Benchmarking the Production of Cellulose Nanofibres: Biomass Feedstock, Mechanical Processing, and Nanopaper Performance

doi:10.21203/rs.3.rs-1852780/v2

Download PDF

Research Article

Benchmarking the Production of Cellulose Nanofibres: Biomass Feedstock, Mechanical Processing, and Nanopaper Performance

https://doi.org/10.21203/rs.3.rs-1852780/v2

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Lignocellulosic biomass plays a vital role in the global shift away from the utilisation of non-renewable petrochemical resources. An emerging class of biomass-derived material is nanocellulose, which are typically generated from the deconstruction of cellulose bundles within the cell wall of terrestrial and aquatic plants, either in the form of cellulose nanocrystals (CNCs) or cellulose nanofibres (CNFs). However, the utilisation of biomass has an inherent challenge associated with product variability, both in terms of the starting feedstock properties, the wide range of processing routes available to generate nanocellulose, and the fabrication of nanocellulose into a diverse range of different product formats. As a result, it is difficult to accurately characterise and benchmark the wide variety of nanocellulose materials described within the literature. To address this challenge, this study presents a threefold benchmarking assessment of CNF-based material, including: (1) CNFs generated from different biomass sources (sorghum, banana, sugarcane, spinifex, and softwood); (2) CNFs generated through different mechanical processing methods (Silverson mixing, twin-screw extrusion, bead milling, and high pressure homogenisation); and (3) Energy-standardised nanopaper mechanical performance. The biomass benchmarking study highlighted sorghum and banana stem as comparatively sustainable biomass feedstock, while the mechanical process benchmarking study highlighted twin-screw extrusion as a promising fibrillation method with low energy consumption. Lastly, the nanopaper benchmarking study aided in the visualisation of the nanopaper research landscape. Sample benchmarking in this manner provides greater insight into the mechanisms driving nanocellulose material performance and processing sustainability.

Cellulose nanofiber

benchmarking

nanofibrillation

nanopaper

sustainability

Nanocellulose versatility

It is widely recognised that the demand for sustainable materials has dramatically increased since the turn of the century, with lignocellulosic biomass playing a vital role in the global shift away from the utilisation of petrochemical resources. Lignocellulosic biomass can be transformed into a variety of products ranging from biofuels, biochemicals, cosmetics, paper and packaging materials, biomedical materials, and electronic device components [1]. A key contributor to the advancement of biomass-based products for these applications is the development of nanostructured cellulosic materials from lignocellulosic biomass, either in the form of cellulose nanocrystals (CNCs) or cellulose nanofibres (CNFs) [2]. Nanocellulose is a highly versatile material that is able to be produced from a wide variety of feedstock sources, typically through the deconstruction of terrestrial lignocellulosic biomass, or alternatively through bacterial biosynthesis [3]. In addition, nanocellulose can be generated through a range of different processing routes, involving a combination of enzymatic, thermal, chemical, and mechanical treatments, as have been extensively reviewed in the past [2, 4–6]. Finally, nanocellulose materials can be fabricated into a range of different product formats, including in an aqueous suspension or paste, dried into a networked film, spray dried into solid particles, compounded into a polymer matrix, stabilised into an aqueous hydrogel structure, freeze dried into an aerogel structure, or carbonised at high temperatures to generate biomass carbon [7]. Furthermore, nanocellulose materials can be characterised through an array of different methods, which investigate properties that include the fibre-light interaction, fibre-solvent interaction, fibre-fibre interaction, fibre morphology and specific surface area, or intrinsic physical and chemical properties of fibres [8, 9].

Therefore, the versatility of nanocellulose products arises from the diversity in biomass feedstock, processing methods and product formats, which in turn creates challenges associated with benchmarking the performance and sustainability of samples across different research studies or products available on the market [9, 10]. This benchmarking challenge is expected to be projected forward into the industrial context, in terms of comparative product evaluation and manufacturing quality control to minimise batch-to-batch variation [11]. Like other manufacturing industries, product quality control for nanocellulose production will be crucial for business viability [12]. Biomass-based materials elevate the challenge of quality control through three additional sources of variability: (1) Genetic factors of the biomass, either within-plant variability between the different sections, within-family variability between different sorghum varieties, within-kingdom variability across different plant types including wood, non-wood, and aquatic origin, or across-kingdom variability between plant, animal, or bacterial-derived cellulose; (2) Environmental factors that influence the development of the biomass source in a complex, interactive sense, both from the natural environmental conditions or from human intervention through farm management practices; and (3) Processing factors, which act as a pseudo-environmental condition encompassing all of the biomass handling and material processing factors.

Role of benchmarking

Benchmarking is an emerging priority in the nanocellulose research field, driven by the desire of academics and industrial researchers to access and compare the property profile of industrially produced materials, and to compare factors that influence product performance and sustainability [13]. Delepierre et al. (2021) highlighted that with the emergence of industrially produced nanocellulose materials: “there is a demand to [develop a] benchmarking study so that [nanocellulose] users have access to up-to-date information and can continue to select materials that best fit their research and application needs” [13]. Benchmarking can be applied to all of the sources of variability outlined above, with the following list outlining potential questions that may be addressed through benchmarking-style investigation:

For biomass benchmarking: What biomass sources generate higher performance and/or more sustainable nanocellulose material?
For processing benchmarking: What processing methods or conditions generate higher performance and/or more sustainable nanocellulose material?
For product benchmarking: Given a particular product type, what are the highest performing and/or most sustainable nanocellulose samples?
For technology benchmarking: How does nanocellulose technology compare to alternative technologies or materials for a given application or to solve a given problem?
For characterisation benchmarking: How accurately do different characterisation methods predict the ‘true’ property profile of the material? How well does one characterisation metric predict the value of another characterisation metric?

Despite the prominent role of benchmarking in the process of technology commercialisation, few studies in the nanocellulose research field have directly investigated these topics [11, 13–15]. To address this gap in the literature, this study benchmarks the performance of different biomass types (Biomass Benchmarking), the effect of different mechanical processing methods on the resultant CNF-based material (Mechanical Process Benchmarking), and the performance and energy consumption of nanopaper material across different literature studies (Nanopaper Performance Benchmarking), with a focus on highlighting sample properties that contribute to high performance of the material.

Biomass Benchmarking

While the production of biomass-derived nanocellulose is a comparatively sustainable technology when compared to the production of petrochemical-derived materials, there remains an opportunity to optimise nanocellulose sustainability in terms of feedstock selection and utilisation [16]. Biomass sources exhibit differences in their biochemical composition, structural morphology, and cell wall architecture, as well as logistical differences such as biomass density, growth rate, seasonal availability, production cost, and transport cost. These parameters lead to differences in the properties and economic viability of nanocellulose products from different biomass sources. There is large diversity in the biomass types used for nanocellulose production with over 100 unique sources identified for nanocellulose production, sourced predominantly from wood, bacterial and agro-industrial byproducts [1]. However, despite this high degree of biomass variability, the majority of commercial nanocellulose production is undertaken using wood-based biomass [17]. In addition, only a limited number of research studies have thoroughly investigated nanocellulose performance across different biomass sources, as outlined in Table 1. System-level investigation across a wide range of biomass sources is required to elucidate biomass-processing-property-quality relationships for nanocellulose materials. The outcomes from this style of investigation would shed light on the best and worst performing biomass sources, the features of the biomass material that are associated with a higher relative performance, the efficiency of biomass deconstruction through chemical and mechanical treatment, and the nanocellulose properties that are generated from the deconstruction process. Table 1 takes a step toward answering these questions through presenting a summary of existing research studies that compare CNF properties across multiple biomass sources, ranking the relative order of characterisation metric values for each biomass source included within the study.

Table 1

Review of studies comparing multiple biomass types for CNF production
Biomass Types	Characterisation Performed	Biomass Properties	CNF Properties	Relationships	Citation
Coffee residues Sugarcane straw Corn husks Orange bagasse	Biochemical composition Crystallinity NMR TGA	Cellulose = Corn > Sugarcane > Coffee > Orange Lignin = Sugarcane > Coffee > Corn > Orange L:C = Sugarcane > Coffee > Orange ~ Corn H:L = Orange > Corn > Coffee > Sugarcane	Yield = Corn > Coffee > Sugarcane > Orange CrI = Corn > Orange > Sugarcane > Coffee Diameter = Orange > Sugarcane > Corn	A lower L:C ratio was associated with a higher CNF crystallinity	[18]
Oil palm tree: -Trunk -Mesocarp -Empty fruit bunch (EFB) -Palm kernel shell	Biochemical composition FTIR AFM TGA Crystallinity Tensile properties	Cellulose = Mesocarp > EFB > Trunk > Kernel Lignin = Kernel > Mesocarp > EFB > Trunk L:C = Kernel > Mesocarp > Trunk > EFB H:L = Trunk > EFB > Kernel > Mesocarp	CrI = Trunk > EFB > Kernel > Mesocarp Tensile = Trunk > EFB > Kernel > Mesocarp Strain = Mesocarp > Trunk > EFB > Kernel Modulus = Trunk > Kernel > EFB > Mesocarp	A lower H:L ratio was associated with a CNF sheet with lower tensile strength, Young’s modulus and crystallinity, while lower L:C ratio was associated with a higher tensile strength and crystallinity	[19]
Coconut coir Softwood Cotton	Biochemical composition Crystallinity SEM, TEM Rheology	Cellulose = Cotton > Wood > Coir Lignin = Coir > > Wood > > Cotton L:C = Coir > Wood > Cotton H:L = Cotton > > Wood > Coir CrI = Cotton > Wood > Coir	Yield = Cotton > Softwood > Coir CrI = Coir > Wood > Cotton Average Diameter = Cotton > Softwood = Coir	The high cellulose content and low hemicellulose and lignin content of cotton yielded a high CrI biomass, but led to the largest diameter and lowest CrI CNFs	[20]
Wheat straw Soy hulls	Biochemical composition SEM, TEM FTIR XRD TGA	Cellulose = Soy > Wheat L:C = Wheat > Soy H:L = Wheat > > Soy CrI (raw) = Soy > Wheat CrI (treated) = Wheat > Soy	Average Diameter = Soy > Wheat Average Length = Wheat > Soy	The higher H:L ratio and lower raw biomass CrI was associated with higher aspect ratio CNFs	[21]
Sugarcane bagasse Rice straw	Biochemical composition Degree of polymerisation Optical Microscopy SEM, TEM, AFM EDX Tensile properties Opacity	Cellulose = Bagasse > Rice Pentosans = Bagasse > Rice Ash = Rice > > Bagasse Degree of Polymerisation = Bagasse ~ Rice	Tensile strength = Bagasse > Rice Young’s Modulus = Bagasse > Rice Average Diameter (TEM/AFM) = Bagasse ~ Rice Opacity = Rice > > Bagasse	The high ash content in rice straw reduced the tensile strength properties and transparency of nanopaper sheets compared to those of the bagasse pulp."	[22]
Coconut husk Banana peel Oil palm trunk	Biochemical composition SEM, TEM XRD FTIR TGA	Cellulose = Oil palm > Coconut > > Banana H:L = Banana > > Oil palm > Coconut L:C = Banana > Coconut > Oil Palm CrI = Banana > Oil palm > Coconut T_max (TGA) = Banana > Coconut > Oil palm	CrI = Oil palm > Coconut > Banana T_max (TGA) = Coconut > Banana > Oil palm Diameter = Oil palm > Coconut > Banana Length = Coconut > Oil Palm > > Banana	The high H:L ratio in banana peel was associated with smaller diameter fibres	[23]
Banana rachis Sisal Kapok Pineapple leaf Coconut coir	FTIR SEM, TEM, AFM Birefringence XRD IGC TGA	T_{decomposition} (TGA) = Coconut > Kapok > Pineapple > Sisal > Banana Weight loss (TGA) = Kapok > Sisal > Banana > Coconut > Pineapple Residual char (TGA) = Coconut > Banana > Kapok > Sisal > Pineapple	Yield = Pineapple > Sisal > Kapok > Banana > Coconut CrI = Pineapple > Sisal > Kapok > Coconut > Banana Width (TEM) = Pineapple > Banana > Kapok > Coconut > Sisal	Nanofibre crystallinity is associated with treatment yield and the inverse of residual char, which is indicative of a higher cellulose content.	[24]
Abaca Sisal Hemp Jute Flax	Biochemical composition Degree of polymerisation AFM DMTA XRD Tensile properties Transparency	Hemicellulose = Sisal > Abaca > Jute > Hemp > Flax	Yield = Sisal > Abaca > Jute > Hemp > Flax CrI = Hemp = Flax > > Jute > Abaca = Sisal µ = Abaca > > Hemp > Jute > Sisal > Flax DP = Abaca > > Hemp > Jute > Sisal > Flax Width (AFM) = Flax > Jute ~ Hemp > Abaca > Sisal Nanocomposite tensile strength (10%) = Flax > Abaca ~ Jute > Sisal	A higher hemicellulose content was associated with a smaller width nanofibres, while flax with the lowest hemicellulose content had the lowest intrinsic viscosity and degree of polymerisation. However, the high nanocomposite reinforcement efficiency of flax was unexpected.	[25]
Wheat straw Barley straw Corn straw Oat straw	Biochemical composition FTIR XRD TGA DLS ZP	Cellulose (raw) = Corn > Wheat > Oat > Barley Hemicellulose (raw) = Corn ~ Wheat > Oat ~ Barley Ash (raw) = Barley > Wheat > Oat > Corn L:C (raw) = Barley ~ Wheat ~ Oat ~ Corn H:L (raw) = Oat > Wheat ~ Barley ~ Corn Hemicellulose (treated) = Wheat > Corn > Barley > Oat L:C (treated) = Oat > Barley > Wheat > Corn H:L (treated) = Wheat > Corn > Barley > Oat SR (treated) = Wheat > Oat > Corn > Barley Yield (treated) = Barley = Oat > Wheat > Corn DP (treated) = Corn > Wheat > Oat > Barley Length (treated) = Barley > Oat > Wheat > Corn Diameter (treated) = Oat > Corn ~ Barley ~ Wheat	SSA = Wheat > > Oat > Barley > Corn Diameter = Barley > Corn > Oat > Wheat DP = Corn ~ Wheat > Oat > Barley Yield = Wheat > > Oat > Corn > Barley	Wheat biomass generated CNFs that were significantly better than all other biomass in terms of SSA, fibre diameter, DP, and nanofibrillation yield. This was associated with the highest Hemicellulose content in the biomass and pulp stages, 2nd and 1st highest H:L ratio in the biomass and pulp stages respectively, and the highest Schopper-Riegler degree in the pulp stage.	[26]

