3. Emerging green routes
Reviewing research published between 2015 and 2020 concerning new CNF/CNC preparation routes, Li and co-workers identified 69 studies.19 All new processes included two or three pretreatment steps, followed by chemical and mechanical treatment. Amid the mechanical treatment processes, ultrasonication was found to be dominant. 
Another recent comprehensive review on new extraction methods from cellulosic biowaste, reported that sugarcane bagasse with its high lignocellulose content and loose structure, and algae residues are the biowastes most suited to extract nanocellulose.32
Poor in lignin and available in huge yearly amount, citrus processing waste (CPW) obtained from the citrus juice industry is another ideal cellulose feedstock. The annual production of citrus fruits exceeds 120 million tons. More than half  of the fruit is non-edible and discarded as waste. Alone, the global volume of citrus fruits yearly squeezed at citrus juice plants exceeds 31 million tonnes, 50-60% of which is waste.33
Unfortunately, the routes to citrus nanocellulose starting from CPW based on enzymatic,34 microwave-assisted hydrothermal treatment,35 and acid hydrolysis,36 all present significant technical limitations. For example, the nanocellulose fibrils obtained via multi-step microwave-assisted extraction of dried depectinated orange peel are deeply colored in brown.35 This is due both to caramelized sugars and to the Maillard reaction between sugars and residual proteins at the high temperatures (120°C to 180°C) required for extraction.35
One entirely green route to nanocellulose in CNC form was reported by Huang and co-workers in 2018.37 The team used ball milling of cellulose in water only, followed by centrifugation.  The process quickly affords cellulose disintegration down to elementary fibrils 3-10 nm wide and 120-400 nm long.
To understand the relevance of this discovery, it is instructive to compare it to enzymatic hydrolysis with cellulase enzymes. Applied to eucalyptus cellulose pulp using a new commercial enzymatic complex at cellulose loading of 10% (w/v) and enzyme loading of 10 mg/g,38 or even relying on a cellulolytic enzymatic complex produced on-site by Aspergillus niger followed by sonication,39 the optimal enzymatic process affords a 24.6% yield in CNC (after 96 h of enzymatic hydrolysis of cellulose pulp, followed by 5 min sonication).  
As mentioned above, the TEMPO-mediated carboxylation of cellulose and production of nanocellulose has already been commercialized.26 Numerous practically relevant discoveries using heterogenized TEMPO catalysts showed evidence, on laboratory scale, how shifting the CNF production process from using TEMPO in solution to new generation heterogeneous TEMPO catalysts has substantial benefits in terms of enhanced quality of the resulting CNF and reduced environmental impact.40
Interestingly, Turbak recognized that cavitation was among the “high forces of the homogenizer”2 responsible for cellulose fibrillation using the high pressure homogenizer. More than 40 years after Turbak and co-workers discovery,1 in 2010 Pandit and Pinjari were the first to report the outcomes of acoustic cavitation (AC) and hydrodynamic cavitation (HC) applied to a 1% w/v (0.5 kg in 50 L water) aqueous suspension of 63 mm cellulose microparticles.41