cellulose was completely nanofibrillated producing the long and flexible nanofibrillated cellulose with tens of nanometer in diameter. No obvious center ‘tree trunk’ sections were retained, implying the hydrogen bonds and cell wall structures were entirely broken down. As a result, the amount of fibrillation was increased and size in diameter was decreased with ongoing grinding treatment. The aspect ratio of cellulosic fiber was obviously increased with increasing mechanical fibrillation, resulting in the flexible and elongated nature of cellulosic fiber. Moreover, the completely nanofibrillated cellulose with a strong capacity of entanglement was aggregated and twisted, causing a highly entangled fiber network structure. It played an important role in the rheological behavior and stability of cellulosic fiber suspensions.
Rheological properties of cellulose suspension
Steady state viscosity versus shear rate in controlled shear mode of cellulosic fiber suspensions with the different times of grinding treatment are shown in Fig. 3. All curves exhibited the typical shear-thinning behavior. This was because the original fiber network was destroyed, leading to the fiber orientation movement along flow lines upon the shear force, as shown in Fig. 4. Strong evidence of increasing viscosity of suspension was found when the concentration of cellulosic fiber was improved. There was an entangled fiber network structure in the serried fiber suspension. Importantly, it was observed that the mechanical fibrillation leaded to obvious increment of viscosity at the same concentration. For example, at a shear rate of 0.1 s− 1, the viscosities of suspensions for A10, A20, and A40 were 5.73 Pa.s, 23.33 Pa.s, and 54.21 Pa.s, respectively. As previous study reported, the viscosity of cellulose suspension was increased as the homogenization treatment (Taheri and Samyn 2016). It is generally accepted that the viscosity of cellulose suspensions obtained by mechanical treatments mainly depends on the strength of the fiber network structure. The large gap of viscosity appeared between the sample A10 and A40, indicating the formation of stronger fiber network structure of completely nanofibrillated cellulose. It was attributed to the strong capacity of entanglement of the long and flexible fiber obtained by mechanical fibrillation.
The oscillatory strain sweep and frequency sweep were measured in order to clarify the viscoelastic behavior and fiber network structure of cellulosic fiber suspensions. Oscillatory strain tests were usually used for studying the linear viscoelastic region. As shown in Fig. 5a-c, linear viscoelastic response to strain-independent of all suspensions were found at low strain region. At this region, the storage modulus (G’) was much greater than the loss modulus (G’’). Once strain exceeded the critical strain, G’ decreased rapidly and intersected with G’’ due to the destroyed fiber network structure, transforming into nonlinear behavior (Nechyporchuk et al. 2016). The oscillatory frequency tests were carried out at the linear viscoelastic region. In order to evaluate the viscoelastic behavior, the value of G’ and G’’ were compared according to the previous report of Li et al. (Li et al. 2015). At low angular frequency, G’’>> G’ and tan δ > 1, both G’ and G’’ obviously improved with the increasing angular frequency, suggesting the viscous rheological behavior. On the contrary, G’ > G’’ and tan δ < 1, the storage modulus was independent of frequency, displaying the gel-like behavior. The viscoelastic modulus of cellulosic fiber suspensions as functions of frequency (ω) and the loss tangent value are shown in Fig. 2d-i. The concentration of cellulosic fiber had a significant effect on the viscoelastic behavior. The cellulosic fiber suspensions presented the typical viscous rheological behavior at low cellulose concentration (0.25% and 0.50%) due to the weak fiber network. The loss tangent value was distinctly decreased with the increased concentration of cellulose, indicating that the phase of suspension was transformed to a strong elastic behavior. For A10 suspensions, the weak elastic behavior was found only in the cellulosic fiber suspension with a concentration of 1.0%, in which G’ and G’’ almost overlapped at low frequency regions, while G’ was larger than G’’ at high frequency regions. Therefore, the phase transition occurred between the cellulose concentration of 0.75–1.0%. In the same way, as the cellulose concentration increased, the phase transition of A20 suspension and A40 suspension happened between the concentration of 0.50–0.75%. But the loss tangent value of A40 suspension was smaller than A20 suspension at the same concentration, implying the stronger gel-like behavior. Consequently, as the same fiber concentration, the viscoelastic behavior of cellulosic fiber suspensions was dominated by the morphological properties of cellulose. As the mechanical fibrillation increased, the cellulose was completely nanofibrillated, resulting in the individually long and flexible fiber had the strong capacity of entanglement. The completely nanofibrillated cellulose was aggregated and twisted to form an entangled fiber network structure.