While this qualitative style of information does not allow statistical modelling of the data, it does provide an overview of the literature that allows the reader to investigate relationships without getting lost in the quantitative data. However, this summary table does not provide information around the relative processing sustainability between the different biomass sources investigated, which is an increasingly important facet of the nanocellulose research field. It also does not inform upon the relative magnitude of values between different biomass sources for each characterisation metric.

One of the most commonly investigated characterisation methods is the biochemical composition of the biomass feedstock, which has been associated with the amenability and recalcitrance of the nanocellulose production process. A higher hemicellulose content has been linked with nanocellulose processing amenability, as hemicellulose is a relatively amorphous, hydrophilic component that promotes fibre swelling and chemical pretreatment efficiency. On the other hand, lignin is a relatively rigid, hydrophobic component that has been linked with nanocellulose processing recalcitrance [27–29]. Previous studies have proposed that the ratio between hemicellulose to lignin (H:L) has a positive impact on nanofibrillation efficiency, while the lignin to cellulose (L:C) has a negative impact on nanofibrillation efficiency [30]. In addition, biomass crystallinity index (CrI) has been proposed to have an inverse relationship with nanofibrillation efficiency, such that biomass with a lower CrI will enhance penetration of chemical pretreatment agents into the fibre structure and promote fibrillation during mechanical processing [31]. However, despite a drive to enhance processing sustainability and the widespread understanding that different biomass feedstocks generate differential outcomes in nanofibre properties, there is a paucity of systematic investigation into the biomass properties that encourage sustainable nanocellulose production, and which biomass types exhibit these properties.

Mechanical Process Benchmarking

Similar to different biomass feedstock, it is well established that different processing routes for nanocellulose production can lead to different nanocellulose properties, industrial scalability, and sustainability outcomes in terms of energy and water consumption. Alternative mechanical treatments have been extensively outlined in the literature, including milling, microfluidisation, disk refining, ultrafine grinding, high-intensity ultra-sonication, extrusion, cryocrushing, blending, and steam explosion [5, 6, 32, 33]. Different mechanical treatments exert a unique mechanism of action and force profile on the lignocellulosic material, which impacts the subsequent CNF morphology. Consequently, different mechanical treatments operate with different energy demand, commonly defined as the unit of energy consumed per dry weight of material processed (kWh/t or kWh/kg). The selection of a mechanical fibrillation method that demonstrates energy efficient processing is a key component of sustainability for the production of fibrillated cellulose materials. While the concept of sustainability encompasses a broader scope than merely processing sustainability, this is the central aspect of this benchmarking study. Li et al. (2021) highlighted this concept when stating that: “The sustainability and energy requirements for the production of fibrillated cellulose are tied not only to the biomass source but also to the processing methods employed. A resource such as fibrillated cellulose is only truly sustainable when its processing is also sustainable” [16]. A comparison between the alternative mechanical treatment processes used for the production of CNF suspensions is outlined in Table 2.

Table 2

Comparison of mechanical processing methods for CNF production
Mechanical Processing Method	Mechanism of Action	Solids Content	Citations
High Pressure Homogenisation	A dilute pulp slurry is pumped through a spring loaded valve assembly under high pressure (50–2000 bar), such that the fibre-water suspension undergoes intense micronisation caused by a rapid drop in pressure, generation of cavitation and shockwaves forces, turbulent flow, and high mechanical shearing as the material passed through a small gap between the homogenising valve and the impact ring. The generated fibres typically have a high specific surface area and a high aspect ratio.	0.5–1 wt. %	[33, 34]
Ball Milling	A dilute pulp slurry is pumped into a hollow cylindrical container, partially filled with ceramic, zirconia or metal balls of 0.4–1 mm diameter. When the container rotates, the heavy, millimetric-sized balls impart a high degree of impaction force through collisions between the balls, fibres, and container wall. Ball milled nanofibres are typically more branched with a lower aspect ratio.	0.5–2 wt. %	[4]
Microfluidization	A dilute pulp slurry is pumped through a N-shaped, Y-shaped, or Z-shaped channel at a speed of several Mach and a high pressure of 200–300 MPa, such that the material is subjected to large shear forces when passed through the tortuous flow channel. Unlike homogenisation, which operates at a constant processing volume, microfluidization operates at constant shear rate, which reduces the likelihood of clogs.	0.5–2 wt. %	[34, 35]
Disk Refining	A dilute pulp slurry is passed between two disks fitted with grooved surfaces, one that is stationary while the other rotates. Fibres are sheared to smaller dimensions through repeated cyclic stress as the disk rotates, with the gap between the grinding disks continuing to shrink during processing to promote delamination and fibrillation.	0.5–10 wt. %	[35]
Ultrafine Grinding / Microgrinding	A dilute pulp slurry is passed between a static and rotating grinding disk, generating shearing force that delaminates and fibrillates fibres. The grinding process is very similar to disk refining, whereas the main difference is the possibility of a lower gap between the discs for these grinders.	0.5–10 wt. %	[4]
Twin Screw Extrusion	A pulp paste is fibrillated by two intermeshing, co-rotating screws mounted in a closed barrel. The material is subjected to a thermo-physico-chemical method involving heat, compression, and shear force that promote the physical disruption of fibres as they pass through the extruder.	15–40 wt. %	[4]
Blending / High Shear Mixing	A dilute pulp slurry undergoes high speed rotor-stator mixing to impact a high degree of shear force on the material. High shear mixed fibres typically have a lower aspect ratio than homogenised fibres due to the cutting of fibres, and lower degree of fibrillation.	0.2–3 wt. %	[36]
Ultrasonication	High-energy acoustic waves sent through the cellulose fibre suspension promote growth and rupture of flow field bubbles (i.e. cavitation) that generate strong hydrodynamic shear forces during bubble implosion, leading to further fibrillation and dispersion of fibres.	0.5–4 wt. %	[35]
Steam Explosion	A dilute pulp slurry is exposed to pressurized steam, followed by a rapid release of pressure as the material is vented into the collection vessel, which promotes hydrolytic rupturing of the fibre structure and partial hydrolysis of non-cellulosic components.	10–30 wt. %	[4]
Cryocrushing	A pulp slurry is frozen with liquid nitrogen and crushed through the application of shear force, such that the mechanical force causes ice crystals to exert pressure on the fibre structure, leading to partial fibrillation.	~ 2 wt. %	[4]
Aqueous counter collision	Two streams of pulp slurry collide against each other under high pressure, resulting in pulverization and fibrillation of fibres.	< 0.4 wt. %	[4]

Nanopaper Performance Benchmarking

Cellulose nanopaper is a strong and tough fibre network material composed primarily of cellulose nanofibres, exhibiting a dual elastic and inelastic stress-strain curve when subjected to tensile testing [37]. Although pure nanopaper consists of 100% CNFs, there is still diversity in their mechanical properties due to the variability in biomass source and processing conditions, as outlined above. Benítez & Walther (2017) highlight this challenge in their unifying review of nanopaper mechanical performance, stating that: “The material space in the CNF world is unfortunately very wide, which makes it profoundly difficult to understand property space and derive general conclusions […] Important points to consider are first the origin and the processing of CNFs, which leads to variation in the degree of polymerization (DP), crystallinity (X_c), diameter, aspect ratio and hemicellulose/ lignin content” [38]. Moreover, the nanopaper fabrication route can impact the resulting mechanical properties, which is typically achieved through film casting or vacuum filtration, and an optional hot-pressing step [38]. To establish a comparative evaluation of CNF material generated within the academic literature, this study benchmarks the performance of CNFs in the simple product format of a vacuum filtered nanopaper handsheet. Nanopaper performance is assessed through mechanical tensile testing, while standardising the performance by the energy consumption required to process the CNF sample.

Materials

Sodium hydroxide (NaOH, pellets, 98% purity) was purchased from Chem Supply, Australia. Reverse Osmosis (RO) purified water was used throughout experimentation. Sorghum (Sorghum bicolor L. Moench) biomass of the Sugargraze variety was grown from seeds provided by the Department of Agriculture and Fisheries (DAF), Queensland, Australia. Softwood (Pinus radiata) biomass was provided by Hume Doors, Queensland, Australia, in the form of coarse wood chips. Sugarcane (Saccharum officinarum) bagasse was provided fresh from the Bundaberg Sugar Bingera sugar mill (Bundaberg, Queensland, Australia) after cane crushing. Sugarcane mulch (Oreco Sweet Garden 26m² Organic Sugar Cane Mulch) was purchased from Bunnings Warehouse, Australia, primarily consisting of sugarcane trash (tops and leaves). Banana (Musa acuminata ‘Cavendish’) biomass was provided from a private banana farm in Kiamba, Queensland, Australia. Spinifex (Triodia pungens) biomass was collected from native arid grasslands surrounding Camooweal, Queensland, Australia, and provided by the Dugalunji Aboriginal Corporation (DAC).

Biomass Preparation

Biomass was washed in distilled water at approximately 80°C and subsequently dried in a convection oven at 60°C until bone dry. Dried biomass was ground using a Retsch SM300 mill at 3000 rpm with a 1 mm trapezoidal mesh screen. Oversized material was separated from the ground biomass with a 0.71 mm aperture sieve. Ground biomass was dispersed and stirred overnight in deionised water at a solid ratio of 1:20 (20 g of water for every 1 g of ground biomass). Chemical pretreatment was performed using a 2% NaOH solution (w/v) at 80°C for 2 h. NaOH treated (delignified) grass was separated from the waste liquor through a fine mesh sieve (53 µm aperture) and rinsed extensively until the filtrate pH was below 8. The delignified grass suspension was set at the desired solids content for mechanical treatment using a Mettler Toledo moisture analyser.

Biochemical Composition

Near infrared spectroscopy (NIRS) analysis of biochemical composition was performed by Celignis Biomass Analysis Laboratories, Limerick (Ireland) on ground biomass samples with visible and near infrared radiation (400–2500 nm) using the FOSS XDS monochromator equipped with a Rapid Content Analyser (RCA) and Vision 3.5 software (Hayes, 2012).

High Pressure Homogenisation

High pressure homogenisation (HPH) was conducted with a Niro-Soavi Panda Plus 2000 machine (GEA, Italy). Delignified biomass passed through the homogeniser at a solids content of 0.5% (w/v) were collected at three sequential energy levels: Low energy – 1 \(\times\) 400 bar pass (L), Medium energy – 1 \(\times\) 400 bar pass + 1 \(\times\) 700 bar pass (M), High energy – 1 \(\times\) 400 bar pass + 1 \(\times\) 700 bar pass + 2 \(\times\) 1100 bar passes (H).

Silverson Blending

Blending was conducted with a L5M-A Laboratory Mixer (Silverson, UK). Delignified biomass samples at 1% (w/v) solid content were passed at 5000 rpm for 5 min with a small square mesh for the low energy level (L), 8000 rpm for 5 min at 1% (w/v) solids content with a small square mesh for the medium energy level (M), and 8000 rpm for 5 min at 1% (w/v) solid content with a small circular mesh for the high energy level (H).

Bead Milling

Milling was conducted with a LabStar Laboratory Bead Mill (Netzsch, Germany). Delignified biomass samples at 0.7% (w/v) solids content were passed through the 400 mL bead mill containing 1 mm zirconium oxide beads (ZetaBeads® Plus, Netzch, Germany) for 1 pass through a large mesh (0.4 mm) for the low energy level (L), and an additional pass for 20 min/L of suspension at 0.5% (w/v) solid content through a small mesh (0.2 mm) for the high energy level (H). The sample pump operated at 60 rpm and the mill speed was nominally set at 1500 rpm.

Twin Screw Extrusion

Extrusion was conducted with a HAAKE PolyLab OS Extruder (Rheomex PTW 16/40 ThermoScientific, England). Delignified biomass samples at ~ 15% (w/v) solids content were passed through the co-rotating intermeshing twin-screw extruder with a screw diameter of 16 mm and a barrel length to diameter ratio (L/D) of 40:1 for one pass for the low energy level (L), two passes for the medium energy level (M), and three passes for the high energy level (H). The extruder was driven by a motor of 7 kW fitted to a gear box with a 1:5.4 gear ratio. The screw profile used was based on the one employed by Rol et al. (2017) but modified with additional kneading blocks to increase mixing intensity.

Energy Consumption

The total energy consumption for the homogenisation, milling and extrusion processes were obtained from a FLUKE 1736 power logger. The energy consumption (kW) of each energy level was recorded and transformed into total energy (kWh) based on the time to pass. The energy consumption of the Silverson process was obtained from the voltage output of the machine. The final energy consumption values were presented on a specific basis as kWh per kg of dry material required to make a single nanopaper at a grammage of 90 gsm.

MorFi Fibre Analysis

10 g of CNF suspension at 0.5% (w/v) was added to approximately 1 L of water for fibre analysis using the MorFi Compact analyser (Techpap, France) equipped with CCD video camera, a high magnification optical flow cell and MorFi R.10.07 automatic analysis software. Four replicate analyses were run for each sample.

Thermogravimetric Analysis (TGA)

Thermogravimetric analysis (TGA) of biomass powder was conducted with a TGA/DSC1 thermogravimetric analyser (Mettler Toledo, United States) equipped with STAR^e evaluation software. A sample of approximately 5 mg was placed in a 40 µL aluminium crucible and weighed with an automatic precision microbalance (Mettler Toledo, United States). The crucible was sealed and subsequently perforated at the top with a pin. Each sample was heated from 25 to 500°C with a heating rate of 5°C/min under an inert nitrogen atmosphere. Raw data were analysed using the Mettler STAR^e Evaluation Software v.9.10 to obtain the initial step change (%), decomposition onset temperature (T_o, °C), the temperature at each of the derivative thermogravimetric (DTG) peaks (\({{\text{T}}_{\text{P}}}_{1-3}\), °C), and the residue mass remaining (%).

X-ray Diffraction (XRD)

The X-ray diffraction (XRD) analysis of the biomass powder was conducted on a Bruker D8 Advance X-ray diffractometer (Bruker, Germany). Graphite-filtered Cu-Kα radiation was generated at 30 kV voltage and 20 mA current. Samples were scanned over a range 2θ = 5°–80° with 0.02° steps at 1.2 s per step. The degree of crystallinity defined as the Crystallinity Index (CrI) was estimated from Eq. 1 proposed by Segal et al. (1959).

\(CrI=\frac{{I}_{200}-{I}_{Am}}{{I}_{200}} \times 100\) Eq. 1

where I₂₀₀ is the intensity of the peak at 2θ = 22.8°, which corresponds to the 200 lattice of the crystalline part of native cellulose Iβ, and I_Am is the intensity of the minimum region (the valley) at 2θ ≈ 18°, which corresponds to the non-crystalline component of cellulose (French 2014; Segal et al. 1959).

Nanopaper Mechanical Properties

CNF suspension was prepared into a nanopaper handsheet using an automatic Rapid Köthen handsheet former (Xell, Austria) according to the ISO 5269-2:2004 standard operating procedure [39]. Handsheet grammage was aimed at approximately 90 gsm (g/m²). The wet cake was dried under vacuum at 130°C for 15 min, or until the handsheet was completely dried. The fabricated nanopaper handsheets were conditioned in the laboratory (25°C, 50–75% relative humidity) for at least 1 day prior to mechanical testing. Up to eight rectangular nanopaper strips (L = 150 mm, W = 15 mm) were cut out from each handsheet. Prior to testing, the thickness of each strip was determined using an automated micrometer, taking the average of 10 measurements taken along the length of each strip. The density of each nanopaper strip was determined with a milligram scale balance (Mettler Toledo, USA) and previous dimension measurements, as shown in Eq. 2.

\({\rho }_{nanopaper}\left(\frac{g}{c{m}^{3}}\right)=\frac{w}{L \times W \times T}\) Eq. 2

Similarly, the grammage calculation used previously determined dimensions as shown in Eq. 3.

\(Grammage\left(\frac{g}{{m}^{2}}\right)=\frac{w}{L \times W}\) Eq. 3

Tensile properties of nanopaper samples were measured using an Instron model 5543 universal testing machine (Instron Pty Ltd., Melbourne, Australia), fitted with a 500 N load cell. Nanopaper strips were tested at a crosshead speed of 1 mm/min with a gauge length of 100 mm. The tensile index\(\left({{\sigma }}_{\text{w}}^{\text{T}}\right)\), otherwise known as the specific tensile strength per unit weight, was calculated with Eq. 4 using the ultimate tensile strength and nanopaper density. In this study, nanopaper tensile index is standardised by the energy consumption.

\({\sigma }_{w}^{T}\left(\frac{Nm}{g}\right)=\frac{UTS}{{\rho }_{nanopaper}}\) Eq. 4

Toughness was calculated as a numerical approximation of the energy absorbed by the nanopaper strip, from the zero-strain starting point until the nanopaper failure strain\(\left({\text{ϵ}}_{\text{f}}\right)\), according to Eq. 5.

\({U}_{T}\left(\frac{MJ}{{m}^{3}}\right)={\int }_{0}^{{ϵ}_{f}}f\left(x\right)dx\approx \sum _{i=2}^{n}\frac{1}{2}\left[{x}_{n}-{x}_{n-1}\right]\left[f\left({x}_{n}\right)+f\left({x}_{n-1}\right)\right]\) Eq. 5

Bibliometric Benchmarking of Nanopaper

Scope

The scoping criteria for the literature benchmarking study of CNF nanopaper, considering the energy consumption, pretreatment intensity, and biomass source used to generate the nanopaper material, is presented in Table 3.

Table 3

Scoping criteria for the nanopaper performance benchmarking review
In scope (all criteria required)	Out of scope
Studies generating films/nanopaper from CNF samples	A full life cycle assessment for CNF materials
Studies reporting the mechanical properties of CNF films/nanopaper	Consideration of the end use application and associated environmental impacts
Studies reporting the energy consumption of CNF processing	Energy/chemical consumption for raw material processing prior to receipt by researchers
Studies reporting the biomass source used for CNF processing	Scaling factor to adjust for lab-scale processing

Assumptions

A series of assumptions were used to process the data from the quantitative literature review, as outlined below:

Assumption 1: Specific energy consumption is reported in the units of kWh per kg of microfibrillated cellulose (MFC) or CNF material produced on a dry basis.
Assumption 2: Total energy consumption is limited to the mechanical treatment and pretreatment processes (not including energy consumption of biomass procurement, temperature requirements, stirring requirements, or product fabrication).
Assumption 3: For studies involving chemical pretreatment, reaction yield is used to convert the reagent input reported per kg of raw material to reagent input per kg of CNF material.
Assumption 4: The material yield of mechanical processing is 100%.
Assumption 5: Ultimate tensile strength is equivalent to tensile index when the nanopaper density was not reported.
Assumption 6: The threshold for acceptable material performance is a tensile index of 50 Nm/g.
Assumption 7: For energy consumption values <1 kWh/kg, no scaling of the TI value is performed for the TI/energy ratio.

Publications

Based on a thorough exploration of the relevant literature, the publications that fulfill all the scoping criteria from Table 3 are outlined in Table 4.