The measurements of response time from the retarded strain recovery after creep were measured to investigated the stress-dependent rheological properties. The creep compliance as function of time are shown in Fig. 5a-c. The creep compliance value of cellulosic fiber suspensions was obviously reduced with the increasing of fiber concentration. For example, the maximum creep compliance value of A40 suspensions with the fiber concentration of 0.25%, 0.50%, 0.75%, and 1.0% were 21.27 Pa− 1, 2.40 Pa− 1, 0.32 Pa− 1, and 0.04 Pa− 1, respectively. The creep compliance result is the ratio of strain to stress, thus suggesting the deformation in per unit stress (Yuan et al. 2021). More fibers participated in the formation of rigid fiber network that can sustain its stability under the small deformation. Moreover, the mechanical fibrillation had significant effect on the creep compliance and the strain recovery. It was obvious that more times of grinding treatments resulted in lower value of creep compliance at the same concentration. The step wise augment of mechanical fibrillation of A10, A20, and A40 leaded to the maximum creep compliance of suspension of 0.72 Pa− 1, 0.14 Pa− 1, and 0.04 Pa− 1, respectively. The small creep compliance value stands for high capacity of anti-deformation. Thus, the extremely entangled fiber network can maintain the structure and limit fiber orientation movement under the low shear rate as shown in Fig. 4. The strong fiber network structure and stable cellulosic fiber suspension can be obtained by more mechanical fibrillation.
Apparent yield stress is considered as the significant rheological parameter of cellulosic fiber suspensions in the process of production and utilization. It is the minimum stress that drives the cellulose suspension to flow. As shown in Fig. 5d, the apparent yield stress was dramatically increased with the increasing of fiber concentration, suggesting that the suspension with high fiber concentration needs more external stress to break the fiber network structure for driving flow. The morphological properties of cellulose play an important role in the apparent yield stress of cellulosic fiber suspension. At fiber concentration of 1.0%, the yield stress of suspensions for A10, A20, and A40 were 0.35 Pa, 4.3 Pa, and 6.1 Pa, respectively. The apparent yield stress was obviously increased after more mechanical fibrillation. As mentioned above, the long and flexible nanofibrillated cellulose formed the strong fiber network structure which could resist the excess stress.
Figure 5e gives the alteration of dynamic modulus during dynamic time sweeps for dynamic stability of cellulosic fiber suspensions. For A20 and A40 suspensions, when the strain level was at 500%, the original fiber network structure was disrupted immediately, resulting in the viscous fluid behavior. However, at the second strain level of 0.1%, it was clear that the modulus was obviously reduced compared with the initial levels. And the subsequent dynamic modulus could maintain the same level during continuous dynamic time sweeps. It indicated that the cellulosic fiber suspensions obtained by mechanical fibrillation had outstanding dynamic stability and good repeatability, due to the outstanding self-assembly ability and strong capacity of entanglement of cellulose. However, there was a slight difference of the viscoelastic modulus in the process of destruction and reconstruction of fiber network between A10 and other samples. It suggested that the squashed cellulose could reform a new fiber network structure, but with a weak repetitive stability. This was due to the limited self-assembly ability and capacity of entanglement of the large and rigid squashed cellulose produced by less mechanical fibrillation. Consequently, more mechanical fibrillation could dramatically fibrillated cellulose resulting in the long and flexible nanofibrillated cellulose, which created an extremely entangled fiber network structure.
Figure 2f shows the effect of temperature on the shear viscosity of cellulosic fiber suspensions with different mechanical fibrillation. It was well known that high temperature could introduce energy to disrupt the entangled fiber network structure, improving the fibers to move freely and resulting in the decreased viscosity. As A20 and A40 possessed the highly entangled fiber network structure, their viscosities were marginally decreased as the temperature improved due to the decreased viscosity of water and the slight increment of fiber mobility under high temperature. But the sample of A10 exhibited an obvious decrease of viscosity as the temperature increased. This was because the squashed cellulose had high mobility without the limitation of entangled fibril network structure under high temperature. It indicated that mechanical fibrillation could improve the stability of viscosity of cellulosic fiber suspension under high temperature.