Table 4

Focal publications for nanopaper benchmarking study
Authors	Year	Title	Citation
Spence et al.	2011	A comparative study of energy consumption and physical properties of microfibrillated cellulose produced by different processing methods	[40]
Josset et al.	2018	Energy consumption of the nanofibrillation of bleached pulp, wheat straw and recycled newspaper through a grinding process	[41]
Kargupta et al.	2021	Sustainable production process of mechanically prepared nanocellulose from hardwood and softwood: A comparative investigation of refining energy consumption at laboratory and pilot scale	[42]
Ang et al.	2019	Effect of refining and homogenization on nanocellulose fiber development, sheet strength and energy consumption	[43]
Desmaisons et al.	2017	A new quality index for benchmarking of different cellulose nanofibrils	[14]
Kepa et al.	2019	Evaluation of properties and specific energy consumption of spinifex-derived lignocellulose fibers produced using different mechanical processes	[44]
Rol et al.	2017	Pilot-scale twin screw extrusion and chemical pretreatment as an energy-efficient method for the production of nanofibrillated cellulose at high solid content	[45]
Berglund et al.	2016	Production potential of cellulose nanofibers from industrial residues: efficiency and nanofiber characteristics	[46]
Jonoobi et al.	2012	Producing low-cost cellulose nanofiber from sludge as new source of raw materials	[47]

Biomass Benchmarking

Biochemical Composition

The initial facet of the biomass benchmarking study was the comparison of biochemical composition across different biomass sources and different stages of biomass processing. Biomass sources under investigation included the sorghum Sugargraze variety (leaf, sheath, upper stem, lower stem), banana (leaf, stem), sugarcane (mulch, bagasse), spinifex (leaf), and softwood (stem). The different stages of biomass processing for which biochemical composition was assessed included before (raw) and after (treated) mild alkali treatment. The key features for the biochemical comparison, as presented in Fig. 1, include:

The lignin content of the softwood biomass in raw form is the highest across all biomass samples at > 30% on a dry weight basis. Furthermore, this high lignin content is maintained after delignification treatment, which indicates low chemical pretreatment efficiency for this biomass source. The complete biomass biochemical composition data is presented in the Supplementary Information.
The chemical pretreatment was most effective for the sorghum sheath section compared to any other biomass source, exhibiting a 3.3 fold reduction in lignin content after treatment. Interestingly, it appears that the chemical treatment had a minimal effect on the banana and softwood lignin content, as their relative content remained relatively constant before and after chemical treatment.
The ash content is effectively reduced to below 7.5% for all biomass sources and plant sections after mild alkali treatment. It is interesting that the chemical pretreatment significantly reduced the softwood ash content, but appeared to have minimal impact on the hemicellulose and lignin content.

Considering the impact of chemical treatment on the ratio between different biochemical components, Fig. 2 demonstrates that sorghum biomass had a significant increase in the C:L ratio of up to an order of magnitude for the sheath section (11.1), which itself was almost 50% greater than the C:L ratio of the Sugargraze leaf and stem sections that exhibited an average C:L ratio of 7.5. This result was also mirrored for the H:L ratio of sorghum biomass after chemical pretreatment.

Interestingly, the raw banana stem biomass had a greater C:L ratio than the leaf section, and this relationship sorghum leaf and stem sections was also mirrored with the banana stem, which exhibited a 2.8 fold greater C:L ratio than the leaf section after pretreatment (Fig. 2). These results may have led to the sustainable processability of banana stem biomass as observed in the nanopaper mechanical property results later in the chapter.

Thermal Stability

Raw data from the thermogravimetric analysis (TGA) experimental series was evaluated using the Mettler STAR^e Evaluation Software to generate the metrics of the initial step change (%), the onset temperature for the first decomposition peak (°C), and the residual material remaining after thermal analysis (%). The stages of thermal decomposition for lignocellulosic biomass can be summarised in the following order:

Region 1: Vapourisation - Vapourisation of bound water initially in equilibrium with the environmental conditions, depending on the relative humidity, typically in a temperature range between 30 ~ 140°C. In addition, low molecular weight volatile components may be released within this region, which are described as extractives in the Biochemical Composition section.
Region 2: Onset degradation - Initial mass loss of solid material, at a typical onset temperature within the range of 140 ~ 300°C, predominantly attributed to degradation of hemicellulose and a smaller degree of lignin components [24].
Region 3: Major degradation - The region where the majority of solid material degradation occurs, typically within the temperature range of 300 ~ 400°C, predominantly attributed to the pyrolysis of cellulose components, in addition to some degree of lignin decomposition.
Region 4: Residual degradation - For the temperature range of greater than ~ 400°C, gradual lignin decomposition occurs until the completion of thermal analysis at 480°C.

The most interesting finding from this analysis was that the leaf and sheath sections exhibited a substantially different thermal stability profile than the stem sections, as described by the degradation onset temperature. As noted in Table 1, a handful of biomass benchmarking studies within the literature have investigated TGA as a biomass comparison tool, especially in the context of their use as the reinforcing phase of polymeric material, which requires elevated temperatures during melt processing [24]. The thermal stability of lignocellulosic biomass has been proposed to be a function of the cellulose polymorphism [48], the cellulose crystallinity [49], in addition to the interaction between macromolecular components potentially associated with sample particle size [50] and biochemical composition [51]. In addition, thermal stability of biomass powder is altered throughout the processing series from raw biomass to chemically treated pulp and to mechanically treated nanofibres. For example, a higher cumulative fibre area, associated with the transformation of cellulose bundles into nanofibres, led to a faster rate of thermal decomposition [52]. The authors noted that the shearing action of mechanical treatment also impacts the crystallinity of the resultant nanofibre material, which introduces another factor of influence to the thermal decomposition profile.

In addition, spinifex leaf exhibited the highest degradation onset temperature, while sugarcane bagasse exhibited the highest initial step change, indicating a high degree of ambient moisture retention for this biomass source. A summary of the key thermal stability metrics is visually described in Fig. 3, while the detailed results from the analysis are outlined in Table 5. The full TGA curve portfolio for all biomass sources is presented in the Supplementary Information.

Table 5

Thermogravimetric analysis (TGA) benchmarking of all biomass sources
Biomass	Section	Step (%)	Onset (°C)	Peak 1 (°C)	Peak 2 (°C)	Residue (%)
Sorghum	Leaf	-5.8	239.4	282.8	346.2	31.5
Sorghum	Sheath	-5.6	236.8	278.6	344.9	30.0
Sorghum	Stem(< 1m)	-3.0	158.7	204.9	321.0	32.9
Sorghum	Stem(> 1m)	-2.9	154.4	200.0	331.5	28.5
Softwood	Stem	-7.2	255.0	-	352.5	27.2
Sugarcane	Mulch	-6.5	231.6	273.6	342.9	32.0
Sugarcane	Bagasse	-0.5	245.6	270.5	345.7	23.8
Banana	Stem	-6.8	225.4	-	296.4	35.4
Banana	Leaf	-5.2	206.1	150.3	307.0	37.6
Spinifex [50]	Leaf	-5.0	262.8	286.0	347.4	28.5

Crystallinity

Similar to the biochemical composition, investigation of biomass crystallinity is relevant for assessment of bulk biomass properties that have been proposed to influence the efficiency of nanofibrillation. In a general sense, the crystallinity of untreated cellulose fibre has been described as distinctly different between wood and non-wood sources, ranging from 55–70% and 90–95%, respectively [53]. Crystallinity and crystallite aspect ratio has been found to be associated with an increase in the stiffness of the plant cell wall for softwood biomass, which can be extrapolated to the stiffness of the cellulose fibres themselves [54]. The increase in cellulose stiffness associated with crystallinity is demonstrated in the linear region of the tensile stress–strain curve, where elastic deformation of the material dominates and the failure of individual fibres and interfibrillar bonds is minimal. Hence, fibre crystallinity dominates the value of material stiffness, alongside the density of the material due to the efficiency of crystalline nanocellulose packing during nanomaterial fabrication [55].

Both chemical and mechanical treatment can impact fibre crystallinity. Through chemical treatment, biomass pulping can swell both amorphous and crystalline regions of the fibre bundles, leading to a reduction in crystallite order and extraction of recalcitrant components such as lignin. This principle promotes a higher efficiency of subsequent mechanical treatment and ultimately an improvement in nanofibre performance [31]. Through mechanical treatment, thermal and mechanical shear forces facilitate the fibrillation of cellulose fibre bundles, in turn partially deconstructing both the crystalline and amorphous regions of the fibre population, leading to a certain reduction in crystallinity [56].

To summarise the XRD curve from each biomass sample into a single value, Fig. 4 presents the crystallinity index calculated from each X-ray diffraction pattern. The crystallinity index (CrI) results present a couple of interesting findings for the comparison between different biomass samples and the implication this has on biomass processing amenability. First, there is a lower CrI value for the leaf section of sorghum biomass. For leaf biomass, the shoulder of the XRD curve at an approximate 2θ angle between 15–18° is less prevalent (see the Supplementary Information for the full portfolio of XRD curves for all biomass sources). This indicates a lower distinction between the amorphous and crystalline components of the cellulose fibres and therefore a higher proportion of amorphous regions. This result fits with our previous investigation, where the leaf section of multiple different sorghum varieties were amenable to the production of a high fine material content after low energy processing [57]. This processing amenability is likely to be contributed to in part by the relatively low fibre crystallinity index for these samples, which enhances accessibility of cellulose fibres to water for increased swelling, enhances accessibility of chemical reagents during pretreatment for increased lignin removal, and promotes cellulose chain disintegration during mechanical treatment, all of which will enhance the nanofibrillation of the lignocellulosic biomass [31].

A second finding is the divergent XRD curve conformation for the banana leaf and stem samples compared to all other biomass types, exhibiting significantly less distinct peaks for the amorphous and crystalline regions of the biomass material. In addition, a handful of sharp intensity peaks exist for both banana samples, which are not found at anywhere near the same extent in the other biomass samples. These peaks have also been previously identified for XRD curves of banana biomass within the literature, which have been attributed to the presence of inorganic matter [58, 59]. Considering that these peaks are also observed in banana biomass samples collected from different countries around the world, as highlighted in the XRD curves from studies investigating the crystallinity of banana leaf samples collected from Egypt (Fig. 5a) and Brazil (Fig. 5b), this component appears endogenous to the plant [58]. However, further investigative work is required to deduce the chemical nature of this component, potentially through energy-dispersive X-ray (EDX) spectroscopy analysis.

Fibre morphology

As established in our previous work, population-level analysis of fibre morphology data is a valuable approach for CNF property evaluation [8]. This approach employs the methodology of principal component analysis (PCA) to visualise the grouping and variance of fibre morphology data across the entire Biomass Benchmarking sample population in a reduced dimensionality format. This is achieved by projecting the data onto a two-dimensional space that maximises the variance between the data points. The PCA diagram in Fig. 6 visualises the difference in fibre morphology between the different biomass sources, with the ellipse around each group of sample points indicating the default 68% normal probability region [61]. The first key features demonstrated within Fig. 6 is the separation of softwood samples from the remainder of the biomass sample population, with fibre morphology exhibiting relatively high fibre width (fibre_w) and fibre coarseness (fibre_coarse). In addition, the sorghum and spinifex samples are predominantly distinct from the other biomass samples, exhibiting high fibre content (fibre_cont) and fibre length (fibre_L), which are associated with high aspect ratio CNF samples.

As a subsequent step to the principal component analysis of CNF morphology, hierarchical clustering was performed to generate biomass groups of greatest similarity. This approach used the ‘Average Linkage’ method to cluster biomass samples together based on their multivariate data similarity. For this clustering visualisation, the location for cutting of the clustering tree was arbitrarily selected such that five clusters were generated, as shown in Fig. 7. This clustering method demonstrates that the softwood biomass sample is the most distally linked to any other biomass type in the sample population, forming a cluster on its own. This result makes sense, considering the relative inefficiency of the mild chemical treatment method on this recalcitrant biomass type, and the subsequent inefficiency of the mechanical nanofibrillation process. Most of the remaining biomass sources formed clusters of their own with the exception to spinifex, which formed a similarity cluster alongside the sorghum biomass sample. The exception to this was sorghum leaf biomass, which showed greater dissimilarity in CNF morphology to the other sections of the same plant type than spinifex did to sorghum biomass. This finding gives weight to the claim that there is meaningful difference in CNF morphology outcomes for different sections of the same plant, and therefore potentially significant difference in the material performance.

Nanopaper mechanical properties

Following principal component analysis and hierarchical clustering of the Biomass Benchmarking fibre morphology data, performance benchmarking of CNF-based nanopaper was conducted against the different biomass types. Biomass samples processed through consistent conditions were fabricated into CNF-based nanopaper and tensile tested to assess nanopaper material performance. The mechanical properties of sorghum nanopaper, especially for the leaf and sheath sections, substantially outperformed nanopaper from sugarcane, banana leaf, spinifex, and softwood biomass at the low and medium energy levels (Fig. 8). Interestingly, the banana stem biomass generated competitive nanopaper performance at the low and medium energy level, potentially attributed to the high water retention capacity observed during chemical pretreatment and the high H:L ratio of banana stem biomass after delignification pretreatment. On the other hand, the softwood biomass presented a substantially inferior paper performance, with samples processed under low and medium energy conditions unable to even form a stable paper structure for tensile testing.

Through utilising this mechanical performance data, this study reapplies the Processing Sustainability Triangle methodology initially developed in our previous work [30]. The Processing Sustainability Triangle, which concisely presents the nanopaper mechanical performance across the processing energy series in terms of biomass processing sustainability, upgradability, and recalcitrance, is shown in Fig. 9. A couple of key conclusions highlighted in this diagram include: (1) softwood biomass as highly recalcitrant to mechanical fibrillation under mild chemical pretreatment conditions, remaining situated in the bottom left vertex of the triangle; (2) spinifex exhibited a very low intercept value, indicating poor processing sustainability at low energy conditions, but exhibiting some degree of upgradability at high energy conditions; (3) the sugarcane bagasse, mulch, and banana leaf samples exhibited poor processing sustainability and upgradability compared to other biomass samples in the study; and (4) banana leaf exhibited a high degree of processing sustainability at the low energy condition but minimal upgradability in subsequent processing, while the sorghum sheath and upper stem sections exhibited high processing sustainability and upgradability due to being situated on the hypotenuse line.

Mechanical Process Benchmarking

Nanopaper mechanical properties

To assess the influence of mechanical treatment type on CNF properties, sorghum biomass was mechanically treated with four alternative processes: High pressure homogenisation, high energy bead milling, twin screw extrusion, and Silverson blending. Tensile index properties range from 32.2 Nm/g for the low energy Silverson process, to 98.9 Nm/g for the high energy HPH. The rank order of mechanical treatment types in terms of nanopaper strength is: HPH > Bead milling > Extrusion > Silverson, as shown in Fig. 10(a). As expected, the tensile index of CNF-based nanopaper increased from low to high energy processing for each treatment methods.

Once nanopaper tensile index values were standardised by the energy demand of the mechanical treatment process, Silverson becomes the most favourable mechanical process at an energy standardised tensile index of 4528–7593 Nmg-1 / kWh. However, while this treatment type presents the highest energy standardised tensile index values, if a minimum material strength is required for the application of interest and the Silverson properties are below the particular threshold (e.g. 50 Nm/g), then the process remains unviable and the sustainable energy consumption becomes irrelevant. Nevertheless, the relatively low energy fibrillation treatments including Silverson blending and twin-screw extrusion show promise as a pretreatment step in combination with addition mechanical treatment.

Fibre morphology

Utilising the previously introduced method of principal component analysis, the results in Fig. 11 demonstrate the comparison of different mechanical treatment methods in terms of their CNF morphology. Here, the population-level fibre analysis conducted through the MorFi device (Techpap, France) and the subsequent PCA visualisation elucidates a couple of hidden trends within the data. First, the processing sample projections onto the first two principal components are quite effective at differentiating the CNF samples in terms of their fibre morphology parameters, with > 72% of the variance explained across the first two principal components.

Second, the homogenisation treatment in yellow is quite effective at maintaining a relatively high fibre length for the material throughout processing, as was highlighted in Table 2 regarding generation of high specific surface area and aspect ratio fibre material [33]. The notion that homogenisation generates high fibre aspect ratio, as a result of the aligned shear force for material flowing through the small homogenisation orifice is confirmed in Fig. 11. On the other hand, Silverson and bead milling operate in a similar manner of high impact shearing force, which results in a high fine content but substantially lower fibre length. This is shown in Fig. 11, with the Silverson and bead mill sample ellipses localised near the fine length and fine area MorFi parameter arrows. These findings concur with the previously established mechanisms outlined in Table 2, associated with lower fibre aspect ratio material for these processing methods [6].

Third, twin-screw extrusion at increasingly high feedstock solids content led to an increase in the content of kinked fibres, which has previously been associated with a higher degree of nanocellulosic fibre deconstruction [62]. This mechanism of action for extrusion operates in an opposing manner compared to Silverson blending, which generates a relatively high proportion of fine material, but with minimal observation of kinked fibres. This result is intuitive, as the shear forces exerted through the Silverson blending process are substantially higher than in extrusion, where the dominant mechanism of fibrillation is pulp kneading between rotating screw elements and the barrel wall. The ability to extract this level of information through the presentation of two summarised diagrams highlights the power of the PCA method to visualise population-level differences between groups. The PCA method is especially when the disparity between groups is high, as is the case for CNF material generated through processing methods with disparate mechanisms of action.

Nanopaper Performance Benchmarking

In addition to the importance of benchmarking different biomass types and mechanical processing methods in an attempt to understand factors that influence product quality and processing sustainability, it is also important to synthesise these results with a product-focused benchmarking study. This section of the study outlines a bibliometric methodology to survey the CNF literature to extract articles that present relevant information on both nanopaper mechanical performance and processing energy consumption data. The aim of this study is to elucidate the energy-standardised nanopaper property landscape portrayed within the literature. Here, CNF samples could be derived from any biomass source or processed through any fibre deconstruction route, with the comparative results between the relevant studies providing additional insights into promising biomass sources and processing methods. Therefore, this study is situated within the spectrum between a purely qualitative literature review, and a comprehensive, quantitative life cycle assessment.

The quantitative results outlining the energy consumption and mechanical property of nanopaper materials generated within the literature are presented in Table 6. Working through this tables in descending order, there are a number of interesting results worth highlighting from each of the relevant publications. Firstly, the results from the mechanical process benchmarking study performed in the current study are comparable with the previously reported benchmarking data in the literature. In this study, tensile index ranged from 32.2–98.9 Nm/g for the Silverson-L and HPH-H samples, respectively, with the energy consumption for these samples ranging between 1.7–38.2 kWh/kg. For this study, the feedstock was untreated sorghum biomass (Sugargraze variety - all sections in equal proportions), which could be considered economic burden free (no procurement cost) if sourced as the byproduct of grain, sugar, or fodder production in an industrial setting. Pretreatment involved low energy knife milling and a mild alkali treatment (2% NaOH). From the range of mechanical processing types, low energy Silverson blending (SS-L) generated nanopaper material with the highest TI/energy ratio of 19.1 (\(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\) ). However, considering a material performance threshold of 50 Nm/g, the highest performing sample was three pass extrusion (X-H) with a TI/energy ratio of 8.3 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)).

Spence et al. (2011) presented a comprehensive comparative study of microfibrillated cellulose (MFC) films produced from different biomass sources and processed through different mechanical treatment routes, including Masuko micro-grinding, microfluidisation, and homogenisation [40]. High mechanical strength films of up to 135 Nm/g were generated through microfluidisation, with a TI/energy ratio of 57.7 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)). In addition, Masuko treated bleached hardwood pulp without mechanical pretreatment reached a tensile index of 113 Nm/g and a TI/energy ratio of 72.9 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)). However, one caveat for this study is that the MFC films were produced through a casting-evaporation technique, which is time consuming and not an industrially viable method of film preparation.

Josset et al. (2018) benchmarked the mechanical properties of three different biomass sources processed though Masuko micro-grinding [41]. Elemental Chlorine Free (ECF) bleached softwood pulp generated the highest performing film material of 96.1 Nm/g. However, ECF4 was the sample with the highest TI/energy ratio at 66.5 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)), indicating that low energy Masuko treatment can generate competitive materials without excessive energy input. In addition, another relevant outcome from this study is the superior mechanical performance of bleached softwood pulp in comparison to recycled newspaper and wheat straw, which has different degrees of pretreatment prior to evaluation in this study, a factor that has not been quantified in the current benchmarking literature review. Although, the authors estimate that the procurement costs of softwood pulp, recycled newspaper, and wheat straw were €700, €150, and €400, respectively.

Kargupta et al. (2021) investigated the energy consumption and mechanical performance of three different wood feedstock for the production of nanopaper handsheets using PFI milling and disc refining as the central mechanical treatment process [42]. While results presented within the article indicate that these processes are sustainable methods for the production of CNFs, a couple of caveats must be considered. First, the disintegration pretreatment process that was performed twice before handsheet fabrication was not included in the total energy consumption calculation, but is estimated to have a significant energy consumption of approximately 20.8 kWh/kg per batch (W. Batchelor, personal communication, December 13, 2021). Second, the energy consumption value for the disc refiner process presented within the article was calculated as a net power consumption value, neglecting the contribution of the no-load power (water flowing through the refiner only). However, as the study aimed to estimate the total energy consumption of the CNF production process, this is an untenable assumption. In light of this, the updated total energy consumption of the disc refining process is estimated to be at a significantly higher range than the reported data. The authors of this study were unable to provide an updated figure for the total energy consumption when contacted (W. Batchelor, personal communication, December 13, 2021).

Ang et al. (2019) compared the mechanical performance of nanopaper handsheets prepared from never-dried bleached eucalyptus kraft (BEK) pulp using a combination of disintegration, PFI milling and disc refining [43]. Handsheet performance ranged from 30.1–118.2 Nm/g and a TI/energy ratio at 1.4–3.9 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)), with the highest TI/energy ratio achieved by the lowest energy PFI milled sample (BEK10k). Kepa et al. (2019) demonstrated that the highest TI/energy ratio (above the 50 Nm/g threshold) was achieved by the lower energy bead milling sample (M20) at 4 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)) [44].

Rol et al. (2017) achieved the production of nanopaper handsheet material with a TI/energy ratio of 5–7.1 ( \(\frac{\text{N}\text{m}}{\text{g}}/ \frac{\text{k}\text{W}\text{h}}{\text{k}\text{g}}\)) for samples above the 50 Nm/g threshold [45]. However, these samples underwent an enzymatic and TEMPO pretreatment, which has a long pretreatment time and high chemical reagent cost, respectively. However, these factors were not incorporated into the TI/energy ratio metric, but are factors that should be considered when evaluating these results.

Berglund et al. (2016) sourced industrial byproducts in the form of carrot residue and brewers spent grain (BSG) to generate incredibly high performance nanopaper material, reaching a maximum tensile index of 181 Nm/g for the carrot residue-derived sample. However, it should be noted that this study employed a couple of an additional processing steps that may contribute to the high performance of their nanopaper materials. First, the pulp feedstock materials were sourced as a by-product from industrial food and beverage production facilities, with an unquantifiable degree of pretreatment applied to these samples. Second, a bleaching chemical pretreatment was applied to the pulp (once for the carrot residue, three times for the BSG), which was not included in the current energy-standardised nanopaper performance metric due to challenges in assessing chemical costs for studies operating at different production scales. Third, additional mechanical treatments were performed yet not quantified in terms of their specific energy consumption, including a Silverson blending pretreatment (which had to be assumed as approximately 2.5 kWh/kg) and a shear mixing step (10,000 rpm, 10 min) prior to nanopaper fabrication. Fourth, the nanopaper fabrication involved an initial dewatering step through vacuum filtration, followed by room temperature drying for 12 hours and hot pressing at 85°C under a pressure of 30 kPa for 40 min. These additional processing steps challenge the scalability of the process to generate these high performance nanopaper materials, yet were unable to be quantified in the current benchmarking study, which contributed to the outperformance of these samples compared to the other studies presented in this benchmarking work. Lastly, Jonoobi et al. (2012) compared the nanopaper fabrication potential between dissolving cellulose (CF) and dissolving cellulose sludge, denoted as sludge (SF), with the latter achieving a nanopaper tensile index of ~ 97 Nm/g after Silverson blending (3000 rpm, 10 min) and recirculated ultrafine grinding (Masuko, 1440–3500 rpm) until a CNF gel was formed. The sludge material demonstrated a remarkable capacity to achieve a high degree of nanofibrillation, assessed through Congo red adsorption for specific surface area measurement, and to subsequently generate a relatively high strength nanopaper material without additional chemical pretreatment steps.

In summary, Fig. 12 provides a visual representation of the tensile index and energy consumption results, which is presented in the form of a polygon plot (‘polygon’ function, R). This diagram provides a snapshot of the nanopaper processing-property landscape of studies that include both nanopaper mechanical performance and energy consumption.

Table 6

Literature review results for nanopaper performance benchmarking
Biomass	Sample name	Pretreatment	Mechanical treatment	Mechanical properties		Totals		Ref.
Biomass	Sample name	Pretreatment	Mechanical treatment	TI (Nm/g)	E (GPa)	Energy (kWh/kg)	TI / Energy (Nm/g) / (kWh/kg)	Ref.
Sorghum	HPH-L	2% NaOH, 1:20 SC (0.4 kg NaOH/kg) Cutting mill (∼0.2 kW/kg)	HPH (9.2 kWh/kg)	72.87	1.87	9.37	7.78	This study
	HPH-M		HPH (22.1 kWh/kg)	86.89	4.96	22.27	3.90
	HPH-H		HPH (38.0 kWh/kg)	98.93	3.18	38.15	2.59
	BM-L		Bead mill (15.1 kWh/kg)	64.75	2.68	15.29	4.23
	BM-H		Bead mill (20.2 kWh/kg)	73.18	2.29	20.39	3.59
	SS-L		Silverson (1.5 kWh/kg)	32.23	1.07	1.69	19.07
	SS-M		Silverson (3.6 kWh/kg)	39.64	1.29	3.82	10.38
	SS-H		Silverson (3.6 kWh/kg)	46.46	1.39	3.82	12.16
	X-L		Extrusion (3.8 kWh/kg)	49.05	2.42	3.97	12.36
	X-M		Extrusion (4.8 kWh/kg)	48.85	2.09	4.95	9.87
	X-H		Extrusion (6.2 kWh/kg)	53.16	3.06	6.41	8.29
Never-dried bleached Kraft hardwood pulp	B_Pulp	-	-	27	2.53	0.00	27.00	[40]
	B_MG1	-	Masuko (1.6 kWh/kg)	113	4.55	1.55	72.90
	B_V	Valley beater (2.01 kWh/kg)	-	74	5.82	2.01	36.85
	B_HV1		HPH (19.9 kWh/kg)	92	6.31	21.89	4.20
	B_MG1		Masuko (1.6 kWh/kg)	113	4.55	1.55	72.90
	B_MGV1		Masuko (1.6 kWh/kg)	93	4.44	3.56	26.14
	B_MFV1		Microfluidiser (0.9 kWh/kg)	120	5.38	2.94	40.83
	B_MFV2		Microfluidiser (0.3 kWh/kg)	134	5.32	2.34	57.22
	B_MFV3		Microfluidiser (0.3 kWh/kg)	135	5.38	2.34	57.65
	B_MFV4		Microfluidiser (0.8 kWh/kg)	135	4.24	2.83	47.74
	B_MFV5		Microfluidiser (1.1 kWh/kg)	124	5.35	3.12	39.75
Never-dried unbleached Kraft hardwood pulp	Pulp	-	-	27	2.67	0.00	27.00
	MG1	-	Masuko (1.6 kWh/kg)	105	6.29	1.55	67.74
	V	Valley beater (2.01 kWh/kg)	-	69	9.09	2.01	34.36
	HV1		HPH (19.9 kWh/kg)	109	7.71	21.89	4.98
	MFV6		Microfluidiser (1.1 kWh/kg)	148	6.22	3.12	47.44
	MFV7		Microfluidiser (0.9 kWh/kg)	54	4.67	2.88	18.73
	MFV8		Microfluidiser (0.5 kWh/kg)	93	5.42	2.55	36.47
	MFV9		Microfluidiser (0.3 kWh/kg)	138	4.83	2.29	60.36
Elemental Chlorine Free bleached softwood pulp	ECF0	-	-	30.10	3.79	0.00	30.10	[41]
	ECF2		Masuko (0.2 kWh/kg)	45.88	4.83	0.24	45.88
	ECF4		Masuko (1.2 kWh/kg)	81.29	7.99	1.22	66.47
	ECF6		Masuko (2.7 kWh/kg)	91.65	8.66	2.71	33.76
	ECF8		Masuko (4.0 kWh/kg)	96.13	8.47	3.96	24.26
	ECF10		Masuko (5.3 kWh/kg)	94.12	8.05	5.25	17.93
Recycled newspapers	NP0		-	40.31	5.21	0.00	40.31
	NP2		Masuko (1.2 kWh/kg)	50.98	5.90	1.23	41.34
	NP4		Masuko (2.5 kWh/kg)	60.10	6.77	2.49	24.17
	NP6		Masuko (4.0 kWh/kg)	69.54	7.41	3.97	17.52
	NP8		Masuko (5.2 kWh/kg)	73.40	7.74	5.21	14.08
	NP10		Masuko (6.8 kWh/kg)	78.20	8.11	6.75	11.58
Wheat straw	WS0		-	6.91	1.48	0.00	6.91
	WS2		Masuko (0.7 kWh/kg)	15.41	2.65	0.74	15.41
	WS4		Masuko (2.0 kWh/kg)	25.31	3.97	1.99	12.75
	WS6		Masuko (3.0 kWh/kg)	35.98	5.24	2.98	12.08
	WS8		Masuko (4.5 kWh/kg)	46.65	6.30	4.47	10.44
	WS10		Masuko (5.8 kWh/kg)	45.26	6.04	5.75	7.87
Unrefined Aus. Bleached Euc. Kraft pulp	ABEK15k	Disintegrator x2 (41.67 kWh/kg)	PFI mill (5.9 kWh/kg)	59.09	-	47.59	1.24	[42]
	ABEK30k		PFI mill (5.9 kWh/kg)	87.93		53.99	1.63
	ABEK50k		PFI mill (5.9 kWh/kg)	113.53		59.11	1.92
Dried unrefined Brazilian Bleached Eucalyptus Kraft	BBEK15k		PFI mill (5.9 kWh/kg)	66.10		47.59	1.39
	BBEK30k		PFI mill (5.9 kWh/kg)	77.08		53.99	1.43
	BBEK50k		PFI mill (5.9 kWh/kg)	88.67		59.11	1.50
	BBEK1		Disc refiner (??? kWh/kg)	~ 20.48		???	???
	BBEK2		Disc refiner (??? kWh/kg)	~ 52.38		???	???
	BBEK3		Disc refiner (??? kWh/kg)	~ 57.62		???	???
	BBEK4		Disc refiner (??? kWh/kg)	~ 85.24		???	???
Unrefined Northern Bleached Softwood Kraft pulp	NBSK15k		PFI mill (5.9 kWh/kg)	41.24		47.59	0.87
	NBSK30k		PFI mill (5.9 kWh/kg)	63.97		53.99	1.18
	NBSK50		PFI mill (5.9 kWh/kg)	75.79		59.11	1.28
	NBSK1		Disc refiner (??? kWh/kg)	~ 17.14		???	???
	NBSK2		Disc refiner (??? kWh/kg)	~ 60.48		???	???
	NBSK3		Disc refiner (??? kWh/kg)	~ 100.95		???	???
	NBSK4		Disc refiner (??? kWh/kg)	~ 110.0		???	???
	NBSK5		Disc refiner (??? kWh/kg)	-		???	???
Never-dried unrefined bleached eucalyptus Kraft pulp	BEK0k	Disintegrator (20.8 kWh/kg)	-	30.05	2.61	20.83	1.44	[43]
	BEK10k		PFI mill (3.5 kWh/kg)	95.63	7.85	24.38	3.92
	BEK30k		PFI mill (10.6 kWh/kg)	105.48	8.59	31.46	3.35
	BEK50k		PFI mill (17.7 kWh/kg)	101.35	10.02	38.54	2.63
	BEK10k1P		PFI mill (3.5 kWh/kg) Disc refiner (7.2 kWh/kg)	118.16	8.53	31.58	3.74
	BEK30k1P		PFI mill (10.6 kWh/kg) Disc refiner (7.2 kWh/kg)	116.30	9.75	38.66	3.01
	BEK50k3P		PFI mill (17.7 kWh/kg) Disc refiner (21.6 kWh/kg)	114.63	10.89	60.14	1.91
Spinifex grass	H	2% NaOH, 1:10 SC (0.2 kg NaOH/kg) Cutting mill (∼0.2 kW/kg)	HPH (43.3 kWh/kg)	84	-	43.3	1.9	[44]
	X1		Extrusion (0.6 kWh/kg)	32		0.8	32
	Extrusion (1.3 kWh/kg)		40	1.5		30.8
	X6		Extrusion (2.9 kWh/kg)	40		3.1	13.8
	X10		Extrusion (5 kWh/kg)	46		5.2	9.2
	X3M20		Extrusion (1.3 kWh/kg) Bead mill (20.2 kWh/kg)	77		21.5	3.6
	X6M20		Extrusion (2.9 kWh/kg) Bead mill (20.2 kWh/kg)	87		23.1	3.8
	X10M20		Extrusion (5 kWh/kg) Bead mill (20.2 kWh/kg)	79		25.2	3.1
	M20		Bead mill (20.2 kWh/kg)	80		20.2	4.0
	M40		Bead mill (33.2 kWh/kg)	82		33.2	2.5
Eucalyptus bleached kraft pulp	R-E-X1	Enzymatic Disc refiner (0.68 kWh/kg)	Extrusion (1.3 kWh/kg)	23.04	10.1	2.00	11.52	[45]
	R-E-X7	Enzymatic (300 ECU/g) Disc refiner (0.68 kWh/kg)	Extrusion (5.7 kWh/kg)	37.58	15.1	6.34	5.93
	R-E-M	Enzymatic (300 ECU/g) Disc refiner (0.68 kWh/kg)	Masuko (15.9 kWh/kg)	82.22	21.3	16.53	4.97
	R-T-X1	TEMPO Disc refiner (0.68 kWh/kg)	Extrusion (4.4 kWh/kg)	44.59	1.4	5.06	8.81
	R-T-X7	TEMPO Disc refiner (0.68 kWh/kg)	Extrusion (12.3 kWh/kg)	36.28	11.2	13.00	2.79
	R-T-M	TEMPO Disc refiner (0.68 kWh/kg)	Masuko (11.2 kWh/kg)	87.73	20.8	12.38	7.09
Carrot residue	C	2% NaOH, 1:10 SC 1.7% NaClO₂, 1:10 SC Silverson (∼2.5 kWh/kg) Shear mixer (10k rpm, 10min)	-	160.36	7.6	2.5	64.14	[46]
Carrot residue	C-M		Masuko (0.9 kWh/kg)	181.03	12.9	3.4	53.24
Brewer spent grains	BSG		-	53.64	4.1	2.5	21.46
Brewer spent grains	BSG-M		Masuko (21 kWh/kg)	111.57	9.9	23.5	4.75
Dissolving cellulose	CF	Silverson blending (∼2.5 kWh/kg)	-	69.70	1.7	2.5	27.88	[47]
Dissolving cellulose	CNF		Masuko (1.7 kWh/kg)	62.84	5.1	4.2	14.96
Paper sludge	SF		-	43.24	1.3	2.5	17.30
Paper sludge	SNF		Masuko (1.3 kWh/kg)	97.36	8.0	3.8	25.62

The biomass benchmarking study compared the nanofibre and nanopaper performance of different feedstock types, including sorghum, softwood, spinifex, sugarcane (bagasse, mulch), and banana (leaf, stem) biomass. The mechanical properties of biomass-derived nanopaper material revealed the high relative performance of sorghum and banana stem biomass processed under low energy conditions. These results open the door to further investigation into the physiological mechanisms that underpin sustainable processability of sorghum and banana stem biomass, which was initially investigated through characterisation of biomass chemical composition, crystallinity, and thermal degradation profile. However, the conventionally desirable traits of high hemicellulose content and low lignin content were only loosely correlated with nanopaper strength at low energy processing. This highlights that biomass features outside of the standard suite of characterisation, relating to plant cell wall organisation and structural orientation, are more likely to influence processing amenability. Mechanical process benchmarking was primarily conducted to evaluate the relative processing energy efficiency across different nanofibrillation unit operations as a function of nanopaper mechanical properties, in addition to differentiating between the mechanisms of action for each mechanical process type. Given a minimum nanopaper performance baseline of 50 Nm/g tensile index, the twin-screw extrusion process generates the most energy efficient fibrillated cellulose material, followed by low energy homogenisation. In addition, the investigation of fibre morphology across difference mechanical process methods, as assessed through PCA visualisation, reinforces existing understanding about the relationship between processing type and the resultant fibre morphology. For example, bead milled CNFs exhibited a lower mean fibre aspect ratio, while a relatively high fibre aspect ratio was maintained for homogenised samples. Lastly, the benchmarking of nanopaper performance portrayed the mechanical strength and energy consumption landscape for CNF-based nanopaper material. This visual summary indicated that sorghum-derived CNF materials were situated within the heart of the nanopaper performance landscape, while not achieving tensile index properties in excess of 100 Nm/g at relatively low energy consumption as was demonstrated for a couple of studies [40, 46]. This exercise provided interesting information on the energy-standardised nanopaper performance across applicable research studies within the nanocellulose literature, of which there were a single digit number of publications that fulfilled the scoping criteria. This highlights an important gap within the literature relating to the lack of comprehensive techno-economic data presented in CNF research studies, which limits the investigation of product benchmarking.

CNF – Cellulose nanofiber

CNC – Cellulose nanocrystal

MFC – Microfibrillated cellulose

DL – Delignification

HPH - High pressure homogenisation

PCA – Principal component analysis

XRD – X-ray diffraction

CrI – Crystallinity index

TGA – Thermogravimetric analysis

TI – Tensile index

Funding

This research was supported by the Australian Government Grains Research and Development Corporation (GRDC) Research Scholarship (Ref: UOQ1903-005RSX). The funding source had no involvement in the preparation, writing or submission of this article.

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Author Contributions

Jordan Pennells: Conceptualization, Methodology, Software, Investigation, Writing - Original Draft, Writing - Reviewing and Editing, Visualization, Project administration, Funding acquisition. Céline Chaléat: Resources, Supervision. Darren Martin: Supervision, Funding acquisition, Writing - Reviewing and Editing.

Supplementary Information

The supplementary information provides comprehensive data on the biochemical composition values, x-ray diffraction curves, and thermogravimetric analysis curves for all biomass samples. In addition, benchmarking of alternative mechanical property results besides nanopaper tensile index are provided, including nanopaper toughness and Young’s modulus.

Acknowledgements

The authors gratefully acknowledge the Australian Government for their provision of a Research Training Program (RTP) scholarship, the University of Queensland for their provision of a living allowance scholarship and continental scholarship, and the Grains Research and Development Corporation (GRDC) for their support of this research through the provision of their GRDC Research Scholarship (GRS). We acknowledge Sunshine Coast banana farmer Ray Nelson for his contribution of banana biomass, and A/Prof Zhanying Zhang from Queensland University of Technology (QUT) for his contribution of sugarcane bagasse for this study. The authors also acknowledge the contribution of Dugalunji Aboriginal Corporation (DAC) on behalf of the Indjalandji-Dhidhanu peoples through use of their equipment and in-kind support.

Pennells J, Godwin ID, Amiralian N, Martin DJ (2020) Trends in the production of cellulose nanofibers from non-wood sources. Cellulose 27:575–593. https://doi.org/10.1007/s10570-019-02828-9
Dhali K, Ghasemlou M, Daver F et al (2021) A review of nanocellulose as a new material towards environmental sustainability. Sci Total Environ 775:145871. https://doi.org/10.1016/j.scitotenv.2021.145871
Dufresne A (2017) Nanocellulose: from nature to high performance tailored materials. Walter de Gruyter GmbH & Co KG
Nechyporchuk O, Belgacem MN, Bras J (2016) Production of cellulose nanofibrils: A review of recent advances. Ind Crops Prod 93:2–25. https://doi.org/10.1016/j.indcrop.2016.02.016
Jonoobi M, Oladi R, Davoudpour Y et al (2015) Different preparation methods and properties of nanostructured cellulose from various natural resources and residues: a review. Cellulose 22:935–969. https://doi.org/10.1007/s10570-015-0551-0
Kargarzadeh H, Mariano M, Gopakumar D et al (2018) Advances in cellulose nanomaterials. Cellulose 25:2151–2189. https://doi.org/10.1007/s10570-018-1723-5
Thomas B, Raj MC, Athira BK et al (2018) Nanocellulose, a Versatile Green Platform: From Biosources to Materials and Their Applications. Chem Rev 118:11575–11625. https://doi.org/10.1021/acs.chemrev.7b00627
Pennells J, Heuberger B, Chaléat C, Martin DJ (2022) Assessing cellulose micro/nanofibre morphology using a high throughput fibre analysis device to predict nanopaper performance. Cellulose. https://doi.org/10.1007/s10570-021-04405-5
Balea A, Blanco A, Delgado-Aguilar M et al (2021) Nanocellulose Characterization Challenges. BioResources 16:4382–4410. https://doi.org/10.15376/biores.16.2.Balea
Ahvenainen P, Kontro I, Svedström K (2016) Comparison of sample crystallinity determination methods by X-ray diffraction for challenging cellulose I materials. Cellulose 23:1073–1086. https://doi.org/10.1007/s10570-016-0881-6
Reid MS, Villalobos M, Cranston ED (2017) Benchmarking Cellulose Nanocrystals: From the Laboratory to Industrial Production. Langmuir 33:1583–1598. https://doi.org/10.1021/acs.langmuir.6b03765
Kim KS, Knotts TL, Jones SC (2008) Characterizing viability of small manufacturing enterprises (SME) in the market. Expert Syst Appl 34:128–134. https://doi.org/10.1016/j.eswa.2006.08.009
Delepierre G, Vanderfleet OM, Niinivaara E et al (2021) Benchmarking Cellulose Nanocrystals Part II: New Industrially Produced Materials. Langmuir 37:8393–8409. https://doi.org/10.1021/acs.langmuir.1c00550
Desmaisons J, Boutonnet E, Rueff M et al (2017) A new quality index for benchmarking of different cellulose nanofibrils. Carbohydr Polym 174:318–329. https://doi.org/10.1016/j.carbpol.2017.06.032
Kriechbaum K, Munier P, Apostolopoulou-Kalkavoura V, Lavoine N (2018) Analysis of the porous architecture and properties of anisotropic nanocellulose foams - A novel approach to assess the quality of cellulose nanofibrils (CNFs). ACS Sustain Chem Eng 6:11959–11967. https://doi.org/10.1021/acssuschemeng.8b02278
Li T, Chen C, Brozena AH et al (2021) Developing fibrillated cellulose as a sustainable technological material. Nature 590:47–56. https://doi.org/10.1038/s41586-020-03167-7
Sharma PR, Sharma SK, Lindström T, Hsiao BS (2020) Nanocellulose-Enabled Membranes for Water Purification: Perspectives. Adv Sustain Syst 4:1–28. https://doi.org/10.1002/adsu.201900114
Mariño MA, Cypriano D, Tasic L (2021) Agroindustry Residues as a Source for Cellulose Nanofibers Production. J Braz Chem Soc 32:878–888. https://doi.org/10.21577/0103-5053.20200239
Okahisa Y, Furukawa Y, Ishimoto K et al (2018) Comparison of cellulose nanofiber properties produced from different parts of the oil palm tree. Carbohydr Polym 198:313–319. https://doi.org/https://doi.org/10.1016/j.carbpol.2018.06.089
Yue D, Qian X (2018) Isolation and Rheological Characterization of Cellulose Nanofibrils (CNFs) from Coir Fibers in Comparison to Wood and Cotton. Polym. 10
Alemdar A, Sain M (2008) Isolation and characterization of nanofibers from agricultural residues – Wheat straw and soy hulls. Bioresour Technol 99:1664–1671. https://doi.org/10.1016/j.biortech.2007.04.029
Hassan ML, Mathew AP, Hassan EA et al (2012) Nanofibers from bagasse and rice straw: process optimization and properties. Wood Sci Technol 46:193–205. https://doi.org/10.1007/s00226-010-0373-z
Yahya M, Chen YW, Lee HV, Hassan WHW (2018) Reuse of Selected Lignocellulosic and Processed Biomasses as Sustainable Sources for the Fabrication of Nanocellulose via Ni(II)-Catalyzed Hydrolysis Approach: A Comparative Study. J Polym Environ 26:2825–2844. https://doi.org/10.1007/s10924-017-1167-2
Deepa B, Abraham E, Cordeiro N et al (2015) Utilization of various lignocellulosic biomass for the production of nanocellulose: a comparative study. Cellulose 22:1075–1090. https://doi.org/10.1007/s10570-015-0554-x
Chaker A, Mutje P, Vilaseca F, Boufi S (2013) Reinforcing potential of nanofibrillated cellulose from nonwoody plants. Polym Compos 34:1999–2007. https://doi.org/https://doi.org/10.1002/pc.22607
Espinosa E, Sánchez R, Otero R et al (2017) A comparative study of the suitability of different cereal straws for lignocellulose nanofibers isolation. Int J Biol Macromol 103:990–999. https://doi.org/10.1016/j.ijbiomac.2017.05.156
Lee HV, Hamid SBA, Zain SK (2014) Conversion of Lignocellulosic Biomass to Nanocellulose: Structure and Chemical Process. Sci World J 2014:631013. https://doi.org/10.1155/2014/631013
Balea A, Merayo N, De La Fuente E et al (2017) Assessing the influence of refining, bleaching and TEMPO-mediated oxidation on the production of more sustainable cellulose nanofibers and their application as paper additives. Ind Crops Prod 97:374–387. https://doi.org/10.1016/j.indcrop.2016.12.050
Delgado-Aguilar M, González I, Tarrés Q et al (2016) The key role of lignin in the production of low-cost lignocellulosic nanofibres for papermaking applications. Ind Crops Prod 86:295–300. https://doi.org/10.1016/j.indcrop.2016.04.010
Pennells J, Mellor V, Martin DJ (2022) Novel Methodology to Visualize Biomass Processing Sustainability & Cellulose Nanofiber Product Quality. ACS Sustain Chem Eng. https://doi.org/10.1021/acssuschemeng.1c08476
Ghasemi M, Alexandridis P, Tsianou M (2018) Dissolution of Cellulosic Fibers: Impact of Crystallinity and Fiber Diameter. Biomacromolecules 19:640–651. https://doi.org/10.1021/acs.biomac.7b01745
Lavoine N, Desloges I, Dufresne A, Bras J (2012) Microfibrillated cellulose - Its barrier properties and applications in cellulosic materials: A review. Carbohydr Polym 90:735–764. https://doi.org/10.1016/j.carbpol.2012.05.026
Osong SH, Norgren S, Engstrand P (2016) Processing of wood-based microfibrillated cellulose and nanofibrillated cellulose, and applications relating to papermaking: a review. Cellulose 23:93–123. https://doi.org/10.1007/s10570-015-0798-5
Djafari Petroudy SR, Chabot B, Loranger E et al (2021) Recent Advances in Cellulose Nanofibers Preparation through Energy-Efficient Approaches: A Review. Energies 14:6792. https://doi.org/10.3390/en14206792
Yi T, Zhao H, Mo Q et al (2020) From cellulose to cellulose nanofibrils—a. comprehensive review of the preparation and modification of cellulose nanofibrils
Silva LE, dos Santos A, de Torres A et al (2021) Redispersion and structural change evaluation of dried microfibrillated cellulose. Carbohydr Polym 252. https://doi.org/10.1016/j.carbpol.2020.117165
Mao R, Meng N, Tu W, Peijs T (2017) Toughening mechanisms in cellulose nanopaper: the contribution of amorphous regions. Cellulose 24:4627–4639. https://doi.org/10.1007/s10570-017-1453-0
Benítez AJ, Walther A (2017) Cellulose nanofibril nanopapers and bioinspired nanocomposites: A review to understand the mechanical property space. J Mater Chem A 5:16003–16024. https://doi.org/10.1039/c7ta02006f
International Organization for Standardization (2001) BS EN ISO 5269-2:2001 - Pulps. Preparation of laboratory sheets for physical testing. Rapid-Kothen method
Spence KL, Venditti RA, Rojas OJ et al (2011) A comparative study of energy consumption and physical properties of microfibrillated cellulose produced by different processing methods. Cellulose 18:1097–1111. https://doi.org/10.1007/s10570-011-9533-z
Josset S, Orsolini P, Siqueira G et al (2014) Energy consumption of the nanofibrillation of bleached pulp, wheat straw and recycled newspaper through a grinding process. Nord Pulp Pap Res J 29:167–175. https://doi.org/10.3183/npprj-2014-29-01-p167-175
Kargupta W, Seifert R, Martinez M et al (2021) Sustainable production process of mechanically prepared nanocellulose from hardwood and softwood: A comparative investigation of refining energy consumption at laboratory and pilot scale. Ind Crops Prod 171:113868. https://doi.org/10.1016/j.indcrop.2021.113868
Ang S, Haritos V, Batchelor W (2019) Effect of refining and homogenization on nanocellulose fiber development, sheet strength and energy consumption. Cellulose 26:4767–4786. https://doi.org/10.1007/s10570-019-02400-5
Kępa K, Chaléat CM, Amiralian N et al (2019) Evaluation of properties and specific energy consumption of spinifex-derived lignocellulose fibers produced using different mechanical processes. Cellulose 26:6555–6569. https://doi.org/10.1007/s10570-019-02567-x
Rol F, Karakashov B, Nechyporchuk O et al (2017) Pilot-Scale Twin Screw Extrusion and Chemical Pretreatment as an Energy-Efficient Method for the Production of Nanofibrillated Cellulose at High Solid Content. ACS Sustain Chem Eng 5:6524–6531. https://doi.org/10.1021/acssuschemeng.7b00630
Berglund L, Noël M, Aitomäki Y et al (2016) Production potential of cellulose nanofibers from industrial residues: Efficiency and nanofiber characteristics. Ind Crops Prod 92:84–92. https://doi.org/10.1016/j.indcrop.2016.08.003
Jonoobi M, Mathew AP, Oksman K (2012) Producing low-cost cellulose nanofiber from sludge as new source of raw materials. Ind Crops Prod 40:232–238. https://doi.org/10.1016/j.indcrop.2012.03.018
Yue Y, Zhou C, French AD et al (2012) Comparative properties of cellulose nano-crystals from native and mercerized cotton fibers. Cellulose 19:1173–1187. https://doi.org/10.1007/s10570-012-9714-4
Wang Z, McDonald AG, Westerhof RJM et al (2013) Effect of cellulose crystallinity on the formation of a liquid intermediate and on product distribution during pyrolysis. J Anal Appl Pyrolysis 100:56–66. https://doi.org/https://doi.org/10.1016/j.jaap.2012.11.017
Kepa K (2021) Cellulose nanofibre and nanopaper:. structure-property-processing relationship and green surface modification
Maiti S, Jayaramudu J, Das K et al (2013) Preparation and characterization of nano-cellulose with new shape from different precursor. Carbohydr Polym 98:562–567. https://doi.org/https://doi.org/10.1016/j.carbpol.2013.06.029
Santucci BS, Bras J, Belgacem MN et al (2016) Evaluation of the effects of chemical composition and refining treatments on the properties of nanofibrillated cellulose films from sugarcane bagasse. Ind Crops Prod 91:238–248. https://doi.org/https://doi.org/10.1016/j.indcrop.2016.07.017
Madsen B, Gamstedt EK (2013) Wood versus Plant Fibers: Similarities and Differences in Composite Applications. Adv Mater Sci Eng 2013:1–14. https://doi.org/10.1155/2013/564346
Hofstetter K, Hellmich C, Eberhardsteiner J (2005) Development and experimental validation of a continuum micromechanics model for the elasticity of wood. Eur J Mech - A/Solids 24:1030–1053. https://doi.org/https://doi.org/10.1016/j.euromechsol.2005.05.006
Fukuzumi H, Saito T, Isogai A (2013) Influence of TEMPO-oxidized cellulose nanofibril length on film properties. Carbohydr Polym 93:172–177. https://doi.org/https://doi.org/10.1016/j.carbpol.2012.04.069
Qing Y, Sabo R, Zhu JY et al (2013) A comparative study of cellulose nanofibrils disintegrated via multiple processing approaches. Carbohydr Polym 97:226–234. https://doi.org/10.1016/j.carbpol.2013.04.086
Pennells J, Cruickshank A, Godwin ID, Martin DJ (2021) Sorghum as a novel biomass for the sustainable production of cellulose nanofibers. Ind Crop Prod 171. https://doi.org/10.1016/j.indcrop.2021.113917
Guimarães JL, Frollini E, da Silva CG et al (2009) Characterization of banana, sugarcane bagasse and sponge gourd fibers of Brazil. Ind Crops Prod 30:407–415. https://doi.org/https://doi.org/10.1016/j.indcrop.2009.07.013
Meng F, Zhang X, Yu W, Zhang Y (2019) Kinetic analysis of cellulose extraction from banana pseudo-stem by liquefaction in polyhydric alcohols. Ind Crops Prod 137:377–385. https://doi.org/https://doi.org/10.1016/j.indcrop.2019.05.025
Mohamed GR, Mahmoud RK, Fahim IS et al (2021) Bio-composite Thermal Insulation Materials Based on Banana Leaves Fibers and Polystyrene: Physical and Thermal Performance. J Nat Fibers 1–16. https://doi.org/10.1080/15440478.2020.1870628
Prager CM, Boelman NT, Eitel JUH et al (2020) A mechanism of expansion: Arctic deciduous shrubs capitalize on warming-induced nutrient availability. Oecologia 192:671–685. https://doi.org/10.1007/s00442-019-04586-8
Aguado R, Moral A, López P et al (2016) Morphological analysis of pulps from orange tree trimmings and its relation to mechanical properties. Meas J Int Meas Confed 93:319–326. https://doi.org/10.1016/j.measurement.2016.06.063

No competing interests reported.

SupplementaryInformationJPE.docx

Download PDF

Editorial decision: Major revision
30 Sep, 2022
Reviews received at journal
19 Sep, 2022
Reviewers agreed at journal
09 Sep, 2022
Reviews received at journal
09 Sep, 2022
Reviewers agreed at journal
29 Aug, 2022
Reviewers invited by journal
29 Aug, 2022
Editor assigned by journal
26 Aug, 2022
Submission checks completed at journal
26 Aug, 2022
First submitted to journal
26 Aug, 2022

You are reading this latest preprint version

Benchmarking the Production of Cellulose Nanofibres: Biomass Feedstock, Mechanical Processing, and Nanopaper Performance

Status:

Version 2

Abstract

Figures

Introduction

Experimental Methodology

Scope

Assumptions

Publications

Results & Discussion

Biochemical Composition

Thermal Stability

Crystallinity

Fibre morphology

Nanopaper mechanical properties

Mechanical Process Benchmarking

Nanopaper mechanical properties

Fibre morphology

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 2